Gut microbiota in brain tumors: An emerging crucial player

Abstract In recent decades, various roles of the gut microbiota in physiological and pathological conditions have been uncovered. Among the many interacting pathways between the host and gut flora, the gut–brain axis has drawn increasing attention and is generally considered a promising way to understand and treat brain tumors, one of the most lethal neoplasms. In this narrative review, we aimed to unveil and dissect the sophisticated mechanisms by which the gut‐brain axis exerts its influence on brain tumors. Furthermore, we summarized the latest research regarding the gastrointestinal microbial landscape and the effect of gut–brain axis malfunction on different brain tumors. Finally, we outlined the ongoing developing approaches of microbial manipulation and their corresponding research related to neuro‐malignancies. Collectively, we recapitulated the advances in gut microbial alterations along with their potential interactive mechanisms in brain tumors and encouraged increased efforts in this area.


| INTRODUC TI ON
Central nervous system (CNS) tumors cause a considerable portion of cancer-related mortality in adults and children. 1 Recent data have demonstrated that the average annual age-adjusted incidence rate of all malignant and non-malignant brain and other CNS tumors was 24.25 per 100,000 people, of which 29.1% are malignant. The most frequently reported type was meningiomas (39.0%), followed by tumors of the pituitary (17.1%), and glioblastoma (14.3%). 2 Previous studies have identified various genetic alterations that may constitute the molecular basis for these diseases; however, therapies targeting these pathways were either unsuccessful or had minimal survival benefits. 3 The gut microorganisms comprise the assembly of commensal bacteria, archaea, fungi, and viruses in the gastrointestinal tract, which is referred to as gut microbiome. 4 Given the diversity among individuals and blurred definition of healthy gut flora, our understanding of the nature and function of the microbiota is still nascent.
Nevertheless, as sequencing technology and the utilization of germfree (GF) mice advance, the gut microbiota has been demonstrated to play a critical role in physiological activities and pathological conditions. 5 The gut-brain axis is the bidirectional communication axis between the CNS and gastrointestinal (GI) tract, involving the gastrointestinal microbiota, enteric nervous system, neuroendocrine mediators, autonomic nervous system, and the CNS. 6 These components function as a multidirectional network that influences biological homeostasis. 7 Exploring the role of the gut-brain axis in the development or progression of brain tumors may offer novel insights into the molecular etiology of these diseases and provide potential treatment targets. 8 We hope that this review will promote further studies in this area.

| MECHANIS MS UNDERLYING THE G UT-B R AIN A XIS
Considerable researches have revealed the intricate networks of the gut-brain axis. Given its substantial influence on the CNS both physically and pathologically, meticulous delineation is warranted. [9][10][11] Herein, we categorized the mechanisms into three chapters (immunological modulation, microbial metabolites-mediated modulation, and direct invasion), as shown in in the initiation, instruction, regulation, and implementation of the immune system. 12 Additionally, nearly 70% of all immune cells, comprising lymphoid, and lymphocytes, are located in the intestinal lymphoid tissue. 13 In the presence of certain gastrointestinal bacteria, dendritic cells mediate the polarization of Th (T helper) cells into Th1 and Th17 cells. 14 Choi et al. concluded that gut microbiota-derived molecules might act as antigens in the brain, inducing intestinal Treg cells to migrate to the brains. 15 These brain-resident T cells can regulate the fetal-to-adult transition of microglia, the primary resident immune cells of the brain. 16 GF animals develop severe immunodeficiency with impaired intestinal lymphoid architecture and immune functions, indicating that the commensal gut microbiota is essential for the activation and function of immune cells. 17,18 The commensal Bacteroides fragilis-derived polysaccharide can influence murine immune system maturation by correcting systemic Th1/Th2 imbalance and T-cell deficiencies in lymphoid tissues. 19 F I G U R E 1 Stratification of the mechanisms involved in the gut-brain axis. Three mechanisms of gut-brain axis were identified, as indicated by the arrows: (1) immunological modulation, (2) microbial metabolites-mediated modulation, and (3) direct invasion. Gastrointestinal interface between microbiota and host is magnified.

| Effect on immune cell function
Peripheral immune cells are influenced by the gut microbiota. GF mice have fewer proinflammatory Th17 cells in the lamina propria of their gastrointestinal tract than specific pathogen-free controls. 20 Furthermore, several bacterial species (e.g., Lactobacilli, Proteobacteria, Clostridium difficile, Enterococci, Bacteroides fragilis) are known to stimulate specific pro-and antiinflammatory immune cell populations. 21 For example, Bacteroides fragilis can induce the differentiation of interleukin-10 (IL-10)-secreting T regulatory cells (Tregs), 22 which impairs anticancer Th1 immunity and promotes glioma progression and aggressiveness. 23 Furthermore, particular microbiota can drive immune suppression, which is commonly observed in malignancies, including glioblastomas. 24 The CNS was once considered an immune-privileged site; however, the permeable nature of the brain-blood barrier (BBB) and the recent discovery of functional lymphatic vasculature indicate that immune cells play a physiological and pathological role in the CNS. 25,26 In addition to glial cells, resident immune cells (e.g., macrophages, CD8+ T cells, Tregs, and other CD4+ Th cells) actively engage in innate and adaptive immune responses. 27 Among these subsets, microglia, which migrate to the brain during development and self-renew until adulthood, are wellstudied. 28 Apart from protecting the brain from pathological conditions, microglia can regulate synaptic transmission, synaptic pruning, and neuronal circuit formation. 29,30 The microbiome influences microglial properties and function. In GF mice, microglia exhibit altered morphology and transcriptomic profiles, accompanied by inhibition of the maturation state, as indicated by increased immature microglia in the brain cortex. 31 Similarly, antibiotic-treated mice present with increased naïve microglia, although the total number of microglia remains unchanged. 31,32 Immature microglia in GF mice tend to display impaired immune activation and responses to challenges, accompanied by reduced inflammatory factors and inhibited innate immune signaling. 8,31 Remarkably, the microbial deficiency-induced immunosuppressive phenotype in GF mice can be reversed by postnatal administration of microbial short chain fatty acids (SCFAs), suggesting that certain microbial species and metabolites are involved in the maturation and homeostasis of microglia. One study revealed that increased intestinal Ruminococcus induced the antiinflammatory polarization of microglia and attenuated neuronal degeneration and necrosis caused by epilepsy in LPS-induced mice model. 33 Moreover, GPR43 in innate immune cells mediates inflammatory responses by interacting with SCFAs, and mice with GPR43 deficits display severe morphological defects in microglia, similar to the defects observed in GF mice. 31 Since GPR43 is intimately related to inflammasomes, the interplay among them may contribute to maintaining microgliamediated immunological homeostasis.
Chronic inflammation and an imbalanced immune environment have long been associated with oncogenesis and tumor progression. For example, a high antiinflammatory/pro-inflammatory ratio is found in glioblastoma (GBM) and correlates with poor survival, and antiinflammatory cells can secrete IL-10, epidermal growth factor (EGF), and vascular endothelial growth factor (VEGF). 34 Interestingly, gut microbiota-derived metabolites can drive microglial polarization towards antiinflammatory phenotype 35 and Tregs can be promoted by diet-derived SCFAs. 36 Tregs can exacerbate the immunosuppressive microenvironment by producing IL-10 and transforming growth factorβ (TGFβ). 17 Accordingly, Treg levels are correlated with brain tumor grade, and in vivo depletion of Tregs improves survival in mice. 37 Besides, previous research showed that microbiome-mediated inflammatory reactions exist in central and peripheral organs, probably serving as a disease biomarker and a therapeutic target for stroke. 38 With respect to brain tumors, inspections of the inflammatory response in peripheral organs and the potential involvement of gut microbiome should be conducted in future research.
The nuclear factor kappa-B (NF-κB) pathway plays a substantial role in the production of inflammatory cytokines such as IL-6 and IL-8, thereby regulating the tumor microenvironment. In cancer cells, abnormalities of the NF-κB pathway activate survival genes.
Specifically, IL-6 can induce NF-κB in GBM, resulting in the activation of signal transducer and activator of transcription 3 (STAT3) and increased tumor aggressiveness. 39 High NF-κB expression is associated with poor survival in mesenchymal GBM. 40 Given that circulating SCFAs can enter the CNS especially when the BBB is disrupted, 41 it is reasonable to speculate that gut microbiota-related metabolites, such as SCFAs, could influence the NF-κB function of cancer and immune cells inside the brain.
In conclusion, the gut flora is critical for the establishment of

| Microbiota-derived metabolites and products
Microbial-derived metabolites and products are major contributors to the microbiota-gut-brain axis, exerting their effects primarily via receptor-mediated interactions in host tissues or cells.
Among them, SCFAs and endogenous tryptophan are the most well-documented metabolites. SCFAs, derived from microbial decomposition of carbohydrates, are related to glucose homeostasis, mucosal serotonin release, lymphocyte function, learning, and memory acquisition via the preservation of brain integrity. [42][43][44] As mentioned above, some circulating SCFAs can enter the CNS.
GF mice have been documented as having increased BBB permeability. Recolonization of the same mice with SCFA-producing bacteria recovered BBB integrity, which further supports the role of SCFAs in CNS homeostasis. 45 Mechanistically, studies have reported interactions between SCFAs and G-protein coupled receptors (GPR), such as GPR41 and GPR43, 46 potentially serving as a connection between SCFA and the CNS. On the other hand, SCFAs, as illustrated above, can modulate the immune system.
Mechanistically, SCFAs can influence metabolism and inhibition of histone deacetylases (HDACs), thereby modifying histone acetylation. 47 The suppression of histone deacetylase activity inhibits the activities of tumor necrosis factorα (TNFα) and NF-κB. 48 Further, the chemokine pattern of dendritic cells (DC) is regulated by SCFAs such as acetate and propionate, 49 and neutrophils, B cells, and T cells can also be modulated by SCFAs. 50 The gut microbiota can directly convert tryptophan into metabolites with immunomodulatory capacity (e.g., indolic acids, indole, and tryptamines). 51  The host metabolic status affects the microbial composition, and microbial metabolites can induce epigenetic modifications 55 ; for example, Brecibacteruim spp regulates the balance of ⍺-ketoglutarate (⍺KG) and glutamate. 56 The IDH1/2 mutation in glioma inhibits multiple ⍺KG-dependent enzymes, leading to aberrant DNA methylation. Thus, the gut microbiota and its derived metabolites might modify the epigenetic status, contributing to glioma progression. 57 Microbiota-derived SCFAs also regulate neuronal, glial, and tumoral epigenetics. Acetate participates in acetyl-CoA production and subsequently affects the acetylation of rapamycin-insensitive companion of mTOR (RICTOR), potentially promoting glioma tumorigenesis. 58 Besides, accumulating evidence reveals that the circadian rhythm has an impact on glioma pathophysiology, and the internal characteristics concerning the circadian clock in glioma involve stemness, metabolism, radiotherapy sensitivity, and chemotherapy sensitivity. 59 Meanwhile, alterations of microbiome-derived metabolites such as ergothioneine were observed to be associated with circadian function. 60 Yet, it remains to be addressed whether the gut microbiota may affect tumorigenesis through disruption of circadian rhythm.
In addition to microbial metabolites, microbiota-derived products also play major roles in the microbiota-gut-brain axis through interactions with toll-like receptors (TLRs) in the enteric nervous system (ENS) and CNS. For example, lipopolysaccharide (LPS), released by gram-negative bacteria are recognized by TLRs expressed on CNS microglia, subsequently inducing proinflammatory cytokine production and proliferation. 61 Remarkably, this immune response has been found to result in neuroinflammation, microglial activation, and neuronal cell loss, triggering cognitive impairments, anxiety, and depression. 62 Polysaccharide A, another common microbial product, is secreted by B. fragilis and can be recognized by TLRs, 5 provoking a protective CNS antiinflammation response. 63

| Neuroactive molecules/neurotransmitters
Aside from metabolites and products, bacteria produce a variety of major neurotransmitters such as dopamine, noradrenaline, serotonin, gamma-aminobutyric acid (GABA), acetylcholine, histamine, and tryptophan metabolites, sustaining the gut-brain axis. 64,65 Serotonin, GABA and tryptophan metabolites cannot directly affect the CNS, owing to the inability to cross the BBB but can influence the nervous system by interacting with cells in the ENS. 46 Gut microbes also regulate the abundance of host neurotransmitters such as dopamine, norepinephrine, and serotonin. 66 Studies in GF mice have demonstrated that the absence of microbiota modulates the neurotransmitter turnover in the CNS and ENS. For example, GF mice display increased turnover rates for dopamine and norepinephrine in the brain. 67 A combination of 46 Clostridium species was shown to restore dopamine and norepinephrine levels in the cecal lumen of GF mice. 68 However, whether this effect is due to direct production of neurotransmitters or results from the modulation of host production remains unclear. Similarly, gavage of Enterococcus faecium and Lactobacillus rhamnosus in young mice increased brain dopamine levels. 69 Moreover, the gut microbiota can produce certain cofactors such as tetrahydrobiopterin (BH4), thereby promoting tyrosine hydroxylase (TH) activity in the brain, consequently leading to increased dopamine levels. 70 GF mice display an increased turnover rate of serotonin in the brain, 67 and substantially decreased peripheral serotonin compared with control mice. 71 However, studies investigating the influence of microbiota on serotonin using GF mice obtained inconsistent results; one reported increased turnover and unchanged levels of serotonin in the striatum, 67 while another showed increased levels of both serotonin and 5-hydroxyindoleacetic acid (5-HIAA) in the hippocampal regions. 72 Correspondingly, supplementation of Lactobacillus plantarum to GF mice significantly increased serotonin and dopamine levels in the striatum. 73 The gut microbiota also affects the circulating GABA, with GF animals demonstrating reduced GABA levels in the gut lumen and serum. 74 The administration of Lactobacillus rhamnosus increased brain GABA level 75 and alleviated depressive and anxiety-like behaviors, along with alterations in the transcriptomic profile related to cerebral GABA receptors. 76 Furthermore, a recent study showed that Lactobacillus rhamnosus JB-1 supplementation increased brain GABA and glutamate/glutamine levels, indicating that the gut microbiota might regulate glutamate production pathways in the brain. 75,77 Despite the interactions mentioned above, how these neuroactive molecules affect the CNS remains unclear.

| Gut hormones
Enteroendocrine cells (EECs), which constitute the interface between the gut microbiota and host, are modulated by the diversity and composition of the gut bacteria, and thus generate fluctuations in secreted hormones, consequently participating in gut-brain crosstalk. 78 It has been shown that bacterial metabolites (e.g., LPS, SCFAs, and tryptophan) can stimulate EECs of the gut epithelium to produce neuropeptides (peptide YY, neuropeptide Y, cholecystokinin, glucagon-like peptide (GLP)-1 & 2, and substance P). These neuropeptides enter the bloodstream to reach local receptors and influence ENS neurons and the vagal nerve. 79 Additionally, the gut microbiota may modulate the production of neuromodulators, as evidenced by increased levels of GLP-1, corticosterone, and adrenocorticosterone in GF mice. 66 Collectively, these results suggest that the gut microbiota regulate the release of gut hormones from EECs through metabolites or bacterial components.
Notably, gut hormones also affect the microbiota, altering the microbial profile of the gut. 80 Future studies are warranted to determine the cause and effect of the gut hormone-gut microbiota association.

| Direct invasion
Although the detailed mechanisms remain poorly understood, direct microbial invasion beyond the BBB is possible. Regardless of heterogeneous requirements, common prerequisites for invasion include asymptomatic colonization of host mucosal surfaces, sustained survival in the bloodstream, resistance or escape from immune responses, and breaching the BBB through various pathways. 81 During tumor development, disorganized and leaky vasculature may allow entry of circulating bacteria, for which the immunosuppressed environment may provide a refuge. 82 Indeed, recent studies have confirmed the presence of intratumor microbiome in brain tumors 83,84 ; however, the detailed mechanisms underlying this direct hostbacteria interaction remain to be clarified.

| AB ERR ANT MICROB IAL L ANDSC APE S AND DYS REG UL ATED MI CROB I OTA-G UT-B R AIN A XIS IN B R AIN TUMOR S
Studies focusing on brain lesions specific alterations in gut microbiota have emerged in recent years, with most focusing on the three most common brain tumors, glioma, pituitary adenoma, and meningioma. The aforementioned changes have been summarized and described in Figure 2, and detailed variations are listed in Table 1.

| Glioma
Glioma is the most common primary malignant tumor of the CNS in Although state-of-art therapies such as immunotherapy are emerging, many barriers remained unsolved. 85 Hence, further research expanding the field of tumor cell-host interactions to identify potential therapeutic targets is required.

| Dysregulated gut-brain axis components
The microbiota has been shown to influence the status of natural immu-

| Dysregulated gut-brain axis components
The same study reported the downregulation of genes related to the metabolism of D-glutamine and D-glutamate, nucleotide excision repair (NER), and endocytosis, likely affecting inflammatory and immune responses of the meningioma patients. 89 Given that several studies have suggested a close interaction between the gut-brain axis and hypothalamus-pituitary-adrenal axis, [104][105][106] it is rational to speculate that the gut microbiota may play a significant role in the pathological process of Cushing's disease (a subtype of pituitary adenomas). Further research is needed to clarify this hypothesis.

| Others
The scope of neuro-oncology is much wider than that of the brain tumor types covered in this review, with CNS lymphoma, neurofibroma, metastatic tumors, etc. However, there are currently no studies exploring their specific gut microbiota alterations. Hence, we call for future studies to inspect these disease-specific microbial profiles.

| Fecal microbiota transplantation (FMT)
FMT can ameliorate neurological diseases, such as seizures. 107 Nevertheless, merits of FMT need to be tested further because of

| Probiotics
Probiotics are live microorganisms which exert health benefits when consumed or applied to the body. They are found in yogurt, fermented foods, and dietary supplements. Modified probiotics, genetically engineered to maximize the beneficial effects, have been introduced and included in more traditional pharmaceutical routes. 44,113,114 In recent years, probiotic administration has been shown to regulate neuroendocrine homeostasis. For example, in a randomized, performance and changes in brain activity on electroencephalography. 104 In conclusion, manipulation of the gut microbiota regulates the neuroendocrine system, implying a potential application in the management of endocrinopathies in pituitary tumors. However, heterogeneity in composition, stability, and authenticity, lack of consensus on dosage, duration, and specific strains to use, and host colonization resistance must be resolved before expanding the application of probiotic-based therapies to brain tumors.

| Antibiotics
Manipulating gut microbiota through antibiotics also deserves ad-

| Bacteriophages
Bacteriophages, a class of prokaryotic viruses that have evolved specifically to infect and replicate within bacteria, can be designed to specifically target detrimental bacteria. 118 One study showed that a bioinorganic hybrid bacteriophage could target and kill

| Engineered microbiomes
The application of genetically engineered microbes, which have higher specificity than FMT has shown promising antitumor effects in preclinical models. 120 Recent studies harnessed the targeting abilities of certain strains (Salmonella and Clostridium) as carriers for the delivery and induction of immune stimulants to delay tumor growth and metastasis in melanoma, renal cell carcinoma, and osteosarcoma. 121 In addition, SYNB1891, designed based on the biology of Escherichia coli, has been modified to express the STING agonist cyclic adenosine diphosphate ribose (CADPR), thereby stimulating expression of IFNs and achieving antitumor effects in tumor-bearing mouse models (melanoma, lymphoma, mammary carcinoma, and colon carcinoma). 122 Further, another study showed that the engineered Escherichia coli Nissle 1917 strain could colonize tumor sites, converting ammonia to L-arginine, thereby increasing the intracellular concentration of L-arginine in subcutaneous colonic cancer cells, which further triggered intra-tumoral infiltration of CD4+ and CD8+ T cells, thus exerting synergistic antitumor effects when combined with anti-PD-L1. 123 Collectively, engineered microbiomes show potential as an adjuvant therapy against a few kinds of tumors, and future research to evaluate their potency in brain tumors is merited.

| Diet
Diet also constitutes a significant component of microbiota manipulation. Accumulating evidence indicates that changes to dietary regimens can rapidly alter gut microbial profiles. 124 Meanwhile, dietary intervention also plays a role in the management of brain tumors. For example, ketogenic diets, comprising high-fat, lowcarbohydrate, and adequate-protein, can reduce seizure frequency by increasing GABA and glutamate levels in epilepsy patients. 125 These findings are generalizable to diseases such as gliomas, with one pre-clinical study detecting massive tumor cell death upon concurrent treatment of mice with calorically restricted ketogenic diet and a glutamine antagonist, without obvious host toxicity. 126 Considering the close interaction between gut microbiota and diet patterns, the gut microbiota might serve as an intermediate between dietary intervention and brain tumors, although further research is required.

| CON CLUS ION
As technology develops, our knowledge of host-microbiota interactions is rapidly expanding. Among these critical interactions, the gut-brain axis has attracted interest from and offered novel insights to neuro-oncologists. The mechanisms underlying the gut-brain axis in brain tumors include immunological modulation, microbial metabolite-mediated modulation, and direct in-

CO N FLI C T O F I NTE R E S T
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.

CO N S E NT FO R PU B LI C ATI O N
Not applicable.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analyzed in this study.