Microbiota-modulated spermine promotes brain metastasis in non-small cell lung cancer by regulating microglia M2 polarization via the STAT3 pathway

Microbiota-modulated spermine promotes brain metastasis in non-small cell lung cancer by regulating microglia M2 polarization via the STAT3 pathway | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Microbiota-modulated spermine promotes brain metastasis in non-small cell lung cancer by regulating microglia M2 polarization via the STAT3 pathway huanhuan li, Lichao Liu, Yawen Bin, Hao Zeng, Jiaojiao Wang, Ruiguang Zhang, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-2259805/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Brain metastasis (BM) is associated with high mortality in patients with non-small cell lung cancer (NSCLC). Alterations in the gut microbiota have been implicated in modulation of brain disorders through the gut-brain-axis (GBA). However, the underlying mechanism by which the gut microbiota affects the development of BM in NSCLC remains largely unknown. In patients, we identified 16 genera of differential bacteria positively or negatively correlated with BM in NSCLC patients, as represented by Klebsiella , unclassified_f_Enterobacteriaceae and Alistipes by 16S rRNA gene sequencing. In addition, untargeted metabolomics (LC-MS/MS and GC/MS) identified 76 metabolites, that were associated with BM. The combination of intestinimonas and spermine was considered a potential marker for the diagnosis of BM in NSCLC. Moreover, the plasma metabolite spermine enhanced BM by promoting M2 polarization of microglia via activation of the signal transducer and activator of transcription 3 (STAT3) signaling pathway in vivo and in vitro , suppressing innate immune function, which in turn promoted tumor progression. Overall, our study demonstrated the composition of both the gut microbiota and metabolites changed significantly between groups, and revealed that metabolite spermine promotes BM by skewing the polarity of M2 microglia by activating STAT3 signals. Our results provide a novel perspective regarding host-gut microbiota interplay in BM of NSCLC and highlight a potential risk of spermine in promoting BM. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Background Brain metastasis (BM) is a leading cause of mortality worldwide in patients with advanced non-small cell lung cancer (NSCLC) [ 1 , 2 ]. Accumulating evidence suggests that the gut microbiota is a critical environmental factor associated with the development of a growing number of diseases, including the progression of cancer, inflammation, and metabolic syndrome [ 3 , 4 ]. Studies have revealed that the gut microbiota affects brain function and behavior via the gut-brain axis (GBA), which involves bidirectional communication between the gut microbiome and the brain through neural, endocrine, and immune pathways [ 5 , 6 ]. Metabolites are considered important mediators between the gut microbiota and brain by entering the bloodstream and exerting functional effects [ 7 ]. For example, a study using a multiple sclerosis animal model indicated that some microbial-regulated tryptophan metabolites bind to aryl hydrocarbon receptor, to suppress brain inflammation [ 8 ]. Microbial-derived molecules such as short-chain fatty acids can cross into the brain and alleviate chronic psychosocial stress [ 9 ]. We hypothesized that the gut microbiota can affect the initiation or progression of brain tumors by modulating the circulation metabolism of the host. However, the profiles of the gut microbial community and host metabolites in NSCLC patients with BM have not been systematically characterized. Furthermore, multiple studies have suggested that activation of microglia promotes brain tumor progression [ 10 , 11 ]. Microglia generally polarize into a classical activated microglia phenotype (M1) or an alternative phenotype (M2). The M1 phenotype, is characterized by the production of a variety of pro-inflammatory cytokines, that have anti-tumor effects. By contrast, the M2 microglia function as pro-tumor factors by initiating immunosuppression [ 12 , 13 ]. Notably, the gut microbiome promotes the maturation of microglia and facilitates their function in the brain via the GBA [ 14 , 15 ]. It has been postulated that metabolites regulated by the gut microbiota modulate the function of microglia in the progression of BM in NSCLC, but the mechanism remains to be elucidated. To address these gaps in knowledge, we analyzed the gut microbial characteristics of 67 NSCLC patients with BM (n = 32) and without BM (NBM, n = 35) via 16S rRNA gene sequencing. In addition, untargeted metabolomics analyses of plasms samples, liquid chromatography-tandem mass spectrometry (LC-MS/MS) and gas chromatography-mass spectrometry (GC-MS) were performed to determine the metabolomic profiles of 67 NSCLC patients. We envisaged this study as a systematic and comprehensive interrogation of the microbiome and metabolome of NSCLC patients with BM. The objectives of the study were to characterize microbial diversity and metabolite abundance and decipher the association between BM in NSCLC with the gut microbiota and host metabolites. Notably, we found that the metabolite spermine promotes the polarization of microglia to the M2 phenotype, resulting in the development of BM in NSCLC model mice. Methods Study design and population A total of 70 NSCLC patients diagnosed with BM (n = 34) and NBM (n = 36) by pathological examinations at Wuhan Union Hospital (Wuhan, China) were initially recruited for inclusion in this study. Written informed consent was obtained from all participating patients. The Institutional Revier Board of Huazhong University of Science and Technology approved the study. General clinical data were recorded for all participants, including age, gender, body mass index (BMI), histories of smoking, and serum carcinoembryonic antigen (CEA). All patients were from similar geographic areas and had similar eating habits. Three subjects overall were excluded from the subsequent analyses due to low reads on sequencing of stool samples (n = 2), or hemolysis of plasma sample (n = 1). Finally, 67 cases were subjected to 16S rRNA gene sequencing and untargeted metabolomics analysis. All samples were collected in accordance with standard operating procedures. A flow chart illustrating the enrollment and analysis processes is shown in Additional file 1: Figure S1. Dna Extraction And 16s Rrna Gene Sequencing DNA was extracted from fecal samples using an E.Z.N.A.® soil DNA kit (Omega Bio-tek, Norcross, GA, USA). PCR amplification was performed using the primers 338F (5'-ACTCCTACGGGAGGCAGCAG-3'), 806R (5'-GGACTACHVGGGTWCTAAT-3’) directionally targeting the V3 and V4 regions of the 16S rRNA gene. DNA recycling and purification were performed using an AxyPrep DNA Gel Extraction kit (Axygen Biosciences, Union City, CA, USA). An Illumina Miseq PE300 platform was used for sequencing according to the manufacturer’s specifications. 16s Rrna Gene Sequencing Data Analysis Sequences exhibiting > 97% similarity thresholds were allocated to one operational taxonomic unit (OTU), chimeric sequences were filtered and the alignment threshold was set to 70% for taxonomic identification of species. Alpha diversity was determined to assess the complexity of species diversity for each sample. Beta diversity calculations were subjected to principal coordinate analysis (PCoA) to assess the diversity in samples from different groups in terms of species complexity. The Wilcoxon rank-sum test was used to compare bacterial abundance and diversity. Heat maps were constructed based on the nonparametric Wilcoxon test ( P < 0.05, q < 0.1) at the genus level. Linear discriminant analysis (LDA) coupled with effect size (LEfSe) was applied to evaluate differentially abundant taxa. The data were analyzed on the free online platform of Majorbio Cloud Platform ( www.majorbio.com ), from Shanghai Majorbio Bio-pharm Technology Co.,Ltd. Untargeted Metabolomics Analysis A total of 67 plasma samples were collected from NSCLC patients and sequenced at Novogene (Beijing, China). LC-MS/MS analyses were conducted using a Vanquish UHPLC system (Thermo Fisher) and Orbitrap Q Exactive series mass spectrometer (Thermo Fisher). The exactive series mass spectrometer was operated in positive/negative polarity mode. Metabolite data were analyzed using Compound Discoverer 3.1 (CD3.1, Thermo Fisher). Metabolites were identified based on peak intensities, normalized to the total spectral intensity. Three databases (mzCloud, mzVault and MassList) were used to obtain accurate qualitative and relative quantitation results. Statistical analyses were performed using the statistical software R (R version R-3.4.3), Python (Python 2.7.6 version), and CentOS (CentOS release 6.6). An Agilent 7890 gas chromatograph system coupled with a Pegasus HT time-of-flight mass spectrometer was used for GC-MS analyses. Untargeted Metabolomics Data Analysis LC-MS/MS: KEGG ( http://www.genome.jp/kegg/ ), HMDB ( http://www.hmdb.ca/ ) and Lipidmaps ( http://www.lipidmaps.org/ ) database, were used to annotate the metabolites. Principal components analysis (PCA) and partial least squares discriminant analysis (PLS-DA) were performed using metaX software. Differential metabolites based on by using variable importance in projection (VIP) score > 1 and P < 0.05, and fold- change (FC) ≥ 2 or ≤ 0.5. The functions of these metabolites and metabolic pathways were analyzed using the KEGG database. GC-MS: Chroma TOF 4.3X software (LECO Corp) and the LECO-Fiehn Rtx5 database were used for raw data analysis. Mass spectrum matches and retention index matches were both considered in metabolite identification. Cell Culture And Reagents Murine Lewis lung cancer (LLC) cells were purchased from the ATCC. Mouse microglia BV2 cell were obtained from Procell Life Science & Technology Co., Ltd (Wuhan, China). All cell lines were cultured in DMEM supplemented with 10% FBS at 37°C under 5% CO 2 . BV2 cells were treated with spermine (10 µM, #S3948, Selleck Chemicals, China) alone or plus the STAT3 inhibitor stattic (0.5 µM, #HY-13818, Sigma-Aldrich, St. Louis, MO, USA) for 24 h to observe changes in the BV2 phenotype. Study Animals And Bm Model Six-week-old female mice (C57BL/6J; Beijing Vital River Laboratory Animal Technology Co. Ltd, China) were used. The BM model was described previously [ 16 ]. LLC cells labeled with luciferase (LLC-Luc, 4 × 10 5 in 0.1 mL PBS) were slowly injected into the intracarotid artery of mice. Two weeks later, bioluminescence imaging was performed to observe the growth of brain tumors using an IVIS Lumina imaging system (In Vivo FX PRO, Bruker Corp.) after mice were injected with D-luciferin intraperitoneally (150 mg/kg, Goldbio St. Louis, MO, USA). BM mice received spermine (10 mg/kg, Selleck) 1 week after injection of tumor cells, by intraperitoneal injection every 3 days until the end of the experiment. Clodronate liposomes (F70101C-A-2, FormuMax, Dakewe Biotechnology Co., Ltd, China) was injected intracerebroventricularly using a stereotactic apparatus (10 µL/mouse). All animal study procedures followed the guidelines of the Institutional Animal Care Committee of Tongji Medical College, Huazhong University of Science and Technology, China. Real-time Quantitative Pcr Total RNA samples were prepared using a total RNA extraction kit (Omega Bio-Tek, China). A PrimeScript RT reagent kit (Takara, Beijing, China) was uesd for reverse transcription. For real-time RT-PCR, SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) and primers (Additional file 2: Table S1) were used, with transcript levels determined according to the 2 −ΔΔCt method. Western Blotting Protein was extracted from tumor cells or tissues using 1× Cell Lysis Buffer (Promega). Western blotting was performed as described previously [ 16 ], using antibodies specific for STAT3 (ab68153, Abcam), p-STAT3 (ab76315, Abcam), iNOS (inducible nitric oxide synthase) (ab178945, Abcam), Ym-1 (chitinase-like 3) (ab192029, Abcam), and GAPDH (AC002, ABclonal, China). Flow Cytometry To assay the populations of microglia in the brain, mouse brains were cut into small pieces, kept in PBS on ice, minced, and digested for 1 h with 0.5 mg/mL collagenase type IV and DNase I 20 U/mL at 37°C. The details of the protocol mainly based on reference[ 17 ]. The cells were separated by Percoll (GE Healthcare Life Sciences) using a density gradient (70%, 37%, 30%), and stained with FITC-conjugated anti-CD11b antibodies (101206), Zombie NIR™ Fixable Viability kit (423106), CD45 Brilliant Violet 510 (109837), CD86 PE (159204), CD206 Brilliant Violet 421 (141717). These antibodies were purchased from BioLegend, San Diego, CA. Immunofluorescence Mice were anesthetized to obtain brain tissue, then processed and embedded in paraffin, the details of the protocol have been described previously[ 18 ]. Brain tissue slides were incubated with the primary antibodies anti-arginase-1 (Arg-1) (1:200, Proteintech, Wuhan, China) or anti-ionized calcium-binding adapter molecule-1 (Iba1; 1:200, Abcam) overnight at 4℃, followed by appropriate secondary antibody conjugated with fluorescent dye Alexa Fluor 488 or Alexa Fluor 594 (1:1,000, Invitrogen). Sections were counterstained with 4,6-diamidino-2-phenylindole (DAPI) (10 µg/mL, Sigma-Aldrich) and then scanned with a digital camera and imaging software (3DHISTECH Ltd., Hungary). Phagocytosis Assay BV2 cells were treated with spermine (10 µM) alone or with the STAT3 inhibitor stattic (0.5 µM) for 24 h to observe the changes in BV2 phagocytic ability. Phagocytosis assay were performed as previously described[ 19 ]. The BV2 cells were observed under a confocal microscope (Olympus, Tokyo, Japan) and analyzed using flow cytometry (Becton Dickinson, Franklin Lakes, NJ, USA). Statistical Analyses The predicted functional composition profiles were collapsed into level-3 KEGG pathways based on 16S rRNA gene sequences using PICRUSt. Correlations between the clinical parameters were calculated using Spearman rank correlation and presented using a heatmap. PCA was performed to examine intrinsic clusters within the metabolomics data. A 95% confidence interval was used as the threshold to identify potential outliers in all samples. PCoA was used to compare the gut microbiome profiles between the two groups using ANOSIM based on Bray-Curtis distance. A correlation matrix between the plasma metabolites and gut bacterial species was generated using Pearson’s correlation coefficient. Statistical analyses were carried out using SPSS version 18 (SPSS, Chicago, IL, USA). All continuous variables such as bacterial α-diversity, age, and body mass index (BMI) was presented as the mean ± SEM, unless otherwise indicated, and difference between groups were compared using the student’s t -test. The statistical significance level was set at P < 0.05. Results Characteristics of the patients To investigate the gut microbiota and metabolites in BM and NBM patients, a total of 67 subjects (BM, n = 32 and NBM, n = 35) remained for the final statistical analyses (Additional file 1, Figure S1). The common clinical characteristics of these 67 NSCLC patients are shown in Additional file 2, Table S2. No difference was observed in terms of sex, age, BMI, smoking history, or CEA levels between BM and NBM patients ( P > 0.05). These data suggest that a more valuable biomarker for predicting the occurrence of BM in NSCLC patients is needed. Changes In The Gut Microbiome Between The Bm And Nbm Group To identify the gut microbiome features in NSCLC patients with BM or NBM, we investigated the composition of the gut microbiome in the 67 final patients based on 16S rRNA gene sequences. The species rarefaction curve for all samples supported the adequacy of the sampling efforts (Additional file 3: Figure S2a). No significant differences were observed in the richness and diversity of the gut microbiota between the NBM and BM group as measured using the Sob index ( P = 0.4037) or Shannon index (P = 0.3433) at the genus level (Fig. 1a-b). Venn diagram was constructed to show the overlapping OTU data of the two groups, and revealed that 1840 of the 2978 OTUs were common in the BM and NBM patient samples (Additional file 3: Figure S2b). Furthermore, to assess the overall structure of the gut microbiota, the beta diversity in microbial composition was found no obvious differences in the BM group with the NBM group (P = 0.311) (Fig. 1c). As gut dysbiosis is reportedly involved in the progression of brain tumors [ 5 ], four phyla were examined to assess the relative proportions of dominant taxa using sunburst plots (Additional file 3: Figure S2c). The Firmicutes/Bacteroidetes ratio (F/B) (4.36 verse 4.13) and the content of Proteobacteria (14.9% verse 9.51%), were higher in the NBM group compared with the BM group, indicating the significantly gut dysbiosis in NBM and BM groups[ 20 ]. Furthermore, to explore the correlation of microbial abundance with common clinical indexes, sex, smoking, BMI and age were included in a Spearman’s correlation analysis. The results showed that sex and smoking were significantly correlated with some bacteria genera, whereas BMI and age exhibited no remarkable correlation of bacterial abundance (Additional file 3: Figure S2d). The top 20 genera with significant differences were exhibited in two groups, Alistipes 、 Klebsiella and unclassified_f_Enterobacteriaceae were significantly more abundant in patients with BM compared with NBM patients, however, the levels of Blautia and Megamonas were decreased in the BM group ( P 2). Notably, we identified 18 discriminatory genera as key discriminants in the two groups, unclassified_f_Enterobacteriaceae and Blautia were in greater abundance in the BM group and NBM group, respectively. These data suggest that the differentially abundant microbiota in BM and NBM patients could potentially serve as a biomarker of BM in NSCLC. To characterize the functional changes in the gut microbiota of NSCLC patients with BM, we predicted the functional composition profiles using 16S rRNA gene sequencing data analyzed using PICRUSt. COG function classification statistics box charts indicated that the pathways of carbohydrate transport and metabolism, and amino acid transport and metabolism were the top two most abundant pathways (Additional file 3: Figure S2e). A total of 239 KEGG (level-3) pathways involved in BM and NBM groups (Additional file 2: Table S3). Amino acid metabolism is reportedly as an important factor in nucleotide production and tumor cell proliferation in highly aggressive BM [ 21 ]. Thus, we hypothesized that gut dysbiosis is correlated with BM development via amino acid metabolism pathways. Changes In The Plasma Metabolome Between The Bm And Nbm Groups As circulating metabolites considered pivotal mediators of communication between the gut microbiota and the brain [ 22 ], the untargeted metabolomic were performed via LC-MS/MS and GC-MS to clarify the microbe-host interactions in BM of NSCLC. The LC-MS/MS metabolite profiling data showed that BM patients were significantly separated from NBM patients in the PCA (Additional file 4: Figure S3a) and PLS-DA models (neg: R2Y 0.60; Q2Y 0.40; pos: R2Y0.65; Q2Y 0.31) (Fig. 2a). In the model of GC-MS data, the two groups were also clearly distinguishable in the PCA (Additional file 4: Figure S3b) and PLS-DA score plot (R2Y 0.84; Q2Y 0.78) (Fig. 2b). In addition, PLS-DA validated model performed in Additional file 4: Figure S3c-d, in the LC-MS/MS model and GC-MS model, with R2Y is greater than Q2Y. These data confirmed that the plasma metabolites differed significantly between the BM and NBM groups. Finally, we identified 62 differed metabolites in the LC-MS/MS models, with 26 were upregulated, and 36 downregulated in the BM group (Fig. 2c). Also, in the GC-MS model, 14 metabolites differed between groups, including 11 upregulated and 3 downregulated in the BM group (Fig. 2d). To discriminate BM from NBM patients via the potential plasma metabolites, the clinical significance of different metabolites was calculated using random forest analysis (Additional file 2: Table S4). The KEGG database was used to analyze the pathways of the metabolites that differed between BM and NBM patients (Fig. 2e-f). The main pathways that associated with BM included biosynthesis of secondary metabolites, glutathione metabolism, arginine and proline metabolism in the LC-MS/MS model and oxidative phosphorylation in the GC-MS model. Taken together, these results suggest that the metabolite profiles of BM patients differed significantly from those of NBM patients. Therefore, we hypothesize that host metabolites are modulated by the gut microbiota to regulate the development of BM in NSCLC via an as yet unidentified pathway. Multi-omic network analysis reveals the relationship between the gut microbiota and serum metabolites in BM of NSCLC patients To explore the functional correlation between gut microbiota changes and plasma metabolite perturbations in BM of NSCLC patients, the top 16 differential genera and top 20 metabolites were selected for matrix analysis in the model of LC-MS/MS data (Fig. 3a-b and Additional file 2: Tables S5-6), and top 14 metabolites in the model of GC-MS data (Fig. 3c and Additional file 2: Table S7). The results showed that gut-microbiota was positively or negatively correlated with several host-metabolites in NSCLC patients. Pearson correlation coefficients were calculated, with |rho | ≥ 0.4 and P ≤ 0.05 to explore the correlation of metabolites and gut bacteria. As shown in Fig. 3d, metabolite Com_211_pos was positively correlated with Intestinimonas , whereas metabolite Com_4055_pos was negatively correlated with Holdemania , and metabolite Com_867_neg was positively correlated with Anaerotruncus in the LC-MS/MS model. In the model of GC-MS, metabolite Com_206 was positively correlated with Holdemania (Fig. 3e), whereas metabolite Com_113 was positively correlated with and Anaerotruncus and [Eubacterium]_nodatum_group (Additional file 4: Figure S3e). Moreover, arachidic acid (Com_867_neg), spermine (Com_211_pos) and 3-methyl glutaric acid (Com_206) were selected with higher abundances in the BM group, with the corresponding AUC values for the receiver operating characteristic (ROC) curves were 0.746, and 0.730, and 0.846, respectively (Fig. 4a), indicating that each of these metabolites could be used as a potential signature of BM progression in NSCLC. Furthermore, the ROC curve analysis indicated that the combination of the three metabolites (AUC, 0.9232; p < 0.0001) was significantly associated with BM in NSCLC (Fig. 4b), suggesting that a three-metabolite model could accurately distinguish BM patients from NBM controls. Notably, the metabolite spermine functions in arginine and proline metabolism, consistent with the pathway predicted by the 16S rRNA gene profiling data. Therefore, spermine was deemed an important potential metabolite involved in the progression of BM in NSCLC. Spermine was positively correlated with Intestinimonas , and that the combination of the AUC value was 0.7525, which indicated the biomarker are predictive ( P = 0.0004) (Fig. 4c). We hypothesized that the gut microbiota affects BM progression by interacting with certain host metabolites, metabolite spermine was modulated by Intestinimonas potentially, would be useful for discriminating BM and NBM patients. Subsequently, we recruited another 42 NSCLC patients as an independent validation cohort, consisting of 18 BM and 24 NBM patients (Additional file 2: Table S8). The plasma of NSCLC patients was collected for spermine analysis, revealed that the level of spermine was significantly increased in the BM group compared with the NBM group ( P = 0.008) (Fig. 4d), consist with previous observations from discovery cohorts. Consistently, as indicated by the AUC value of spermine with 0.7014 ( P = 0.0209) (Fig. 4e), further confirming spermine as a potential indicator for predicting BM in NSCLC. Spermine Promotes Bm In Nsclc Via Polarizing Microglia To The M2 Phenotype A BM mouse model was established to explore the function of spermine. As shown in Fig. 5a, the size and the occurrence of brain tumors in spermine-treated (10mg/kg) mice increased remarkably compared with the control group (Fig. 5b-c). Microglia was reportedly is associated with brain tumor progression, with the M2 phenotype acts as pro-tumor factor by immunosuppression [ 12 , 23 , 24 ]. Moreover, spermine reportedly promotes macrophage M2 polarization [ 25 ]. Therefore, we hypothesized that spermine regulates the development of BM by promoting M2 polarization of microglia. Immunofluorescence analyses showed that abundant activated microglia (Iba1 + ) infiltrated brain metastatic lesions, and the number of M2 microglia (Arg-1 + ) was increased in the spermine group, suggesting that spermine promotes microglia to polarize to the M2 phenotype (Fig. 5d). In addition, we analyzed the microglia in brain tumor tissue using flow cytometry (Fig. 5e). These results showed that the percent of CD45 low CD11b + CD206 + (M2) microglia was significantly increased, whereas that of CD45 low CD11b + CD86 + (M1) microglia was decreased in spermine-treated mice compared with control mice (Fig. 5f-g). These results suggest that spermine promotes BM by promoting the polarization of microglia toward the M2 phenotype. To verify this observation, we depleted microglial cells via intracerebroventricular injection of clodronate liposomes (10 µL/mouse) in BM model mice [ 26 ]. Notably, clodronate liposomes significantly suppressed spermine-induced BM (Fig. 5h-i). These results suggest that spermine-induced BM is indeed mediated by the polarization of microglia. Spermine-induced M2 microglia polarization and suppression of phagocytic ability via activation of the STAT3 pathway. To further investigate whether spermine regulates microglia polarization directly, we measured the expression of various M1/M2 markers under spermine (10 µM) treatment of cultured mouse BV2 cells. qRT-PCR and Western blot revealed that the M2 markers were greatly increased in the spermine group compared with the control group, whereas M1 markers were significantly decreased (Fig. 6a-b). Thus, verified that spermine promotes the polarization of microglia to the M2 phenotype, which exerts pro-tumor effects. Studies have reported that activation of the STAT3 pathway is involved in M2 polarization [ 27 ] and related to the promotion of immunosuppression, leading to pro-tumor effects [ 28 ]. To further elucidate the underlying mechanism by which spermine promotes M2 polarization of microglia, phosphorylation of STAT3 was analyzed, which revealed that spermine indeed enhanced the expression of p-STAT3 (Figure. 6c). In addition, activation of STAT3 in the spermine group was significantly suppressed, and M2 polarization of microglia was also significantly suppressed by using STAT3 inhibitor static (0.5 µm) (Fig. 6d). Moreover, inhibition of STAT3 reversed the spermine-mediated up-regulation of M2 CD206 + cells (Fig. 6e-f), suggesting that spermine promotes M2 microglial polarization via activation of the STAT3 pathway. These results showed that spermine enhances the development of BM in NSCLC by promoting microglia M2 polarization, and the STAT3 pathway might be involved in this process. The phagocytic ability of microglia, which exhibit anti-tumor activity, was examined using a fluorescent microsphere phagocytosis assay and flow cytometry. Microglia showed strong phagocytic activity, whereas spermine significantly compromised this activity. Notably, the down-regulation of the phagocytic ability of microglia induced by spermine was suppressed in the stattic group (Fig. 6h-k). The above results suggest that spermine promotes microglia M2 polarization and suppresses the phagocytic ability of microglia via activation of the STAT3 pathway, which in turn promotes BM progression. Discussion BM is a severe complication of advanced NSCLC and accompanied by low quality of life and poor prognosis [ 29 , 30 ]. Clarification of the underlying mechanism of the development of BM in NSCLC is the cornerstone of improving the overall survival of patients. Accumulating evidence suggests that the gut microbiota contributes to brain tumor development and to the success of therapy via GBA [ 5 , 6 ]. Dysregulation of the gut microbiota composition and function is associated with disorders of host metabolism, altering circulating metabolites is particularly effective way to regulate the brain microenvironment [ 9 , 31 ]. Characterization of the gut microbiota and metabolites in NSCLC patients with BM could facilitate the development of biomarkers that could lead to beneficial approaches for therapy. Thus, in our study, we conducted a global untargeted metabolomic analysis of plasma samples in combination with 16S rRNA gene sequencing of fecal sample, to decipher the connection between gut bacterial and host metabolism and generate the basis for novel treatment for BM in NSCLC patients. An imbalance among gut bacteria is a common hallmark of numerous human diseases [ 32 , 33 ], such as low diversity, an increased ratio of Firmicutes/Bacteroidetes , and increased levels of Proteobacteria in the gut microbiota [ 20 , 34 ]. We identified 16 bacterial genera in the two groups that differed significantly in abundance. Interestingly, the genera Klebsiella , Enterobacter , and unclassified_f__Enterobacteriaceae , which belong to Proteobacteria phylum, were more abundant in the BM group, suggesting associated with the development of BM. In addition, genera such as Blautia were reduced the BM group, and these organisms are a source of short-chain fatty acids (SCFA) that relieve inflammation [ 35 ]. The genera Alistipes and Klebsiella , as the classic pathogenic bacteria [ 36 , 37 ], were shown increased in BM group. Notably, the intestinimona genus, belongs to the Ruminococcaceae family, which is associated with the transport of secondary metabolites such as secondary bile acids, was higher in the BM group, indicating that gut dysbiosis might affect the host disease status by altering metabolite levels. Therefore, the tumor-promoting effects of the microbiome in NSCLC patients with BM could derive from holistic dysbiosis and hypothesized that the gut microbiota affects the brain via the GBA, in part by altering circulating metabolites. Using LC-MS/MS and GC/MS approach, our study identified 76 differentially abundant metabolites between the BM and NBM groups, primarily consisting of lipids and lipid-like molecules (e.g., arachidic acid, 3-Methylglutaric Acid and hexadecanedioic acid), organic nitrogen compounds (e.g., spermine, choline and phosphocholine), and organic acids and their derivatives. Perturbations in lipid metabolism can affect the progression of cancers through effects on the stem cell character of tumor cells, angiogenesis, and immune surveillance [ 38 , 39 ]. The metabolites associated with the tumor immune response and thus affect the relationship between the gut microbiota and disease [ 40 ]. In our study, spermine, which belongs to polyamines, have been attributed to human cancers [ 41 , 42 ], was increased in the BM group compared with the NBM group. In addition, the increased concentration of arachidic acid in BM patients, which may be associated with nonalcoholic steatohepatitis [ 43 ], is worthy of further exploration. We also detected altered levels of benzenoids and organoheterocyclic compounds in plasma samples. These compounds exert anti-inflammatory activity [ 44 ] and are related to ameloblastoma biology, respectively. In general, the untargeted plasma metabolomic showed unique and differential metabolic signatures in BM and NBM patients. Using an integrated metabolomics and 16S rRNA gene sequencing approach, we identified the metabolite spermine having a positive correlation with the intestinimonas genus in the BM group. Moreover, ROC curve analysis showed that the model including intestinimonas and spermine is a potential biomarker for predicting the risk of BM in NSCLC patients. An important question is how dose spermine participate in the development of the disease? Spermine is metabolite that is reportedly from both the gut microbiota and the host, and regulates the innate immune response by affecting macrophage activation in the context of host inflammation and carcinogenesis by impairing M1 macrophage responses and inducing an M2-like state [ 25 ]. Microglia are macrophages found in the brain, and their activation affects the microenvironment of brain tumors [ 45 ]. Thus, we hypothesize that spermine promotes the development of BM by activating microglia and inducing polarization to the M2 phenotype, exerting pro-tumor effects. Depletion of microglia via intracerebral injection of clodronate liposomes supported this hypothesis. Furthermore, STAT3 activation also reportedly stimulates microglia polarization to the M2 phenotype, leading to the development of an immunosuppressive microenvironment that promotes brain tumor growth [ 46 ]. Interestingly, extracellular spermine would be able to directly modulate microglial via the polyamine transport system, which is involved in the innate immune system [ 47 ]. Our study illustrated that spermine modulates the polarization of microglia toward the M2 phenotype and suppress the phagocytic ability of these cells by increasing the phosphorylation of STAT3. There are several limitations to our study, including the relatively small sizes of the patient cohorts in the multi-omic analysis. Thus, multi-center studies of NSCLC patients with/without BM will be needed to further validate our study’s findings. Notably, target bacteria also required further verification. In addition, several unknown metabolites that could not be identified in the currently available databases were also found to be significantly associated with several bacterial genera. Further investigations of the effects of the spermine synthesis and degradation pathways are also warranted. Lipid metabolism, including that of arachidic acid and 3-methyl glutaric acid, deserve more attention in future studies as well in order to explore their potential for use as biomarkers in BM patients. We cannot rule out the possibility that spermine activates microglia via other pathways in brain tumor development, and this possibility deserves in-depth study. Despite these shortcomings, our results demonstrate that the combination of spermine and intestinimonas represents a potential therapeutic target for treating BM in NSCLC patients. Spermine are potential biomarkers for discriminating BM and NBM, pending further validation studies. Conclusions We demonstrated the signature microbiota and metabolites that distinguish BM and NBM patients. However, the microbial diversity did not differ significantly compared with the NBM group. Multi-omics analysis revealed a correlation between the differentially abundant genera and various metabolites, and we filtered out three combinations, including spermine and intestinimonas , arachidic acid and Candidatus_Soleaferrea , 3-methyl glutaric acid and Holdemania , suggesting that gut microbiota promote the development of BM in NSCLC in part through altering the metabolome. We identified the intestinimonas genus in NSCLC patients with BM as key commensal intestinal bacteria positively correlated with the plasma metabolite spermine, promoting brain tumor growth. In addition, the model of intestinimonas and spermine is a potential biomarker for predicting the risk of BM in NSCLC. Notably, spermine was identified as a key plasma metabolite that act through STAT3 signaling to promote microglia M2 polarization, exerting a pro-tumor effect in NSCLC patients with BM. In summary, our findings provide novel insights for future investigations into the causal associations between gut dysbiosis and host metabolites in the development and progression of BM in NSCLC. We identified spermine as a potential therapeutic target that could prove useful in the prevention, diagnosis and treatment of NSCLC with BM. Abbreviations BM: Brain metastasis; NSCLC: non-small cell lung cancer; GBA: gut-brain-axis; LC-MS/MS: Liquid chromatography-tandem mass spectrometry; GC/MS: Gas chromatography-mass spectrometry; STAT3: Signal Transducer and Activator of Transcription 3; CEA: serum carcinoembryonic antigen; PCoA: principal coordinate analysis; LDA: Linear discriminant analysis; LEfSe: LDA effect size; PCA: Principal components analysis; PLS-DA: Partial least squares discriminant analysis; LLC: Lewis lung cancer cell lines; LLC-Luc: luciferase-labeled LLC; Iba1: ionized calcium binding adapter molecule-1; Arg-1: arginase-1; iNOS: inducible nitric oxide synthase; Ym-1: chitinase 3-like 3; CD206: mannose receptor; TGF-β: transforming growth factor β; RF: random forest; neg: negative; pos: positive; SCFA: short-chain fatty acids Declarations Ethics approval and consent to participate This study was approved by the Institutional Review Board of Huazhong University of Science and Technology. Written informed consent was obtained from all legal guardians of the patients. All animal experiments were conducted in agreement with the Guide for the Care and Use of Laboratory Animals and were approved by the Committee on Animal Handling of Huazhong University of Science and Technology. Data Availability Statement The authors confirm that the data supporting the findings of this study are available within the article or the supplementary materials. Acknowledgments Not applicable. Authors’ Disclosures No disclosures were reported. Competing interests The authors declare that they have no competing interests. Funding This research was supported by the grants from the National Key R&D Program of China (Grant No 2019YFC1316205, Grant No. 2019YFC1316205) Authors' contributions XRD and NY conceived the study. HHL, LCL, and HZ performed the experiments. HHL, YWB, HZ and RGZ collected clinical samples. HHL, FT and JJW analyzed the data. XRD, FT and HHL wrote the manuscript. All authors read and approved the final manuscript. References Ernani, V. and T.E. Stinchcombe, Management of Brain Metastases in Non-Small-Cell Lung Cancer. J Oncol Pract, 2019. 15;11;563-570. Khalifa, J., A. Amini, S. Popat, L.E. Gaspar, and C. Faivre-Finn, Brain Metastases from NSCLC: Radiation Therapy in the Era of Targeted Therapies. J Thorac Oncol, 2016. 11;10;1627-43. Hall, A.B., A.C. Tolonen, and R.J. Xavier, Human genetic variation and the gut microbiome in disease. Nat Rev Genet, 2017. 18;11;690-699. Durack, J. and S.V. Lynch, The gut microbiome: Relationships with disease and opportunities for therapy. J Exp Med, 2019. 216;1;20-40. Mehrian-Shai, R., J.K.V. Reichardt, C.C. Harris, and A. Toren, The Gut-Brain Axis, Paving the Way to Brain Cancer. Trends Cancer, 2019. 5;4;200-207. Cryan, J.F., K.J. O'Riordan, C.S.M. Cowan, K.V. Sandhu, T.F.S. Bastiaanssen, M. Boehme, et al., The Microbiota-Gut-Brain Axis. Physiol Rev, 2019. 99;4;1877-2013. Nicholson, J.K., J.K. Elaine Holmes, Remy Burcelin, Glenn Gibson, Wei Jia, and S. Pettersson, Host-gut microbiota metabolic interactions. Science, 2012 33;6086;1262-1267. Rothhammer, V., I.D. Mascanfroni, L. Bunse, M.C. Takenaka, J.E. Kenison, L. Mayo, et al., Type I interferons and microbial metabolites of tryptophan modulate astrocyte activity and central nervous system inflammation via the aryl hydrocarbon receptor. Nat Med, 2016. 22;6;586-97. van de Wouw, M., M. Boehme, J.M. Lyte, N. Wiley, C. Strain, O. O'Sullivan, et al., Short-chain fatty acids: microbial metabolites that alleviate stress-induced brain-gut axis alterations. 2018. 596;20;4923-4944. Matcovitch-Natan, O., D.R. Winter, A. Giladi, S. Vargas Aguilar, A. Spinrad, S. Sarrazin, et al., Microglia development follows a stepwise program to regulate brain homeostasis. Science, 2016. 353;6301;aad8670. Gutmann, D.H. and H. Kettenmann, Microglia/Brain Macrophages as Central Drivers of Brain Tumor Pathobiology. Neuron, 2019. 104;3;442-449. Wu, S.Y., F. Xing, S. Sharma, K. Wu, A. Tyagi, Y. Liu, et al., Nicotine promotes brain metastasis by polarizing microglia and suppressing innate immune function. J Exp Med, 2020. 217;8. Wolf, S.A., H.W. Boddeke, and H. Kettenmann, Microglia in Physiology and Disease. Annu Rev Physiol, 2017. 79;619-643. Sherwin, E., K. Rea, T.G. Dinan, and J.F. Cryan, A gut (microbiome) feeling about the brain. Curr Opin Gastroenterol, 2016. 32;2;96-102. Dinan, T.G. and J.F. Cryan, The Microbiome-Gut-Brain Axis in Health and Disease. Gastroenterol Clin North Am, 2017. 46;1;77-89. Wei, C., X. Dong, H. Lu, F. Tong, L. Chen, R. Zhang, et al., LPCAT1 promotes brain metastasis of lung adenocarcinoma by up-regulating PI3K/AKT/MYC pathway. J Exp Clin Cancer Res, 2019. 38;1;95. Martin, E., M. El-Behi, B. Fontaine, and C. Delarasse, Analysis of Microglia and Monocyte-derived Macrophages from the Central Nervous System by Flow Cytometry. J Vis Exp, 2017;124. Li, H., F. Tong, R. Meng, L. Peng, J. Wang, R. Zhang, et al., E2F1-mediated repression of WNT5A expression promotes brain metastasis dependent on the ERK1/2 pathway in EGFR-mutant non-small cell lung cancer. Cell Mol Life Sci, 2021. 78;6;2877-2891. Wang, J., H. Pan, Z. Lin, C. Xiong, C. Wei, H. Li, et al., Neuroprotective Effect of Fractalkine on Radiation-induced Brain Injury Through Promoting the M2 Polarization of Microglia. Mol Neurobiol, 2021. 58;3;1074-1087. Mariat, D., O. Firmesse, F. Levenez, V.D. Guimarăes, H. Sokol, J. Doré, et al., The Firmicutes/Bacteroidetes ratio of the human microbiota changes with age. BMC Microbiology, 2009. 9;1;123. Ngo, B., E. Kim, and V. Osorio-Vasquez, Limited Environmental Serine and Glycine Confer Brain Metastasis Sensitivity to PHGDH Inhibition. 2020. 10;9;1352-1373. Rooks, M.G. and W.S. Garrett, Gut microbiota, metabolites and host immunity. Nat Rev Immunol, 2016. 16;6;341-52. Roesch, S., C. Rapp, S. Dettling, and C. Herold-Mende, When Immune Cells Turn Bad-Tumor-Associated Microglia/Macrophages in Glioma. Int J Mol Sci, 2018. 19;2. Guldner, I.H., Q. Wang, L. Yang, S.M. Golomb, Z. Zhao, J.A. Lopez, et al., CNS-Native Myeloid Cells Drive Immune Suppression in the Brain Metastatic Niche through Cxcl10. Cell, 2020. 183;5;1234-1248 e25. Latour, Y.L., A.P. Gobert, and K.T. Wilson, The role of polyamines in the regulation of macrophage polarization and function. Amino Acids, 2020. 52;2;151-160. Todd, L., I. Palazzo, L. Suarez, X. Liu, L. Volkov, T.V. Hoang, et al., Reactive microglia and IL1beta/IL-1R1-signaling mediate neuroprotection in excitotoxin-damaged mouse retina. J Neuroinflammation, 2019. 16;1;118. Qin, C., W.H. Fan, Q. Liu, K. Shang, M. Murugan, L.J. Wu, et al., Fingolimod Protects Against Ischemic White Matter Damage by Modulating Microglia Toward M2 Polarization via STAT3 Pathway. Stroke, 2017. 48;12;3336-3346. Jones, L.M., M.L. Broz, J.J. Ranger, J. Ozcelik, R. Ahn, D. Zuo, et al., STAT3 Establishes an Immunosuppressive Microenvironment during the Early Stages of Breast Carcinogenesis to Promote Tumor Growth and Metastasis. Cancer Res, 2016. 76;6;1416-28. Andratschke, N., J. Kraft, C. Nieder, R. Tay, R. Califano, R. Soffietti, et al., Optimal management of brain metastases in oncogenic-driven non-small cell lung cancer (NSCLC). Lung Cancer, 2019. 129;63-71. Bacha, S., H. Cherif, D. Rabaa, S. Habibech, S. Cheikhrouhou, H. Racil, et al., Brain metastases of non-small cell lung cancer: prognostic factors and management. Tunis Med, 2018. 96;3;165-171. Roager, H.M. and T.R. Licht, Microbial tryptophan catabolites in health and disease. Nat Commun, 2018. 9;1;3294. Kriss, M., K.Z. Hazleton, N.M. Nusbacher, C.G. Martin, and C.A. Lozupone, Low diversity gut microbiota dysbiosis: drivers, functional implications and recovery. Curr Opin Microbiol, 2018. 44;34-40. Ruppé, É., T. Lisboa, and F. Barbier, The gut microbiota of critically ill patients: first steps in an unexplored world. Intensive Care Med, 2018. 44;9;1561-1564. Shin, N.R., T.W. Whon, and J.W. Bae, Proteobacteria: microbial signature of dysbiosis in gut microbiota. Trends Biotechnol, 2015. 33;9;496-503. Zhang, X., Y. Zhao, M. Zhang, X. Pang, J. Xu, C. Kang, et al., Structural changes of gut microbiota during berberine-mediated prevention of obesity and insulin resistance in high-fat diet-fed rats. PLoS One, 2012. 7;8;e42529. Dorman, M.J. and F.L. Short, Genome watch: Klebsiella pneumoniae: when a colonizer turns bad. Nat Rev Microbiol, 2017. 15;7;384. Moschen, A.R., R.R. Gerner, J. Wang, V. Klepsch, T.E. Adolph, S.J. Reider, et al., Lipocalin 2 Protects from Inflammation and Tumorigenesis Associated with Gut Microbiota Alterations. Cell Host Microbe, 2016. 19;4;455-69. Röhrig, F. and A. Schulze, The multifaceted roles of fatty acid synthesis in cancer. Nat Rev Cancer, 2016. 16;11;732-749. Ma, Y., S.M. Temkin, A.M. Hawkridge, C. Guo, W. Wang, X.Y. Wang, et al., Fatty acid oxidation: An emerging facet of metabolic transformation in cancer. Cancer Lett, 2018. 435;92-100. Kumar, A., C. Smith, C. Jobin, G. Trinchieri, T.K. Howcroft, H. Seifried, et al., Workshop Report: Modulation of Antitumor Immune Responses by Dietary and Microbial Metabolites. J Natl Cancer Inst, 2017. 109;6. Pegg, A.E., The function of spermine. IUBMB Life, 2014. 66;1;8-18. Nowotarski, S.L., P.M. Woster, and R.A. Casero, Jr., Polyamines and cancer: implications for chemotherapy and chemoprevention. Expert Rev Mol Med, 2013. 15;e3. Ye, J.Z., Y.T. Li, W.R. Wu, D. Shi, D.Q. Fang, L.Y. Yang, et al., Dynamic alterations in the gut microbiota and metabolome during the development of methionine-choline-deficient diet-induced nonalcoholic steatohepatitis. World J Gastroenterol, 2018. 24;23;2468-2481. Shie, P.H., S.Y. Wang, H.L. Lay, and G.J. Huang, 4,7-Dimethoxy-5-methyl-1,3-benzodioxole from Antrodia camphorata inhibits LPS-induced inflammation via suppression of NF-κB and induction HO-1 in RAW264.7 cells. Int Immunopharmacol, 2016. 31;186-94. Beggs, S. and M.W. Salter, SnapShot: Microglia in Disease. Cell, 2016. 165;5;1294-1294.e1. Wu, S.Y. and K. Watabe, The roles of microglia/macrophages in tumor progression of brain cancer and metastatic disease. Front Biosci (Landmark Ed), 2017. 22;1805-1829. Soulet, D. and S. Rivest, Polyamines play a critical role in the control of the innate immune response in the mouse central nervous system. J Cell Biol, 2003. 162;2;257-68. Supplementary Files figures7.png Figure. S1 The overview of the workflow integrating gut microbiome and plasma metabolome in NSCLC patients with BM and NBM Recruited 70 patients diagnosed with NSCLC for integrated analysis, in the process of testing and evaluating the quality of the fecal and plasma samples, the paired samples of 3 patients were excluded owing to hemolysis and too low DNA content. Finally, 32 BM and 35 NBM patients performed the integrating analysis of 16S rRNA gene sequencing and metabolic data. figures8.png Figure. S2 The analysis of gut microbiome in BM and NBM patients of NSCLC a Rarefaction curves for the genus numbers in the BM (n = 32) and NBM (n = 35) groups. b Venn diagram shows the common OTU in the BM and NBM groups. c The hierarchical taxonomic composition of the microbial community in the BM and NBM patients analysis via sunburst diagram (upper). From the innermost circle to the outermost circle, the order is domain, kingdom, and phylum. The proportion of different phylum displayed in form (lower). d Heat map of the Spearman’s rank correlation coefficient between the clinical indices and the composition genera of the microbial community in 67 patients, the right legend is the color range of different R values. e COG function classification statistical box chart of different pathways predicted by PICRUSt software in all samples microbial communities. * P < 0.05, ** P < 0.01, *** P ≤ 0.001. figures9.png Figure. S3 The analysis of plasma metabolome from BM and NBM patients. a & b PCA score plots constructed with 397 metabolites from 32 BM samples (bule points) and 35 NBM samples (red points) in negative (neg) ion mode(left) and 676 metabolites in positive (pos) ion mode (right) in LC-MS/MS (a) and 303 metabolites in GC-MS (b). c & d Statistical validation of the PLS-DA using comparison of explained variation (R2Y) and evaluated the prediction ability (Q2Y) of the PLS-DA model in LC-MS/MS (c) and GC-MS (d). e Scatter plot illustrating statistical associations between the relative abundance of differential genera at the level of 16S OTUs of gut microbiota and altered plasma metabolites (|rho| ≥0.4, and P ≤0.05), Com_113 and Anaerotruncus or [Eubacterium]_nodatum_group in the GC-MS model. supplementtable.xlsx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-2259805","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":155290863,"identity":"f9ce94eb-cbf3-49f0-bdf8-05cdcf275f18","order_by":0,"name":"huanhuan li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA10lEQVRIie3RsQrCMBCA4RPBqdo1U/sKEUd9mBShXSw4lY4pBSfBNeDgM/gGp4FOobODQydnxaWLYKqDY+ImmH87uI8LBMDl+sEGfonI6CwAwG60ICNSRdjk8cSeBLCYHBp1jPj7qM3DAClGK0x2haJwzST4W24gfc40OacFV7QnagnkjMYrqMklLUHR/nAlgRJmIhHXRCaDjjzsyByQKcm8jvSsCKk0yeOxgGp5WNeJR04GEm4291tLZ2Eo5L5ps2ngCwP5RPD1mZ7tvs7nXyy7XC7XX/UES/pL1iHEiCgAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-8179-3633","institution":"Huazhong University of Science and Technology Tongji Medical College","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"huanhuan","middleName":"","lastName":"li","suffix":""},{"id":155290864,"identity":"61e818d7-0541-4ec5-ae7f-9395114081cc","order_by":1,"name":"Lichao Liu","email":"","orcid":"","institution":"Huazhong University of Science and Technology Tongji Medical College","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Lichao","middleName":"","lastName":"Liu","suffix":""},{"id":155290865,"identity":"6ea1d227-2096-427d-9729-0a883b8a8ac5","order_by":2,"name":"Yawen Bin","email":"","orcid":"","institution":"Huazhong University of Science and Technology Tongji Medical College","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yawen","middleName":"","lastName":"Bin","suffix":""},{"id":155290866,"identity":"3e615748-39a5-4644-998c-616e75964ed2","order_by":3,"name":"Hao Zeng","email":"","orcid":"","institution":"Huazhong University of Science and Technology Tongji Medical College","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Zeng","suffix":""},{"id":155290867,"identity":"b3a5312f-d9bf-4db3-98ba-c758fb2f4e17","order_by":4,"name":"Jiaojiao Wang","email":"","orcid":"","institution":"Huazhong University of Science and Technology Tongji Medical College","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Jiaojiao","middleName":"","lastName":"Wang","suffix":""},{"id":155290868,"identity":"8b79ea5f-729b-4610-9c9f-fe022966ec5c","order_by":5,"name":"Ruiguang Zhang","email":"","orcid":"","institution":"Huazhong University of Science and Technology Tongji Medical College","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Ruiguang","middleName":"","lastName":"Zhang","suffix":""},{"id":155290869,"identity":"32d02ac2-76cb-4b2e-9cd7-4885d825b5f2","order_by":6,"name":"Fan Tong","email":"","orcid":"","institution":"Huazhong University of Science and Technology Tongji Medical College","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Fan","middleName":"","lastName":"Tong","suffix":""},{"id":155290870,"identity":"45fdef3a-9650-426e-a227-31149f40f90e","order_by":7,"name":"Nong Yang","email":"","orcid":"","institution":"Hunan Cancer Hospital","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Nong","middleName":"","lastName":"Yang","suffix":""},{"id":155290871,"identity":"d8409a11-f234-419b-9794-9d8608b11706","order_by":8,"name":"xiaorong Dong","email":"","orcid":"https://orcid.org/0000-0001-7470-1836","institution":"Huazhong University of Science and Technology Tongji Medical College","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"xiaorong","middleName":"","lastName":"Dong","suffix":""}],"badges":[],"createdAt":"2022-11-10 14:49:38","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-2259805/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-2259805/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":29726846,"identity":"7eadb16d-0d16-4956-8740-5b30663974f2","added_by":"auto","created_at":"2022-11-30 15:47:35","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":328494,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe diversity and composition of gut microbiome in BM and NBM patients of NSCLC.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea \u0026amp; b\u003c/strong\u003e Sob and Shannon indexes were used to estimate the α-diversity of the gut microbiota in the BM and NBM groups. \u003cstrong\u003ec \u003c/strong\u003ePrincipal coordinate analysis (PCoA) of Bray-Curtis analysis of the β-diversity of the two groups at the genus level. Two-tailed Wilcoxon rank-sum test was used to determine significance. \u003cstrong\u003ed \u003c/strong\u003eThe top 20\u003cstrong\u003e \u003c/strong\u003egenera that were differentially enriched in BM and NBM patients (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05). Two-tailed Wilcoxon rank-sum test was used to determine significance. \u003cstrong\u003ee\u003c/strong\u003eLinear discriminant analysis (LDA) integrated with effect size (LEfSe). LDA scores for the differentially abundant bacterial genera between BM and NBM (LDA \u0026gt; 2.0). Red bars indicate taxa enriched in BM, and blue bars indicate taxa enriched in NBM.\u003c/p\u003e","description":"","filename":"figures1.png","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/5db6364d22b3af44ee3f9d35.png"},{"id":29726849,"identity":"e3bc5d15-72c7-4136-a09c-f1d1f614b5d6","added_by":"auto","created_at":"2022-11-30 15:47:35","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":470433,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMetabolomics profiling of plasma from BM and NBM patients.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea \u0026amp; b \u003c/strong\u003ePLS-DA score plots constructed with 398 metabolites from 32 BM samples (red points) and 35 NBM samples (blue points) in negative (neg) ion mode (left) and 677 metabolites in positive (pos) ion mode (right) in LC-MS/MS (PLS-DA models: neg: R2Y 0.60 and Q2Y 0.40; pos: R2Y0.65 and Q2Y 0.31) \u003cstrong\u003e(a)\u003c/strong\u003e and 303 metabolites in GC-MS (PLS-DA models: R2Y 0.84 and Q2Y 0.78) \u003cstrong\u003e(b)\u003c/strong\u003e. \u003cstrong\u003ec \u0026amp; d\u003c/strong\u003e Volcano map showing the differential metabolites of BM significantly up-regulated (red dots) and down-regulated (green dots) compared with the NBM group; the size of the dots represents the VIP value. Twenty-nine differential metabolites identified in neg-ion mode (left) and 33 metabolites in pos-ion mode in LC-MS/MS \u003cstrong\u003e(c)\u003c/strong\u003e, and 14 differential metabolites identified in GC-MS \u003cstrong\u003e(d)\u003c/strong\u003e. \u003cstrong\u003ee \u0026amp; f \u003c/strong\u003eKEGG bubble diagram\u003cstrong\u003e \u003c/strong\u003eshowing metabolic pathway enrichment of top 20 differential metabolites in LC-MS/MS (left: neg mode, right: pos mode) and GC-MS. The size of the points represents the number of differential metabolites in the corresponding pathway. The color of the point represents the \u003cem\u003eP\u003c/em\u003e-value value of the hypergeometric test.\u003c/p\u003e","description":"","filename":"figures2.png","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/aa2a89d38a158db709cfd380.png"},{"id":29726848,"identity":"3d74a2ab-2e59-42f5-8719-39e8e2864f3d","added_by":"auto","created_at":"2022-11-30 15:47:35","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":756179,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eInterrelationship between the gut microbiota and plasma metabolites of BM and NBM patients in LC-MS/MS and GC-MS models.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea \u0026amp; b \u003c/strong\u003ePearson correlation calculated between gut microbiota of top 16 significant genera and top 20 differential metabolites in LC-MS/MS (\u003cstrong\u003ea\u003c/strong\u003e: negative ion mode, \u003cstrong\u003eb\u003c/strong\u003e: positive ion mode). Blue indicates positive correlation, red indicates negative correlation (\u003cem\u003eP\u003c/em\u003e ≤0.05), and blank indicates no statistical significance. \u003cstrong\u003ec\u003c/strong\u003ePearson correlation calculated between gut microbiota of top 16 significant genera and top 14 differential metabolites in GC-MS. Blue indicates positive correlation, red indicates negative correlation (\u003cem\u003eP\u003c/em\u003e ≤0.05), and blank indicates no statistical significance. \u003cstrong\u003ed \u0026amp; e\u003c/strong\u003e Scatter plot illustrating statistical associations between the relative abundance of differential genera at the level of 16S OTUs of gut microbiota and altered plasma metabolites (|rho|≥0.4, and \u003cem\u003eP\u003c/em\u003e≤0.05), including Com_211_pos (Spermine),Com_4055_pos (3-Hydroxyproline), Com_867_neg (Arachidic acid) in the LC-MS/MS model \u003cstrong\u003e(d)\u003c/strong\u003e and Com_206 (3-methyl glutaric acid) in the GC-MS model \u003cstrong\u003e(e)\u003c/strong\u003e.\u003c/p\u003e","description":"","filename":"figures3.png","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/85a1c8a73e37cee47e3798df.png"},{"id":29727663,"identity":"749c7f2f-a53d-49aa-be07-6065b8e1584e","added_by":"auto","created_at":"2022-11-30 15:55:35","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":142142,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAnalysis of potential plasma metabolite markers in BM and NBM patients.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e Receiver operating characteristic (ROC) curve of the random forest analysis of 3 identified plasma metabolites, arachidic acid (AUC=0.746), spermine (AUC=0.730), and 3-methyl glutaric acid (AUC=0.846). \u003cstrong\u003eb \u003c/strong\u003eROC curve for the predictive biomarker. The mix of 3 metabolites (arachidic acid, spermine, and 3-methyl glutaric acid) was selected based on the feature elimination step, and the best predictive model with the highest AUC value (AUC = 0.923, blank curve) was constructed.\u003cstrong\u003e c \u003c/strong\u003eThe mix of spermine and i\u003cem\u003entestinimonas\u003c/em\u003ewere selected for ROC analysis, and the predictive model with the AUC value (AUC = 0.7527, blank curve) was constructed.\u003cstrong\u003e d\u003c/strong\u003e The content of plasma spermine was determined by ELISA in the 42 verification NSCLC patients. Data shown are the average of three independent experiments. Student’s \u003cem\u003et\u003c/em\u003e test was used to analyze the difference between two groups. *, \u003cem\u003eP\u003c/em\u003e\u0026lt; 0.05; **,\u003cem\u003e P\u003c/em\u003e \u0026lt; 0.01. \u003cstrong\u003ee\u003c/strong\u003eROC curve of the random forest analysis of plasma spermine levels in the verification NSCLC group. (AUC = 0.7014, blank curve).\u003c/p\u003e","description":"","filename":"figures4.png","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/bbcb0bc3ae215597c8e6b25b.png"},{"id":29728144,"identity":"2b3d5098-845b-46f2-88b3-8ef1cf36e29d","added_by":"auto","created_at":"2022-11-30 16:03:35","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1764718,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePlasma metabolite spermine promotes BM in NSCLC by facilitating M2 polarization of microglia.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea \u003c/strong\u003eMouse Lewis lung carcinoma cells labeled with luciferase (LLC-luc, ×10\u003csup\u003e5\u003c/sup\u003e cells per mouse) were injected via the intracarotid artery into C57BL/6 mice (n = 6). Seven days after injection of LLC-luc cells, mice received spermine (10 mg/kg) via intraperitoneal injection every 3 d until the end point. \u003cstrong\u003eb \u0026amp; c \u003c/strong\u003eRepresentative bioluminescent images \u003cstrong\u003e(b)\u003c/strong\u003e and quantitative data analysis\u003cstrong\u003e (c)\u003c/strong\u003e in brain metastatic mice from the control or spermine group. The data are expressed as the mean ± SD, *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, **\u003cstrong\u003eP\u003c/strong\u003e \u0026lt; 0.01, as determined by the \u003cem\u003et\u003c/em\u003e-test. \u003cstrong\u003ed\u003c/strong\u003e Representative images of immunofluorescence analysis of Arg-1+ and Iba1+ microglia in the metastatic brain lesions of mice that were treated with or without spermine (n = 6/group) (left). Scale bar, 20 μm. Quantitative analysis of immunofluorescence in Arg-1+ /Iba1+ microglia (right). The data are expressed as the mean ± SEM, *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01, as determined by the \u003cem\u003et\u003c/em\u003e-test. \u003cstrong\u003ee \u003c/strong\u003eSchematic diagram of gate choose by flow cytometry analysis of microglia.\u003cstrong\u003e f \u0026amp; g\u003c/strong\u003e Brain samples in \u003cstrong\u003eb\u003c/strong\u003e were isolated and examined by flow cytometry for M1 (CD45\u003csup\u003elow\u003c/sup\u003eCD11b\u003csup\u003e+\u003c/sup\u003eCD86+) and M2 (CD45\u003csup\u003elow\u003c/sup\u003eCD11b\u003csup\u003e+\u003c/sup\u003eCD206+) microglial phenotypes. \u003cstrong\u003eh \u0026amp; i \u003c/strong\u003eRepresentative bioluminescent images \u003cstrong\u003e(g)\u003c/strong\u003e and quantitative data analysis \u003cstrong\u003e(h)\u003c/strong\u003e in brain metastatic mice from control or spermine plus clodronate liposomes groups. The data are expressed as the mean ± SD, *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01, as determined by the \u003cem\u003et\u003c/em\u003e-test.\u003c/p\u003e","description":"","filename":"figures5.png","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/d430c50181965eaae8dbe040.png"},{"id":29727665,"identity":"4b7a6ef7-1539-483d-967d-a94fc9e68c42","added_by":"auto","created_at":"2022-11-30 15:55:35","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":613928,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSpermine induces M2 polarization of microglia with activation of the STAT3 pathway\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea \u003c/strong\u003eExpression of M1/M2 microglia markers was examined by qRT-PCR in BV2 cells with or without spermine (10 μM). Data were normalized using GAPDH as a control.\u003cstrong\u003e b\u003c/strong\u003e BV2 cells were treated with or without spermine (10 μM), and the expression of M1/M2 markers was examined by Western blotting. \u003cstrong\u003ec\u003c/strong\u003e BV2 cells were treated with or without spermine (10 μM) for 24 h, and the expression of p-STAT3 was examined by Western blotting. Data shown are the average of three independent experiments. Differences between groups were analyzed using the Student’s \u003cem\u003et\u003c/em\u003e-test. *, \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05; **, \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01. \u003cstrong\u003ed \u003c/strong\u003eBV2 cells were treated with or without spermine (10 μM) and/or stattic (0.5 μM) for 24 h, and the expression of M1/M2 markers and STAT3 was examined by Western blotting. \u003cstrong\u003ee \u0026amp; f \u003c/strong\u003eBV2 cells were treated with or without spermine (10 μM) and/or Stattic (0.5 μM) for 24 h, the expression of of M1/M2 marker was examined by flow cytometry, with a blank as a control without antibody labeling. \u003cstrong\u003eh \u0026amp; i \u003c/strong\u003ePhagocytic activity of BV2 cells treated with or without spermine (10 μM) and/or stattic (0.5 μM) for 24 h. Representative fluorescent images showing different phagocytosis ability of microspheres (green) in BV2 cells with phalloidin (red) are showen. Scale bar = 20 μm. The results of statistical analyses of phagocytosis of fluorescent microspheres are shown in \u003cstrong\u003ee\u003c/strong\u003e (n =3/group). \u003cstrong\u003ej \u0026amp; k \u003c/strong\u003ePhagocytosis capacity was analyzed by flow cytometry after co-culture of BV2 cells with microspheres for 3 h, after treatment with or without spermine (10 μM) and/or stattic (0.5 μM) for 24 h. Results of statistical analyses of the number of phagocytic fluorescent microspheres are shown in g (n =3/group). Data shown are the average of three independent experiments. *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01, determined by unpaired two-sided \u003cem\u003et\u003c/em\u003e-test. Data are presented as the mean ± SEM. Differences among multiple groups were examined by one-way ANOVA with post-hoc Tukey honestly significant difference test.\u003c/p\u003e","description":"","filename":"figures6.png","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/2576a278c7f8c0a56a0cb017.png"},{"id":31711747,"identity":"a7d72bf8-dbdb-4ad6-9250-4f61259cc0c2","added_by":"auto","created_at":"2023-01-17 21:33:58","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2406974,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/e0c103db-f3ae-4294-9553-db74cb3c4945.pdf"},{"id":29726845,"identity":"542e611b-a3a5-4467-8a37-f440b9dd385a","added_by":"auto","created_at":"2022-11-30 15:47:35","extension":"png","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":62561,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure. S1 The overview of the workflow integrating gut microbiome and plasma metabolome in NSCLC patients with BM and NBM\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRecruited 70 patients diagnosed with NSCLC for integrated analysis, in the process of testing and evaluating the quality of the fecal and plasma samples, the paired samples of 3 patients were excluded owing to hemolysis and too low DNA content. Finally, 32 BM and 35 NBM patients performed the integrating analysis of 16S rRNA gene sequencing and metabolic data.\u003c/p\u003e","description":"","filename":"figures7.png","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/3b58605dfdf0edeb221c7137.png"},{"id":29727662,"identity":"d95c3ae3-c718-4c89-8890-5ef29c288e0e","added_by":"auto","created_at":"2022-11-30 15:55:35","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":726016,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure. S2 The analysis of gut microbiome in BM and NBM patients of NSCLC\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e Rarefaction curves for the genus numbers in the BM (n = 32) and NBM (n = 35) groups. \u003cstrong\u003eb\u003c/strong\u003e Venn diagram shows the common OTU in the BM and NBM groups. \u003cstrong\u003ec\u003c/strong\u003e The hierarchical taxonomic composition of the microbial community in the BM and NBM patients analysis via sunburst diagram (upper). From the innermost circle to the outermost circle, the order is domain, kingdom, and phylum. The proportion of different phylum displayed in form (lower). \u003cstrong\u003ed\u003c/strong\u003e Heat map of the Spearman’s rank correlation coefficient between the clinical indices and the composition genera of the microbial community in 67 patients, the right legend is the color range of different R values. \u003cstrong\u003ee \u003c/strong\u003eCOG function classification statistical box chart of different pathways predicted by PICRUSt software in all samples microbial communities. *\u003cem\u003e P\u003c/em\u003e \u0026lt; 0.05, ** \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01, *** \u003cem\u003eP\u003c/em\u003e ≤ 0.001.\u003c/p\u003e","description":"","filename":"figures8.png","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/ee705098ef9ffd25293081d4.png"},{"id":29726854,"identity":"eeda3dad-e73d-4184-a746-73a1b9e5f87c","added_by":"auto","created_at":"2022-11-30 15:47:35","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":264543,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure. S3 The analysis of plasma metabolome from BM and NBM patients.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea \u0026amp; b\u003c/strong\u003e PCA score plots constructed with 397 metabolites from 32 BM samples (bule points) and 35 NBM samples (red points) in negative (neg) ion mode(left) and 676 metabolites in positive (pos) ion mode (right) in LC-MS/MS \u003cstrong\u003e(a)\u003c/strong\u003e and 303 metabolites in GC-MS \u003cstrong\u003e(b)\u003c/strong\u003e. \u003cstrong\u003ec \u0026amp; d \u003c/strong\u003eStatistical validation of the PLS-DA using comparison of explained variation (R2Y) and evaluated the prediction ability (Q2Y) of the PLS-DA model in LC-MS/MS \u003cstrong\u003e(c)\u003c/strong\u003e and GC-MS \u003cstrong\u003e(d)\u003c/strong\u003e. \u003cstrong\u003ee\u003c/strong\u003e Scatter plot illustrating statistical associations between the relative abundance of differential genera at the level of 16S OTUs of gut microbiota and altered plasma metabolites (|rho| ≥0.4, and \u003cem\u003eP\u003c/em\u003e≤0.05), Com_113 and \u003cem\u003eAnaerotruncus \u003c/em\u003eor \u003cem\u003e[Eubacterium]_nodatum_group\u003c/em\u003e in the GC-MS model.\u003c/p\u003e","description":"","filename":"figures9.png","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/cebbccaa8f3ad2e193f144be.png"},{"id":29726852,"identity":"3c9d54b5-60f8-4180-a194-c86877b2545d","added_by":"auto","created_at":"2022-11-30 15:47:35","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":53264,"visible":true,"origin":"","legend":"","description":"","filename":"supplementtable.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-2259805/v1/4f243871183845c6b78835b3.xlsx"}],"financialInterests":"","formattedTitle":"Microbiota-modulated spermine promotes brain metastasis in non-small cell lung cancer by regulating microglia M2 polarization via the STAT3 pathway","fulltext":[{"header":"Background","content":"\u003cp\u003eBrain metastasis (BM) is a leading cause of mortality worldwide in patients with advanced non-small cell lung cancer (NSCLC) [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Accumulating evidence suggests that the gut microbiota is a critical environmental factor associated with the development of a growing number of diseases, including the progression of cancer, inflammation, and metabolic syndrome [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Studies have revealed that the gut microbiota affects brain function and behavior via the gut-brain axis (GBA), which involves bidirectional communication between the gut microbiome and the brain through neural, endocrine, and immune pathways [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMetabolites are considered important mediators between the gut microbiota and brain by entering the bloodstream and exerting functional effects [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. For example, a study using a multiple sclerosis animal model indicated that some microbial-regulated tryptophan metabolites bind to aryl hydrocarbon receptor, to suppress brain inflammation [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Microbial-derived molecules such as short-chain fatty acids can cross into the brain and alleviate chronic psychosocial stress [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. We hypothesized that the gut microbiota can affect the initiation or progression of brain tumors by modulating the circulation metabolism of the host. However, the profiles of the gut microbial community and host metabolites in NSCLC patients with BM have not been systematically characterized.\u003c/p\u003e \u003cp\u003eFurthermore, multiple studies have suggested that activation of microglia promotes brain tumor progression [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Microglia generally polarize into a classical activated microglia phenotype (M1) or an alternative phenotype (M2). The M1 phenotype, is characterized by the production of a variety of pro-inflammatory cytokines, that have anti-tumor effects. By contrast, the M2 microglia function as pro-tumor factors by initiating immunosuppression [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Notably, the gut microbiome promotes the maturation of microglia and facilitates their function in the brain via the GBA [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. It has been postulated that metabolites regulated by the gut microbiota modulate the function of microglia in the progression of BM in NSCLC, but the mechanism remains to be elucidated.\u003c/p\u003e \u003cp\u003eTo address these gaps in knowledge, we analyzed the gut microbial characteristics of 67 NSCLC patients with BM (n\u0026thinsp;=\u0026thinsp;32) and without BM (NBM, n\u0026thinsp;=\u0026thinsp;35) via 16S rRNA gene sequencing. In addition, untargeted metabolomics analyses of plasms samples, liquid chromatography-tandem mass spectrometry (LC-MS/MS) and gas chromatography-mass spectrometry (GC-MS) were performed to determine the metabolomic profiles of 67 NSCLC patients. We envisaged this study as a systematic and comprehensive interrogation of the microbiome and metabolome of NSCLC patients with BM. The objectives of the study were to characterize microbial diversity and metabolite abundance and decipher the association between BM in NSCLC with the gut microbiota and host metabolites. Notably, we found that the metabolite spermine promotes the polarization of microglia to the M2 phenotype, resulting in the development of BM in NSCLC model mice.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv class=\"Section2\" id=\"Sec3\"\u003e\n \u003ch2\u003eStudy design and population\u003c/h2\u003e\n \u003cp\u003eA total of 70 NSCLC patients diagnosed with BM (n\u0026thinsp;=\u0026thinsp;34) and NBM (n\u0026thinsp;=\u0026thinsp;36) by pathological examinations at Wuhan Union Hospital (Wuhan, China) were initially recruited for inclusion in this study. Written informed consent was obtained from all participating patients. The Institutional Revier Board of Huazhong University of Science and Technology approved the study. General clinical data were recorded for all participants, including age, gender, body mass index (BMI), histories of smoking, and serum carcinoembryonic antigen (CEA). All patients were from similar geographic areas and had similar eating habits.\u003c/p\u003e\n \u003cp\u003eThree subjects overall were excluded from the subsequent analyses due to low reads on sequencing of stool samples (n\u0026thinsp;=\u0026thinsp;2), or hemolysis of plasma sample (n\u0026thinsp;=\u0026thinsp;1). Finally, 67 cases were subjected to 16S rRNA gene sequencing and untargeted metabolomics analysis. All samples were collected in accordance with standard operating procedures. A flow chart illustrating the enrollment and analysis processes is shown in Additional file 1: Figure S1.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eDna Extraction And 16s Rrna Gene Sequencing\u003c/h3\u003e\n\u003cp\u003eDNA was extracted from fecal samples using an E.Z.N.A.\u0026reg; soil DNA kit (Omega Bio-tek, Norcross, GA, USA). PCR amplification was performed using the primers 338F (5\u0026apos;-ACTCCTACGGGAGGCAGCAG-3\u0026apos;), 806R (5\u0026apos;-GGACTACHVGGGTWCTAAT-3\u0026rsquo;) directionally targeting the V3 and V4 regions of the 16S rRNA gene. DNA recycling and purification were performed using an AxyPrep DNA Gel Extraction kit (Axygen Biosciences, Union City, CA, USA). An Illumina Miseq PE300 platform was used for sequencing according to the manufacturer\u0026rsquo;s specifications.\u003c/p\u003e\n\u003ch3\u003e16s Rrna Gene Sequencing Data Analysis\u003c/h3\u003e\n\u003cp\u003eSequences exhibiting\u0026thinsp;\u0026gt;\u0026thinsp;97% similarity thresholds were allocated to one operational taxonomic unit (OTU), chimeric sequences were filtered and the alignment threshold was set to 70% for taxonomic identification of species. Alpha diversity was determined to assess the complexity of species diversity for each sample. Beta diversity calculations were subjected to principal coordinate analysis (PCoA) to assess the diversity in samples from different groups in terms of species complexity. The Wilcoxon rank-sum test was used to compare bacterial abundance and diversity. Heat maps were constructed based on the nonparametric Wilcoxon test (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, q\u0026thinsp;\u0026lt;\u0026thinsp;0.1) at the genus level. Linear discriminant analysis (LDA) coupled with effect size (LEfSe) was applied to evaluate differentially abundant taxa. The data were analyzed on the free online platform of Majorbio Cloud Platform (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.majorbio.com\u003c/span\u003e\u003c/span\u003e), from Shanghai Majorbio Bio-pharm Technology Co.,Ltd.\u003c/p\u003e\n\u003ch3\u003eUntargeted Metabolomics Analysis\u003c/h3\u003e\n\u003cp\u003eA total of 67 plasma samples were collected from NSCLC patients and sequenced at Novogene (Beijing, China). LC-MS/MS analyses were conducted using a Vanquish UHPLC system (Thermo Fisher) and Orbitrap Q Exactive series mass spectrometer (Thermo Fisher). The exactive series mass spectrometer was operated in positive/negative polarity mode. Metabolite data were analyzed using Compound Discoverer 3.1 (CD3.1, Thermo Fisher). Metabolites were identified based on peak intensities, normalized to the total spectral intensity. Three databases (mzCloud, mzVault and MassList) were used to obtain accurate qualitative and relative quantitation results. Statistical analyses were performed using the statistical software R (R version R-3.4.3), Python (Python 2.7.6 version), and CentOS (CentOS release 6.6). An Agilent 7890 gas chromatograph system coupled with a Pegasus HT time-of-flight mass spectrometer was used for GC-MS analyses.\u003c/p\u003e\n\u003ch3\u003eUntargeted Metabolomics Data Analysis\u003c/h3\u003e\n\u003cp\u003eLC-MS/MS: KEGG (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.genome.jp/kegg/\u003c/span\u003e\u003c/span\u003e), HMDB (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.hmdb.ca/\u003c/span\u003e\u003c/span\u003e) and Lipidmaps (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.lipidmaps.org/\u003c/span\u003e\u003c/span\u003e) database, were used to annotate the metabolites. Principal components analysis (PCA) and partial least squares discriminant analysis (PLS-DA) were performed using metaX software. Differential metabolites based on by using variable importance in projection (VIP) score\u0026thinsp;\u0026gt;\u0026thinsp;1 and \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, and fold- change (FC)\u0026thinsp;\u0026ge;\u0026thinsp;2 or \u0026le;\u0026thinsp;0.5. The functions of these metabolites and metabolic pathways were analyzed using the KEGG database.\u003c/p\u003e\n\u003cp\u003eGC-MS: Chroma TOF 4.3X software (LECO Corp) and the LECO-Fiehn Rtx5 database were used for raw data analysis. Mass spectrum matches and retention index matches were both considered in metabolite identification.\u003c/p\u003e\n\u003ch3\u003eCell Culture And Reagents\u003c/h3\u003e\n\u003cp\u003eMurine Lewis lung cancer (LLC) cells were purchased from the ATCC. Mouse microglia BV2 cell were obtained from Procell Life Science \u0026amp; Technology Co., Ltd (Wuhan, China). All cell lines were cultured in DMEM supplemented with 10% FBS at 37\u0026deg;C under 5% CO\u003csub\u003e2\u003c/sub\u003e. BV2 cells were treated with spermine (10 \u0026micro;M, #S3948, Selleck Chemicals, China) alone or plus the STAT3 inhibitor stattic (0.5 \u0026micro;M, #HY-13818, Sigma-Aldrich, St. Louis, MO, USA) for 24 h to observe changes in the BV2 phenotype.\u003c/p\u003e\n\u003ch3\u003eStudy Animals And Bm Model\u003c/h3\u003e\n\u003cp\u003eSix-week-old female mice (C57BL/6J; Beijing Vital River Laboratory Animal Technology Co. Ltd, China) were used. The BM model was described previously [\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e]. LLC cells labeled with luciferase (LLC-Luc, 4 \u0026times; 10\u003csup\u003e5\u003c/sup\u003e in 0.1 mL PBS) were slowly injected into the intracarotid artery of mice. Two weeks later, bioluminescence imaging was performed to observe the growth of brain tumors using an IVIS Lumina imaging system (In Vivo FX PRO, Bruker Corp.) after mice were injected with D-luciferin intraperitoneally (150 mg/kg, Goldbio St. Louis, MO, USA).\u003c/p\u003e\n\u003cp\u003eBM mice received spermine (10 mg/kg, Selleck) 1 week after injection of tumor cells, by intraperitoneal injection every 3 days until the end of the experiment. Clodronate liposomes (F70101C-A-2, FormuMax, Dakewe Biotechnology Co., Ltd, China) was injected intracerebroventricularly using a stereotactic apparatus (10 \u0026micro;L/mouse). All animal study procedures followed the guidelines of the Institutional Animal Care Committee of Tongji Medical College, Huazhong University of Science and Technology, China.\u003c/p\u003e\n\u003ch3\u003eReal-time Quantitative Pcr\u003c/h3\u003e\n\u003cp\u003eTotal RNA samples were prepared using a total RNA extraction kit (Omega Bio-Tek, China). A PrimeScript RT reagent kit (Takara, Beijing, China) was uesd for reverse transcription. For real-time RT-PCR, SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) and primers (Additional file 2: Table S1) were used, with transcript levels determined according to the 2 \u003csup\u003e\u0026minus;\u0026Delta;\u0026Delta;Ct\u003c/sup\u003e method.\u003c/p\u003e\n\u003ch3\u003eWestern Blotting\u003c/h3\u003e\n\u003cp\u003eProtein was extracted from tumor cells or tissues using 1\u0026times; Cell Lysis Buffer (Promega). Western blotting was performed as described previously [\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e], using antibodies specific for STAT3 (ab68153, Abcam), p-STAT3 (ab76315, Abcam), iNOS (inducible nitric oxide synthase) (ab178945, Abcam), Ym-1 (chitinase-like 3) (ab192029, Abcam), and GAPDH (AC002, ABclonal, China).\u003c/p\u003e\n\u003ch3\u003eFlow Cytometry\u003c/h3\u003e\n\u003cp\u003eTo assay the populations of microglia in the brain, mouse brains were cut into small pieces, kept in PBS on ice, minced, and digested for 1 h with 0.5 mg/mL collagenase type IV and DNase I 20 U/mL at 37\u0026deg;C. The details of the protocol mainly based on reference[\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e]. The cells were separated by Percoll (GE Healthcare Life Sciences) using a density gradient (70%, 37%, 30%), and stained with FITC-conjugated anti-CD11b antibodies (101206), Zombie NIR\u0026trade; Fixable Viability kit (423106), CD45 Brilliant Violet 510 (109837), CD86 PE (159204), CD206 Brilliant Violet 421 (141717). These antibodies were purchased from BioLegend, San Diego, CA.\u003c/p\u003e\n\u003ch3\u003eImmunofluorescence\u003c/h3\u003e\n\u003cp\u003eMice were anesthetized to obtain brain tissue, then processed and embedded in paraffin, the details of the protocol have been described previously[\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e]. Brain tissue slides were incubated with the primary antibodies anti-arginase-1 (Arg-1) (1:200, Proteintech, Wuhan, China) or anti-ionized calcium-binding adapter molecule-1 (Iba1; 1:200, Abcam) overnight at 4℃, followed by appropriate secondary antibody conjugated with fluorescent dye Alexa Fluor 488 or Alexa Fluor 594 (1:1,000, Invitrogen). Sections were counterstained with 4,6-diamidino-2-phenylindole (DAPI) (10 \u0026micro;g/mL, Sigma-Aldrich) and then scanned with a digital camera and imaging software (3DHISTECH Ltd., Hungary).\u003c/p\u003e\n\u003ch3\u003ePhagocytosis Assay\u003c/h3\u003e\n\u003cp\u003eBV2 cells were treated with spermine (10 \u0026micro;M) alone or with the STAT3 inhibitor stattic (0.5 \u0026micro;M) for 24 h to observe the changes in BV2 phagocytic ability. Phagocytosis assay were performed as previously described[\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e]. The BV2 cells were observed under a confocal microscope (Olympus, Tokyo, Japan) and analyzed using flow cytometry (Becton Dickinson, Franklin Lakes, NJ, USA).\u003c/p\u003e\n\u003ch3\u003eStatistical Analyses\u003c/h3\u003e\n\u003cp\u003eThe predicted functional composition profiles were collapsed into level-3 KEGG pathways based on 16S rRNA gene sequences using PICRUSt. Correlations between the clinical parameters were calculated using Spearman rank correlation and presented using a heatmap. PCA was performed to examine intrinsic clusters within the metabolomics data. A 95% confidence interval was used as the threshold to identify potential outliers in all samples. PCoA was used to compare the gut microbiome profiles between the two groups using ANOSIM based on Bray-Curtis distance. A correlation matrix between the plasma metabolites and gut bacterial species was generated using Pearson\u0026rsquo;s correlation coefficient. Statistical analyses were carried out using SPSS version 18 (SPSS, Chicago, IL, USA). All continuous variables such as bacterial \u0026alpha;-diversity, age, and body mass index (BMI) was presented as the mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SEM, unless otherwise indicated, and difference between groups were compared using the student\u0026rsquo;s \u003cem\u003et\u003c/em\u003e-test. The statistical significance level was set at \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eCharacteristics of the patients\u003c/h2\u003e \u003cp\u003eTo investigate the gut microbiota and metabolites in BM and NBM patients, a total of 67 subjects (BM, n\u0026thinsp;=\u0026thinsp;32 and NBM, n\u0026thinsp;=\u0026thinsp;35) remained for the final statistical analyses (Additional file 1, Figure S1). The common clinical characteristics of these 67 NSCLC patients are shown in Additional file 2, Table S2. No difference was observed in terms of sex, age, BMI, smoking history, or CEA levels between BM and NBM patients (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). These data suggest that a more valuable biomarker for predicting the occurrence of BM in NSCLC patients is needed.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eChanges In The Gut Microbiome Between The Bm And Nbm Group\u003c/h3\u003e\n\u003cp\u003eTo identify the gut microbiome features in NSCLC patients with BM or NBM, we investigated the composition of the gut microbiome in the 67 final patients based on 16S rRNA gene sequences. The species rarefaction curve for all samples supported the adequacy of the sampling efforts (Additional file 3: Figure S2a). No significant differences were observed in the richness and diversity of the gut microbiota between the NBM and BM group as measured using the Sob index (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.4037) or Shannon index (P\u0026thinsp;=\u0026thinsp;0.3433) at the genus level (Fig.\u0026nbsp;1a-b). Venn diagram was constructed to show the overlapping OTU data of the two groups, and revealed that 1840 of the 2978 OTUs were common in the BM and NBM patient samples (Additional file 3: Figure S2b). Furthermore, to assess the overall structure of the gut microbiota, the beta diversity in microbial composition was found no obvious differences in the BM group with the NBM group (P\u0026thinsp;=\u0026thinsp;0.311) (Fig.\u0026nbsp;1c).\u003c/p\u003e \u003cp\u003eAs gut dysbiosis is reportedly involved in the progression of brain tumors [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e], four phyla were examined to assess the relative proportions of dominant taxa using sunburst plots (Additional file 3: Figure S2c). The \u003cem\u003eFirmicutes/Bacteroidetes\u003c/em\u003e ratio (F/B) (4.36 verse 4.13) and the content of \u003cem\u003eProteobacteria\u003c/em\u003e (14.9% verse 9.51%), were higher in the NBM group compared with the BM group, indicating the significantly gut dysbiosis in NBM and BM groups[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. Furthermore, to explore the correlation of microbial abundance with common clinical indexes, sex, smoking, BMI and age were included in a Spearman\u0026rsquo;s correlation analysis. The results showed that sex and smoking were significantly correlated with some bacteria genera, whereas BMI and age exhibited no remarkable correlation of bacterial abundance (Additional file 3: Figure S2d).\u003c/p\u003e \u003cp\u003eThe top 20 genera with significant differences were exhibited in two groups, \u003cem\u003eAlistipes\u003c/em\u003e、\u003cem\u003eKlebsiella\u003c/em\u003e and \u003cem\u003eunclassified_f_Enterobacteriaceae\u003c/em\u003e were significantly more abundant in patients with BM compared with NBM patients, however, the levels of \u003cem\u003eBlautia\u003c/em\u003e and \u003cem\u003eMegamonas\u003c/em\u003e were decreased in the BM group (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, Fig.\u0026nbsp;1d). Moreover, the individual taxa exhibiting a differential abundance between the subject groups was selected via LEfSe analysis in Fig.\u0026nbsp;1e (LDA score\u0026thinsp;\u0026gt;\u0026thinsp;2). Notably, we identified 18 discriminatory genera as key discriminants in the two groups, \u003cem\u003eunclassified_f_Enterobacteriaceae\u003c/em\u003e and \u003cem\u003eBlautia\u003c/em\u003e were in greater abundance in the BM group and NBM group, respectively. These data suggest that the differentially abundant microbiota in BM and NBM patients could potentially serve as a biomarker of BM in NSCLC.\u003c/p\u003e \u003cp\u003eTo characterize the functional changes in the gut microbiota of NSCLC patients with BM, we predicted the functional composition profiles using 16S rRNA gene sequencing data analyzed using PICRUSt. COG function classification statistics box charts indicated that the pathways of carbohydrate transport and metabolism, and amino acid transport and metabolism were the top two most abundant pathways (Additional file 3: Figure S2e). A total of 239 KEGG (level-3) pathways involved in BM and NBM groups (Additional file 2: Table S3). Amino acid metabolism is reportedly as an important factor in nucleotide production and tumor cell proliferation in highly aggressive BM [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Thus, we hypothesized that gut dysbiosis is correlated with BM development via amino acid metabolism pathways.\u003c/p\u003e\n\u003ch3\u003eChanges In The Plasma Metabolome Between The Bm And Nbm Groups\u003c/h3\u003e\n\u003cp\u003eAs circulating metabolites considered pivotal mediators of communication between the gut microbiota and the brain [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], the untargeted metabolomic were performed via LC-MS/MS and GC-MS to clarify the microbe-host interactions in BM of NSCLC. The LC-MS/MS metabolite profiling data showed that BM patients were significantly separated from NBM patients in the PCA (Additional file 4: Figure S3a) and PLS-DA models (neg: R2Y 0.60; Q2Y 0.40; pos: R2Y0.65; Q2Y 0.31) (Fig.\u0026nbsp;2a). In the model of GC-MS data, the two groups were also clearly distinguishable in the PCA (Additional file 4: Figure S3b) and PLS-DA score plot (R2Y 0.84; Q2Y 0.78) (Fig.\u0026nbsp;2b). In addition, PLS-DA validated model performed in Additional file 4: Figure S3c-d, in the LC-MS/MS model and GC-MS model, with R2Y is greater than Q2Y. These data confirmed that the plasma metabolites differed significantly between the BM and NBM groups.\u003c/p\u003e \u003cp\u003eFinally, we identified 62 differed metabolites in the LC-MS/MS models, with 26 were upregulated, and 36 downregulated in the BM group (Fig.\u0026nbsp;2c). Also, in the GC-MS model, 14 metabolites differed between groups, including 11 upregulated and 3 downregulated in the BM group (Fig.\u0026nbsp;2d).\u003c/p\u003e \u003cp\u003eTo discriminate BM from NBM patients via the potential plasma metabolites, the clinical significance of different metabolites was calculated using random forest analysis (Additional file 2: Table S4). The KEGG database was used to analyze the pathways of the metabolites that differed between BM and NBM patients (Fig.\u0026nbsp;2e-f). The main pathways that associated with BM included biosynthesis of secondary metabolites, glutathione metabolism, arginine and proline metabolism in the LC-MS/MS model and oxidative phosphorylation in the GC-MS model. Taken together, these results suggest that the metabolite profiles of BM patients differed significantly from those of NBM patients. Therefore, we hypothesize that host metabolites are modulated by the gut microbiota to regulate the development of BM in NSCLC via an as yet unidentified pathway.\u003c/p\u003e \u003cp\u003e \u003cb\u003eMulti-omic network analysis reveals the relationship between the gut microbiota and serum metabolites in BM of NSCLC patients\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo explore the functional correlation between gut microbiota changes and plasma metabolite perturbations in BM of NSCLC patients, the top 16 differential genera and top 20 metabolites were selected for matrix analysis in the model of LC-MS/MS data (Fig.\u0026nbsp;3a-b and Additional file 2: Tables S5-6), and top 14 metabolites in the model of GC-MS data (Fig.\u0026nbsp;3c and Additional file 2: Table S7). The results showed that gut-microbiota was positively or negatively correlated with several host-metabolites in NSCLC patients. Pearson correlation coefficients were calculated, with |rho | \u0026ge; 0.4 and \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026le;\u0026thinsp;0.05 to explore the correlation of metabolites and gut bacteria. As shown in Fig.\u0026nbsp;3d, metabolite Com_211_pos was positively correlated with \u003cem\u003eIntestinimonas\u003c/em\u003e, whereas metabolite Com_4055_pos was negatively correlated with \u003cem\u003eHoldemania\u003c/em\u003e, and metabolite Com_867_neg was positively correlated with \u003cem\u003eAnaerotruncus\u003c/em\u003e in the LC-MS/MS model. In the model of GC-MS, metabolite Com_206 was positively correlated with \u003cem\u003eHoldemania\u003c/em\u003e (Fig.\u0026nbsp;3e), whereas metabolite Com_113 was positively correlated with and \u003cem\u003eAnaerotruncus\u003c/em\u003e and \u003cem\u003e[Eubacterium]_nodatum_group\u003c/em\u003e (Additional file 4: Figure S3e).\u003c/p\u003e \u003cp\u003eMoreover, arachidic acid (Com_867_neg), spermine (Com_211_pos) and 3-methyl glutaric acid (Com_206) were selected with higher abundances in the BM group, with the corresponding AUC values for the receiver operating characteristic (ROC) curves were 0.746, and 0.730, and 0.846, respectively (Fig.\u0026nbsp;4a), indicating that each of these metabolites could be used as a potential signature of BM progression in NSCLC. Furthermore, the ROC curve analysis indicated that the combination of the three metabolites (AUC, 0.9232; p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) was significantly associated with BM in NSCLC (Fig.\u0026nbsp;4b), suggesting that a three-metabolite model could accurately distinguish BM patients from NBM controls. Notably, the metabolite spermine functions in arginine and proline metabolism, consistent with the pathway predicted by the 16S rRNA gene profiling data. Therefore, spermine was deemed an important potential metabolite involved in the progression of BM in NSCLC.\u003c/p\u003e \u003cp\u003eSpermine was positively correlated with \u003cem\u003eIntestinimonas\u003c/em\u003e, and that the combination of the AUC value was 0.7525, which indicated the biomarker are predictive (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0004) (Fig.\u0026nbsp;4c). We hypothesized that the gut microbiota affects BM progression by interacting with certain host metabolites, metabolite spermine was modulated by \u003cem\u003eIntestinimonas\u003c/em\u003e potentially, would be useful for discriminating BM and NBM patients. Subsequently, we recruited another 42 NSCLC patients as an independent validation cohort, consisting of 18 BM and 24 NBM patients (Additional file 2: Table S8). The plasma of NSCLC patients was collected for spermine analysis, revealed that the level of spermine was significantly increased in the BM group compared with the NBM group (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.008) (Fig.\u0026nbsp;4d), consist with previous observations from discovery cohorts. Consistently, as indicated by the AUC value of spermine with 0.7014 (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0209) (Fig.\u0026nbsp;4e), further confirming spermine as a potential indicator for predicting BM in NSCLC.\u003c/p\u003e\n\u003ch3\u003eSpermine Promotes Bm In Nsclc Via Polarizing Microglia To The M2 Phenotype\u003c/h3\u003e\n\u003cp\u003eA BM mouse model was established to explore the function of spermine. As shown in Fig.\u0026nbsp;5a, the size and the occurrence of brain tumors in spermine-treated (10mg/kg) mice increased remarkably compared with the control group (Fig.\u0026nbsp;5b-c). Microglia was reportedly is associated with brain tumor progression, with the M2 phenotype acts as pro-tumor factor by immunosuppression [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Moreover, spermine reportedly promotes macrophage M2 polarization [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Therefore, we hypothesized that spermine regulates the development of BM by promoting M2 polarization of microglia. Immunofluorescence analyses showed that abundant activated microglia (Iba1\u003csup\u003e+\u003c/sup\u003e) infiltrated brain metastatic lesions, and the number of M2 microglia (Arg-1\u003csup\u003e+\u003c/sup\u003e) was increased in the spermine group, suggesting that spermine promotes microglia to polarize to the M2 phenotype (Fig.\u0026nbsp;5d). In addition, we analyzed the microglia in brain tumor tissue using flow cytometry (Fig.\u0026nbsp;5e). These results showed that the percent of CD45\u003csup\u003elow\u003c/sup\u003eCD11b\u003csup\u003e+\u003c/sup\u003e CD206\u003csup\u003e+\u003c/sup\u003e (M2) microglia was significantly increased, whereas that of CD45\u003csup\u003elow\u003c/sup\u003eCD11b\u003csup\u003e+\u003c/sup\u003e CD86\u003csup\u003e+\u003c/sup\u003e (M1) microglia was decreased in spermine-treated mice compared with control mice (Fig.\u0026nbsp;5f-g).\u003c/p\u003e \u003cp\u003eThese results suggest that spermine promotes BM by promoting the polarization of microglia toward the M2 phenotype. To verify this observation, we depleted microglial cells via intracerebroventricular injection of clodronate liposomes (10 \u0026micro;L/mouse) in BM model mice [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Notably, clodronate liposomes significantly suppressed spermine-induced BM (Fig.\u0026nbsp;5h-i). These results suggest that spermine-induced BM is indeed mediated by the polarization of microglia.\u003c/p\u003e \u003cp\u003e \u003cb\u003eSpermine-induced M2 microglia polarization and suppression of phagocytic ability via activation of the STAT3 pathway.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo further investigate whether spermine regulates microglia polarization directly, we measured the expression of various M1/M2 markers under spermine (10 \u0026micro;M) treatment of cultured mouse BV2 cells. qRT-PCR and Western blot revealed that the M2 markers were greatly increased in the spermine group compared with the control group, whereas M1 markers were significantly decreased (Fig.\u0026nbsp;6a-b). Thus, verified that spermine promotes the polarization of microglia to the M2 phenotype, which exerts pro-tumor effects. Studies have reported that activation of the STAT3 pathway is involved in M2 polarization [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] and related to the promotion of immunosuppression, leading to pro-tumor effects [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. To further elucidate the underlying mechanism by which spermine promotes M2 polarization of microglia, phosphorylation of STAT3 was analyzed, which revealed that spermine indeed enhanced the expression of p-STAT3 (Figure. 6c). In addition, activation of STAT3 in the spermine group was significantly suppressed, and M2 polarization of microglia was also significantly suppressed by using STAT3 inhibitor static (0.5 \u0026micro;m) (Fig.\u0026nbsp;6d). Moreover, inhibition of STAT3 reversed the spermine-mediated up-regulation of M2 CD206\u0026thinsp;+\u0026thinsp;cells (Fig.\u0026nbsp;6e-f), suggesting that spermine promotes M2 microglial polarization via activation of the STAT3 pathway. These results showed that spermine enhances the development of BM in NSCLC by promoting microglia M2 polarization, and the STAT3 pathway might be involved in this process.\u003c/p\u003e \u003cp\u003eThe phagocytic ability of microglia, which exhibit anti-tumor activity, was examined using a fluorescent microsphere phagocytosis assay and flow cytometry. Microglia showed strong phagocytic activity, whereas spermine significantly compromised this activity. Notably, the down-regulation of the phagocytic ability of microglia induced by spermine was suppressed in the stattic group (Fig.\u0026nbsp;6h-k). The above results suggest that spermine promotes microglia M2 polarization and suppresses the phagocytic ability of microglia via activation of the STAT3 pathway, which in turn promotes BM progression.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eBM is a severe complication of advanced NSCLC and accompanied by low quality of life and poor prognosis [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Clarification of the underlying mechanism of the development of BM in NSCLC is the cornerstone of improving the overall survival of patients. Accumulating evidence suggests that the gut microbiota contributes to brain tumor development and to the success of therapy via GBA [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Dysregulation of the gut microbiota composition and function is associated with disorders of host metabolism, altering circulating metabolites is particularly effective way to regulate the brain microenvironment [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Characterization of the gut microbiota and metabolites in NSCLC patients with BM could facilitate the development of biomarkers that could lead to beneficial approaches for therapy. Thus, in our study, we conducted a global untargeted metabolomic analysis of plasma samples in combination with 16S rRNA gene sequencing of fecal sample, to decipher the connection between gut bacterial and host metabolism and generate the basis for novel treatment for BM in NSCLC patients.\u003c/p\u003e \u003cp\u003eAn imbalance among gut bacteria is a common hallmark of numerous human diseases [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e], such as low diversity, an increased ratio of \u003cem\u003eFirmicutes/Bacteroidetes\u003c/em\u003e, and increased levels of \u003cem\u003eProteobacteria\u003c/em\u003e in the gut microbiota [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. We identified 16 bacterial genera in the two groups that differed significantly in abundance. Interestingly, the genera \u003cem\u003eKlebsiella\u003c/em\u003e, \u003cem\u003eEnterobacter\u003c/em\u003e, and \u003cem\u003eunclassified_f__Enterobacteriaceae\u003c/em\u003e, which belong to \u003cem\u003eProteobacteria\u003c/em\u003e phylum, were more abundant in the BM group, suggesting associated with the development of BM.\u003c/p\u003e \u003cp\u003eIn addition, genera such as \u003cem\u003eBlautia\u003c/em\u003e were reduced the BM group, and these organisms are a source of short-chain fatty acids (SCFA) that relieve inflammation [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. The genera \u003cem\u003eAlistipes\u003c/em\u003e and \u003cem\u003eKlebsiella\u003c/em\u003e, as the classic pathogenic bacteria [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e], were shown increased in BM group. Notably, the \u003cem\u003eintestinimona\u003c/em\u003e genus, belongs to the \u003cem\u003eRuminococcaceae\u003c/em\u003e family, which is associated with the transport of secondary metabolites such as secondary bile acids, was higher in the BM group, indicating that gut dysbiosis might affect the host disease status by altering metabolite levels. Therefore, the tumor-promoting effects of the microbiome in NSCLC patients with BM could derive from holistic dysbiosis and hypothesized that the gut microbiota affects the brain via the GBA, in part by altering circulating metabolites.\u003c/p\u003e \u003cp\u003eUsing LC-MS/MS and GC/MS approach, our study identified 76 differentially abundant metabolites between the BM and NBM groups, primarily consisting of lipids and lipid-like molecules (e.g., arachidic acid, 3-Methylglutaric Acid and hexadecanedioic acid), organic nitrogen compounds (e.g., spermine, choline and phosphocholine), and organic acids and their derivatives. Perturbations in lipid metabolism can affect the progression of cancers through effects on the stem cell character of tumor cells, angiogenesis, and immune surveillance [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. The metabolites associated with the tumor immune response and thus affect the relationship between the gut microbiota and disease [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. In our study, spermine, which belongs to polyamines, have been attributed to human cancers [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e], was increased in the BM group compared with the NBM group. In addition, the increased concentration of arachidic acid in BM patients, which may be associated with nonalcoholic steatohepatitis [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e], is worthy of further exploration. We also detected altered levels of benzenoids and organoheterocyclic compounds in plasma samples. These compounds exert anti-inflammatory activity [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e] and are related to ameloblastoma biology, respectively. In general, the untargeted plasma metabolomic showed unique and differential metabolic signatures in BM and NBM patients.\u003c/p\u003e \u003cp\u003eUsing an integrated metabolomics and 16S rRNA gene sequencing approach, we identified the metabolite spermine having a positive correlation with the \u003cem\u003eintestinimonas\u003c/em\u003e genus in the BM group. Moreover, ROC curve analysis showed that the model including \u003cem\u003eintestinimonas\u003c/em\u003e and spermine is a potential biomarker for predicting the risk of BM in NSCLC patients. An important question is how dose spermine participate in the development of the disease? Spermine is metabolite that is reportedly from both the gut microbiota and the host, and regulates the innate immune response by affecting macrophage activation in the context of host inflammation and carcinogenesis by impairing M1 macrophage responses and inducing an M2-like state [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Microglia are macrophages found in the brain, and their activation affects the microenvironment of brain tumors [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. Thus, we hypothesize that spermine promotes the development of BM by activating microglia and inducing polarization to the M2 phenotype, exerting pro-tumor effects. Depletion of microglia via intracerebral injection of clodronate liposomes supported this hypothesis. Furthermore, STAT3 activation also reportedly stimulates microglia polarization to the M2 phenotype, leading to the development of an immunosuppressive microenvironment that promotes brain tumor growth [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. Interestingly, extracellular spermine would be able to directly modulate microglial via the polyamine transport system, which is involved in the innate immune system [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. Our study illustrated that spermine modulates the polarization of microglia toward the M2 phenotype and suppress the phagocytic ability of these cells by increasing the phosphorylation of STAT3.\u003c/p\u003e \u003cp\u003eThere are several limitations to our study, including the relatively small sizes of the patient cohorts in the multi-omic analysis. Thus, multi-center studies of NSCLC patients with/without BM will be needed to further validate our study\u0026rsquo;s findings. Notably, target bacteria also required further verification. In addition, several unknown metabolites that could not be identified in the currently available databases were also found to be significantly associated with several bacterial genera. Further investigations of the effects of the spermine synthesis and degradation pathways are also warranted. Lipid metabolism, including that of arachidic acid and 3-methyl glutaric acid, deserve more attention in future studies as well in order to explore their potential for use as biomarkers in BM patients. We cannot rule out the possibility that spermine activates microglia via other pathways in brain tumor development, and this possibility deserves in-depth study. Despite these shortcomings, our results demonstrate that the combination of spermine and \u003cem\u003eintestinimonas\u003c/em\u003e represents a potential therapeutic target for treating BM in NSCLC patients. Spermine are potential biomarkers for discriminating BM and NBM, pending further validation studies.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eWe demonstrated the signature microbiota and metabolites that distinguish BM and NBM patients. However, the microbial diversity did not differ significantly compared with the NBM group. Multi-omics analysis revealed a correlation between the differentially abundant genera and various metabolites, and we filtered out three combinations, including spermine and \u003cem\u003eintestinimonas\u003c/em\u003e, arachidic acid and \u003cem\u003eCandidatus_Soleaferrea\u003c/em\u003e, 3-methyl glutaric acid and \u003cem\u003eHoldemania\u003c/em\u003e, suggesting that gut microbiota promote the development of BM in NSCLC in part through altering the metabolome. We identified the \u003cem\u003eintestinimonas\u003c/em\u003e genus in NSCLC patients with BM as key commensal intestinal bacteria positively correlated with the plasma metabolite spermine, promoting brain tumor growth. In addition, the model of \u003cem\u003eintestinimonas\u003c/em\u003e and spermine is a potential biomarker for predicting the risk of BM in NSCLC. Notably, spermine was identified as a key plasma metabolite that act through STAT3 signaling to promote microglia M2 polarization, exerting a pro-tumor effect in NSCLC patients with BM. In summary, our findings provide novel insights for future investigations into the causal associations between gut dysbiosis and host metabolites in the development and progression of BM in NSCLC. We identified spermine as a potential therapeutic target that could prove useful in the prevention, diagnosis and treatment of NSCLC with BM.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eBM: Brain metastasis; NSCLC: non-small cell lung cancer; GBA: gut-brain-axis; LC-MS/MS: Liquid chromatography-tandem mass spectrometry; GC/MS: Gas chromatography-mass spectrometry; STAT3: Signal Transducer and Activator of Transcription 3; CEA: serum carcinoembryonic antigen; PCoA: principal coordinate analysis; LDA: Linear discriminant analysis; LEfSe: LDA effect size; PCA: Principal components analysis; PLS-DA: Partial least squares discriminant analysis; LLC: Lewis lung cancer cell lines; LLC-Luc: luciferase-labeled LLC; Iba1: ionized calcium binding adapter molecule-1; Arg-1: arginase-1; iNOS: inducible nitric oxide synthase; Ym-1: chitinase 3-like 3; CD206: mannose receptor; TGF-\u0026beta;: transforming growth factor \u0026beta;; RF: random forest; neg: negative; pos: positive; SCFA: short-chain fatty acids \u003c/p\u003e\n"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Institutional Review Board of Huazhong University of Science and Technology. Written informed consent was obtained from all legal guardians of the patients. All animal experiments were conducted in agreement with the Guide for the Care and Use of Laboratory Animals and were approved by the Committee on Animal Handling of Huazhong University of Science and Technology.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors confirm that the data supporting the findings of this study are available within the article or the supplementary materials.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; Disclosures\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo disclosures were reported.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research was supported by the grants from the National Key R\u0026amp;D\u003c/p\u003e\n\u003cp\u003eProgram of China (Grant No 2019YFC1316205, Grant No.\u0026nbsp;2019YFC1316205)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eXRD and NY conceived the study. HHL, LCL, and HZ performed the experiments. HHL, YWB, HZ and RGZ collected clinical samples. HHL, FT and JJW analyzed the data. XRD, FT and HHL wrote the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eErnani, V. and T.E. Stinchcombe, Management of Brain Metastases in Non-Small-Cell Lung Cancer. J Oncol Pract, 2019. 15;11;563-570.\u003c/li\u003e\n\u003cli\u003eKhalifa, J., A. Amini, S. Popat, L.E. Gaspar, and C. Faivre-Finn, Brain Metastases from NSCLC: Radiation Therapy in the Era of Targeted Therapies. J Thorac Oncol, 2016. 11;10;1627-43.\u003c/li\u003e\n\u003cli\u003eHall, A.B., A.C. Tolonen, and R.J. Xavier, Human genetic variation and the gut microbiome in disease. Nat Rev Genet, 2017. 18;11;690-699.\u003c/li\u003e\n\u003cli\u003eDurack, J. and S.V. Lynch, The gut microbiome: Relationships with disease and opportunities for therapy. J Exp Med, 2019. 216;1;20-40.\u003c/li\u003e\n\u003cli\u003eMehrian-Shai, R., J.K.V. Reichardt, C.C. Harris, and A. Toren, The Gut-Brain Axis, Paving the Way to Brain Cancer. Trends Cancer, 2019. 5;4;200-207.\u003c/li\u003e\n\u003cli\u003eCryan, J.F., K.J. O\u0026apos;Riordan, C.S.M. Cowan, K.V. Sandhu, T.F.S. Bastiaanssen, M. Boehme, et al., The Microbiota-Gut-Brain Axis. Physiol Rev, 2019. 99;4;1877-2013.\u003c/li\u003e\n\u003cli\u003eNicholson, J.K., J.K. Elaine Holmes, Remy Burcelin, Glenn Gibson, Wei Jia, and S. Pettersson, Host-gut microbiota metabolic interactions. Science, 2012 33;6086;1262-1267.\u003c/li\u003e\n\u003cli\u003eRothhammer, V., I.D. Mascanfroni, L. Bunse, M.C. Takenaka, J.E. Kenison, L. Mayo, et al., Type I interferons and microbial metabolites of tryptophan modulate astrocyte activity and central nervous system inflammation via the aryl hydrocarbon receptor. Nat Med, 2016. 22;6;586-97.\u003c/li\u003e\n\u003cli\u003evan de Wouw, M., M. Boehme, J.M. Lyte, N. Wiley, C. Strain, O. O\u0026apos;Sullivan, et al., Short-chain fatty acids: microbial metabolites that alleviate stress-induced brain-gut axis alterations. 2018. 596;20;4923-4944.\u003c/li\u003e\n\u003cli\u003eMatcovitch-Natan, O., D.R. Winter, A. Giladi, S. Vargas Aguilar, A. Spinrad, S. Sarrazin, et al., Microglia development follows a stepwise program to regulate brain homeostasis. Science, 2016. 353;6301;aad8670.\u003c/li\u003e\n\u003cli\u003eGutmann, D.H. and H. Kettenmann, Microglia/Brain Macrophages as Central Drivers of Brain Tumor Pathobiology. Neuron, 2019. 104;3;442-449.\u003c/li\u003e\n\u003cli\u003eWu, S.Y., F. Xing, S. Sharma, K. Wu, A. Tyagi, Y. Liu, et al., Nicotine promotes brain metastasis by polarizing microglia and suppressing innate immune function. J Exp Med, 2020. 217;8.\u003c/li\u003e\n\u003cli\u003eWolf, S.A., H.W. Boddeke, and H. Kettenmann, Microglia in Physiology and Disease. Annu Rev Physiol, 2017. 79;619-643.\u003c/li\u003e\n\u003cli\u003eSherwin, E., K. Rea, T.G. Dinan, and J.F. Cryan, A gut (microbiome) feeling about the brain. Curr Opin Gastroenterol, 2016. 32;2;96-102.\u003c/li\u003e\n\u003cli\u003eDinan, T.G. and J.F. Cryan, The Microbiome-Gut-Brain Axis in Health and Disease. Gastroenterol Clin North Am, 2017. 46;1;77-89.\u003c/li\u003e\n\u003cli\u003eWei, C., X. Dong, H. Lu, F. Tong, L. Chen, R. Zhang, et al., LPCAT1 promotes brain metastasis of lung adenocarcinoma by up-regulating PI3K/AKT/MYC pathway. J Exp Clin Cancer Res, 2019. 38;1;95.\u003c/li\u003e\n\u003cli\u003eMartin, E., M. El-Behi, B. Fontaine, and C. Delarasse, Analysis of Microglia and Monocyte-derived Macrophages from the Central Nervous System by Flow Cytometry. J Vis Exp, 2017;124.\u003c/li\u003e\n\u003cli\u003eLi, H., F. Tong, R. Meng, L. Peng, J. Wang, R. Zhang, et al., E2F1-mediated repression of WNT5A expression promotes brain metastasis dependent on the ERK1/2 pathway in EGFR-mutant non-small cell lung cancer. Cell Mol Life Sci, 2021. 78;6;2877-2891.\u003c/li\u003e\n\u003cli\u003eWang, J., H. Pan, Z. Lin, C. Xiong, C. Wei, H. Li, et al., Neuroprotective Effect of Fractalkine on Radiation-induced Brain Injury Through Promoting the M2 Polarization of Microglia. Mol Neurobiol, 2021. 58;3;1074-1087.\u003c/li\u003e\n\u003cli\u003eMariat, D., O. Firmesse, F. Levenez, V.D. Guimarăes, H. Sokol, J. Dor\u0026eacute;, et al., The Firmicutes/Bacteroidetes ratio of the human microbiota changes with age. BMC Microbiology, 2009. 9;1;123.\u003c/li\u003e\n\u003cli\u003eNgo, B., E. Kim, and V. Osorio-Vasquez, Limited Environmental Serine and Glycine Confer Brain Metastasis Sensitivity to PHGDH Inhibition. 2020. 10;9;1352-1373.\u003c/li\u003e\n\u003cli\u003eRooks, M.G. and W.S. Garrett, Gut microbiota, metabolites and host immunity. Nat Rev Immunol, 2016. 16;6;341-52.\u003c/li\u003e\n\u003cli\u003eRoesch, S., C. Rapp, S. Dettling, and C. Herold-Mende, When Immune Cells Turn Bad-Tumor-Associated Microglia/Macrophages in Glioma. Int J Mol Sci, 2018. 19;2.\u003c/li\u003e\n\u003cli\u003eGuldner, I.H., Q. Wang, L. Yang, S.M. Golomb, Z. Zhao, J.A. Lopez, et al., CNS-Native Myeloid Cells Drive Immune Suppression in the Brain Metastatic Niche through Cxcl10. Cell, 2020. 183;5;1234-1248 e25.\u003c/li\u003e\n\u003cli\u003eLatour, Y.L., A.P. Gobert, and K.T. Wilson, The role of polyamines in the regulation of macrophage polarization and function. Amino Acids, 2020. 52;2;151-160.\u003c/li\u003e\n\u003cli\u003eTodd, L., I. Palazzo, L. Suarez, X. Liu, L. Volkov, T.V. Hoang, et al., Reactive microglia and IL1beta/IL-1R1-signaling mediate neuroprotection in excitotoxin-damaged mouse retina. J Neuroinflammation, 2019. 16;1;118.\u003c/li\u003e\n\u003cli\u003eQin, C., W.H. Fan, Q. Liu, K. Shang, M. Murugan, L.J. Wu, et al., Fingolimod Protects Against Ischemic White Matter Damage by Modulating Microglia Toward M2 Polarization via STAT3 Pathway. Stroke, 2017. 48;12;3336-3346.\u003c/li\u003e\n\u003cli\u003eJones, L.M., M.L. Broz, J.J. Ranger, J. Ozcelik, R. Ahn, D. Zuo, et al., STAT3 Establishes an Immunosuppressive Microenvironment during the Early Stages of Breast Carcinogenesis to Promote Tumor Growth and Metastasis. Cancer Res, 2016. 76;6;1416-28.\u003c/li\u003e\n\u003cli\u003eAndratschke, N., J. Kraft, C. Nieder, R. Tay, R. Califano, R. Soffietti, et al., Optimal management of brain metastases in oncogenic-driven non-small cell lung cancer (NSCLC). Lung Cancer, 2019. 129;63-71.\u003c/li\u003e\n\u003cli\u003eBacha, S., H. Cherif, D. Rabaa, S. Habibech, S. Cheikhrouhou, H. Racil, et al., Brain metastases of non-small cell lung cancer: prognostic factors and management. Tunis Med, 2018. 96;3;165-171.\u003c/li\u003e\n\u003cli\u003eRoager, H.M. and T.R. Licht, Microbial tryptophan catabolites in health and disease. Nat Commun, 2018. 9;1;3294.\u003c/li\u003e\n\u003cli\u003eKriss, M., K.Z. Hazleton, N.M. Nusbacher, C.G. Martin, and C.A. Lozupone, Low diversity gut microbiota dysbiosis: drivers, functional implications and recovery. Curr Opin Microbiol, 2018. 44;34-40.\u003c/li\u003e\n\u003cli\u003eRupp\u0026eacute;, \u0026Eacute;., T. Lisboa, and F. Barbier, The gut microbiota of critically ill patients: first steps in an unexplored world. Intensive Care Med, 2018. 44;9;1561-1564.\u003c/li\u003e\n\u003cli\u003eShin, N.R., T.W. Whon, and J.W. Bae, Proteobacteria: microbial signature of dysbiosis in gut microbiota. Trends Biotechnol, 2015. 33;9;496-503.\u003c/li\u003e\n\u003cli\u003eZhang, X., Y. Zhao, M. Zhang, X. Pang, J. Xu, C. Kang, et al., Structural changes of gut microbiota during berberine-mediated prevention of obesity and insulin resistance in high-fat diet-fed rats. PLoS One, 2012. 7;8;e42529.\u003c/li\u003e\n\u003cli\u003eDorman, M.J. and F.L. Short, Genome watch: Klebsiella pneumoniae: when a colonizer turns bad. Nat Rev Microbiol, 2017. 15;7;384.\u003c/li\u003e\n\u003cli\u003eMoschen, A.R., R.R. Gerner, J. Wang, V. Klepsch, T.E. Adolph, S.J. Reider, et al., Lipocalin 2 Protects from Inflammation and Tumorigenesis Associated with Gut Microbiota Alterations. Cell Host Microbe, 2016. 19;4;455-69.\u003c/li\u003e\n\u003cli\u003eR\u0026ouml;hrig, F. and A. Schulze, The multifaceted roles of fatty acid synthesis in cancer. Nat Rev Cancer, 2016. 16;11;732-749.\u003c/li\u003e\n\u003cli\u003eMa, Y., S.M. Temkin, A.M. Hawkridge, C. Guo, W. Wang, X.Y. Wang, et al., Fatty acid oxidation: An emerging facet of metabolic transformation in cancer. Cancer Lett, 2018. 435;92-100.\u003c/li\u003e\n\u003cli\u003eKumar, A., C. Smith, C. Jobin, G. Trinchieri, T.K. Howcroft, H. Seifried, et al., Workshop Report: Modulation of Antitumor Immune Responses by Dietary and Microbial Metabolites. J Natl Cancer Inst, 2017. 109;6.\u003c/li\u003e\n\u003cli\u003ePegg, A.E., The function of spermine. IUBMB Life, 2014. 66;1;8-18.\u003c/li\u003e\n\u003cli\u003eNowotarski, S.L., P.M. Woster, and R.A. Casero, Jr., Polyamines and cancer: implications for chemotherapy and chemoprevention. Expert Rev Mol Med, 2013. 15;e3.\u003c/li\u003e\n\u003cli\u003eYe, J.Z., Y.T. Li, W.R. Wu, D. Shi, D.Q. Fang, L.Y. Yang, et al., Dynamic alterations in the gut microbiota and metabolome during the development of methionine-choline-deficient diet-induced nonalcoholic steatohepatitis. World J Gastroenterol, 2018. 24;23;2468-2481.\u003c/li\u003e\n\u003cli\u003eShie, P.H., S.Y. Wang, H.L. Lay, and G.J. Huang, 4,7-Dimethoxy-5-methyl-1,3-benzodioxole from Antrodia camphorata inhibits LPS-induced inflammation via suppression of NF-\u0026kappa;B and induction HO-1 in RAW264.7 cells. Int Immunopharmacol, 2016. 31;186-94.\u003c/li\u003e\n\u003cli\u003eBeggs, S. and M.W. Salter, SnapShot: Microglia in Disease. Cell, 2016. 165;5;1294-1294.e1.\u003c/li\u003e\n\u003cli\u003eWu, S.Y. and K. Watabe, The roles of microglia/macrophages in tumor progression of brain cancer and metastatic disease. Front Biosci (Landmark Ed), 2017. 22;1805-1829.\u003c/li\u003e\n\u003cli\u003eSoulet, D. and S. Rivest, Polyamines play a critical role in the control of the innate immune response in the mouse central nervous system. J Cell Biol, 2003. 162;2;257-68.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-2259805/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-2259805/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eBrain metastasis (BM) is associated with high mortality in patients with non-small cell lung cancer (NSCLC). Alterations in the gut microbiota have been implicated in modulation of brain disorders through the gut-brain-axis (GBA). However, the underlying mechanism by which the gut microbiota affects the development of BM in NSCLC remains largely unknown. In patients, we identified 16 genera of differential bacteria positively or negatively correlated with BM in NSCLC patients, as represented by \u003cem\u003eKlebsiella\u003c/em\u003e, \u003cem\u003eunclassified_f_Enterobacteriaceae\u003c/em\u003e and \u003cem\u003eAlistipes\u003c/em\u003e by 16S rRNA gene sequencing. In addition, untargeted metabolomics (LC-MS/MS and GC/MS) identified 76 metabolites, that were associated with BM. The combination of \u003cem\u003eintestinimonas\u003c/em\u003e and spermine was considered a potential marker for the diagnosis of BM in NSCLC. Moreover, the plasma metabolite spermine enhanced BM by promoting M2 polarization of microglia via activation of the signal transducer and activator of transcription 3 (STAT3) signaling pathway \u003cem\u003ein vivo\u003c/em\u003e and \u003cem\u003ein vitro\u003c/em\u003e, suppressing innate immune function, which in turn promoted tumor progression. Overall, our study demonstrated the composition of both the gut microbiota and metabolites changed significantly between groups, and revealed that metabolite spermine promotes BM by skewing the polarity of M2 microglia by activating STAT3 signals. Our results provide a novel perspective regarding host-gut microbiota interplay in BM of NSCLC and highlight a potential risk of spermine in promoting BM.\u003c/p\u003e","manuscriptTitle":"Microbiota-modulated spermine promotes brain metastasis in non-small cell lung cancer by regulating microglia M2 polarization via the STAT3 pathway","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2022-11-30 15:47:30","doi":"10.21203/rs.3.rs-2259805/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"bc56d227-3b32-4795-803e-946a976bd220","owner":[],"postedDate":"November 30th, 2022","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2023-01-17T21:33:50+00:00","versionOfRecord":[],"versionCreatedAt":"2022-11-30 15:47:30","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-2259805","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-2259805","identity":"rs-2259805","version":["v1"]},"buildId":"omnImTCwR2MFx8CMYfrG7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}