Altered colonic microflora and its metabolic profile in mice with acute viral myocarditis induced by coxsackievirus B3 Authors: Yimin Xue, Shirong Lin, Mingguang Chen, Jun Ke, Jiuyun Zhang, Qiaolian Fan, Yimei Chen, Feng Chen Published: 2024 ABSTRACT AbstractMounting evidence suggests that the gut-heart axis is critical in the pathogenesis of cardiovascular diseases. The gut serves as the primary pathway through which Coxsackievirus B3 (CVB3) infects its host, leading to acute viral myocarditis (AVMC). However, little is known about the role of gut microflora and its metabolites in the development of AVMC. The AVMC model was established by intraperitoneal injection of CVB3 in mice. Then, 16S ribosomal RNA (16S rRNA) gene sequencing and ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) untargeted metabolomics profiling were performed to analyze the microflora composition and metabolic profile of colonic contents. Compared to the Control mice, the AVMC mice displayed a significant reduction in gut microflora richness and diversity, as revealed by an increased abundance of Proteobacteria and a decreased abundance of Cyanobacteria and Desulfobacterota. LEfSe analysis indicated that the main genera differing between the two groups were Escherichia-Shigella, Lactobacillus, Clostridium_sensu_stricto_1, Prevotellaceae_UCG-001, and Odoribacter. Based on the criterion of OPLS-DA VIP ≥ 1.0 and p-value < 0.05, a total of 198 differential metabolites (DMs) were identified in the gut, including 79 upregulated and 119 downregulated metabolites, of which lipids and lipid-like molecules accounted for the largest proportion. Notably, both altered gut bacterial taxa and metabolites were significantly enriched in the Lipid metabolism pathway, with Traumatic acid (TA), Alpha-Linolenic acid (ALA), Eicosapentaenoic acid (EPA), and Docosahexaenoic acid (DHA) being the key DMs in the pathway. Additionally, significant positive correlations (|r| > 0.80 and p < 0.05) were found between TA levels and Anaerotruncus and Bilophila abundance, between EPA levels and Clostridium_sensu_stricto_1 abundance, and between DHA levels and Escherichia-Shigella abundance, respectively. CVB3 infection leads to notable alterations in gut microflora composition and its metabolic profile, which may participate in AVMC development. Our findings provide important clues for future in-depth studies on AVMC etiology.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12985-024-02571-z.Keywords: Acute viral myocarditis, Coxsackievirus B3, Gut microflora, Metabolomics, Gut-heart axis Introduction Fulminant myocarditis (FM) is a serious diffuse inflammatory disease of the myocardium that often leads to refractory cardiogenic shock, which is a major cause of sudden death in young adults [1]. Cardiotropic infections followed by overactivation of the immune system are the most common reasons for FM in humans [2]. The mortality rate of FM has decreased due to the development of treatment methods, such as mechanical circulatory support and immunomodulation therapy, but remains unacceptably high [2, 3]. If not treated adequately, FM may progress into dilated cardiomyopathy, which eventually leads to terminal heart failure and requires heart transplantation [4]. Therefore, it is imperative to deepen our understanding of the dissemination and pathogenesis of cardiotropic infections to discover novel therapeutic targets. Coxsackievirus B3 (CVB3) is an enterovirus of the Picornaviridae family and is one of the most common infectious agents in FM. Upon infecting the host, CVB3 replicates in large numbers in the gut and spreads into the bloodstream before invading its target cells, cardiomyocytes, via the coxsackievirus-adenovirus receptor (CAR) [5]. As the entry site of CVB3, the gut is critical for its replication and spread. Studying changes in gut function may help in the development of early interventions to block the progression of acute viral myocarditis (AVMC) to FM. The gut serves as the largest immune organ in the human body while also performing digestive functions. The gut microflora is a complex ecological community that plays indispensable roles in regulating various host functions, such as the intestinal barrier, nutrition, and immune homeostasis [6–8]. In normal circumstances, there is a symbiotic relationship between gut microflora and its host. The human host provides a living environment for the gut microflora, which in turn influences distal organ function and the immune system by producing numerous bioactive metabolites [9]. In recent years, an increasing number of studies have focused on the impact of gut microflora-derived metabolites on cardiovascular diseases. For example, trimethylamine-N-oxide (TMAO), a small organic molecule derived from gut microflora, is strongly associated with an increased risk of atherosclerosis, myocardial infarction, myocardial fibrosis, and heart failure [10–13]. Short-chain fatty acids (SCFAs) are the main metabolites produced by the fermentation of dietary fibers, including acetate, propionate, butyrate, and other fatty acids. Recent studies have shown that these metabolites exert essential roles in controlling hypertension, alleviating ischemia/reperfusion injury, reducing atherosclerosis risk, and improving cardiac dysfunction by modulating the immune system and inflammatory response [14–16]. Furthermore, secondary bile acids (BAs), metabolized by gut microflora, can dose-dependently reduce cardiomyocyte contraction via BA receptors on the surface of cardiomyocytes, in addition to their involvement in lipid absorption [17, 18]. Therefore, some scholars suggest the gut-heart axis, which involves bidirectional communication between the gut microflora and the heart [19]. With the rapid development of multi-omics combined analysis technology, the gut microflora and its metabolites have been considered potential targets for the treatment of various cardiovascular diseases [9]. Notably, a recent study by Hu et al. [20] showed increased bacterial richness and diversity in fecal samples of experimental autoimmune myocarditis (EAM) mice, and fecal microflora transplantation (FMT) helped mitigate myocardial damage by restoring microflora composition. Kong et al. [21] also investigated a potential link between gut microflora and myocardial metabolites in AVMC. These findings provide a new therapeutic direction for AVMC. However, to date, the interaction between gut microflora and its related metabolites in AVMC has been insufficiently addressed. In this study, we applied 16S ribosomal RNA (16S rRNA) gene sequencing technology to identify changes in the gut microflora composition of AVMC mice. Meanwhile, an untargeted ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) metabolomics approach was used to explore the metabolome of colonic contents. Finally, the relationship between gut microbiome and its metabolome was analyzed by integrating multi-omics data. The findings of our study may contribute to a better understanding and further research on AVMC in the near future. Materials and methods Mouse models of CVB3-induced AVMC The animal grouping and experimental design are shown in Fig. 1. A total of 20 specific pathogen-free male BALB/c mice (aged 6 weeks and weighing 17–21 g) were purchased from Beijing SiPeiFu Biotechnology Co., Ltd. [License No. SCXK (Jing) 2019-0010]. All animals were housed at Shengli Clinical Medical College of Fujian Medical University and maintained in accordance with standard ethical animal handling guidelines. After 3 days of adaptation, the mice were randomly divided into two groups: Control group (n = 7) and AVMC group (n = 13). The CVB3-induced AVMC mouse model was established according to our previous report [22]. Briefly, the AVMC mice were treated with CVB3 (Nancy strain; 2 × 105 plaque-forming units/mouse) via intraperitoneal administration, and the Control mice were treated with phosphate-buffered saline (PBS). After 7 days, mice that survived were anesthetized using 2% isoflurane and euthanized by cervical dislocation. Their hearts and colonic contents were harvested for histological examination, quantitative RT-PCR (RT-qPCR), 16S rRNA gene sequencing, and untargeted metabolomics profiling, respectively. All animal experiments followed the Principles of Laboratory Animal Care and were approved by the Institutional Animal Care and Use Committee of Fujian Provincial Hospital (Permit No. IACUC-FPH-SL-20231020[0219]). Schematic diagram illustrating the overall experimental workflow Histological examination The hearts were fixed overnight with 4% paraformaldehyde, embedded in paraffin, and cross-sectioned into 5-µm slices. For histological examination, the sections were processed using the conventional hematoxylin and eosin (HE) staining method for visualization under microscopic inspection (Leica DM2000 LED, Wetzlar, Hessian, Germany). Cardiac pathological scores were determined blindly based on the criteria described previously [23]. RT-qPCR Total RNA was isolated from cardiac tissue homogenates (5 samples per group) using TRIzol® reagent (#15596-026, Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Reverse transcription for cDNA synthesis was conducted with a PrimeScript RT Master Mix (#RR036A, TaKaRa, Kusatsu, Shiga, Japan). The cDNA was then subjected to RT-qPCR using PowerUp™ SYBR™ Green Master Mix (#A25742, Thermo Fisher Scientific, Wilmington, DE, USA) on an ABI7500 instrument (Applied Biosystems, Foster City, CA, USA) for 40 cycles. The reaction condition was as follows: 95˚C for 10 min, followed by 40 cycles consisting 95˚C for 15 s, and 60˚C for 1 min. Relative expression levels were calculated using the 2−ΔΔCt method and were normalized to β-actin expression. The primer sequences used were as follows: CVB3, forward primer: CGGTACCTTTGTGCGCCTGT, reverse primer: CAGGCCGCCAACGCAGCC; β-actin, forward primer: GCACCACACCTTCTACAATG, reverse primer: GTGAGGGAGAGCATAGCC. DNA extraction and 16S rRNA gene sequencing Colonic contents were collected and immediately stored at − 80 °C prior to analysis. Microbial DNA was extracted using the HiPure Stool DNA Kit (#D3141, Magen, Guangzhou, Guangdong, China) following the manufacturer’s guidelines. Each purified DNA sample was subjected to quality control using 2% agarose gel electrophoresis (Biowest, Barcelona, Spain) and a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). Subsequently, the 16S rRNA V3-V4 region was amplified by PCR using the 341 F (CCTACGGGNGGCWGCAG) and 806R (GGACTACHVGGGTATCTAAT) primers in a Thermocycler (Applied Biosystem). The PCR conditions were as follows: 95 °C for 5 min, followed by 30 cycles at 95 °C for 1 min, 60 °C for 1 min, 72 °C for 1 min, and a final extension at 72 °C for 7 min. The amplification products were purified using AMPure XP Beads (#A63881, Beckman Coulter, Brea, CA, USA) and quantified using an ABI StepOnePlus Real-Time PCR System (Life Technologies, Foster City, CA, USA) following the protocols. After the library construction, paired-end sequencing (PE250) was performed on an Illumina NovaSeq 6000 platform (Illumina, San Diego, CA, USA). Processing of sequencing data The raw 16S rRNA gene sequencing reads were trimmed using FASTP software (v.0.20.0) to remove low-quality and adapter-contaminated reads [24]. Using FLASH software (v.1.2.7), the PE reads for each sample were then spliced into tags by overlapping, and the resulting clean tags were obtained after filtering [25]. Next, the clustering was performed using USEARCH software (v.8.0.1623) to obtain operational taxonomic unit (OTU, 97% similarity) abundance and OTU representative sequences by removing chimeras [26]. Finally, OTUs were annotated with taxonomic information using a naïve Bayesian classifier at the Ribosomal Database Project (v.2.2) [27]. The OTU-based assessments of the Alpha diversity indices, including community richness (ACE and Chao1) and diversity (Shannon and Simpson), were calculated in QIIME (v.1.9.1) and compared between groups using Welch’s t-test [28]. Beta diversity analysis was performed using the Bray-Curtis and Weighted UniFrac distances to assess structural changes in the bacterial community and visualized by principal coordinate analysis (PCoA) and non-metric dimensional scaling (NMDS) plots. Stacked bar plots generated in the R project ggplot2 package (v.2.2.1) were used to identify the community composition at different classification levels (phylum and genus). Additionally, significant differences in the relative abundance of bacterial taxa (from the phylum down to the genus level) between the Control and AVMC groups were determined based on a linear discriminant analysis (LDA) effect size (LEfSe) method [29]. For the functional prediction of the bacterial community, the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of OTUs was inferred using PICRUSt2 (https://github.com/picrust/picrust2) [30]. Metabolite extraction and UHPLC-MS/MS analysis of colonic contents Metabolites from mouse colonic contents were extracted using a previously described protocol [31]. Specifically, 25 mg of colonic contents were placed in an EP tube, and 500 µL of extraction solution (MeOH: ACN: H2O = 2:2:1, v/v) containing deuterated internal standards was added. Subsequently, the samples were vortexed for 30 s, homogenized at 35 Hz for 4 min, and sonicated for 5 min in an ice-water bath (the latter two steps were repeated 3 times). After incubation for 1 h at − 40℃ to precipitate proteins, the samples were centrifuged (12000 rpm, 4℃) for 15 min. Finally, the resulting supernatants (200 µL) were collected for UHPLC-MS/MS analysis using a Vanquish UHPLC system (Thermo Fisher Scientific) coupled with an Orbitrap Q Exactive™ HF-X mass spectrometer (Thermo Fisher Scientific). 20 µL supernatant of each sample was mixed to generate a pooled quality control (QC) sample, and the extraction solution was designated as the blank sample. Data processing and untargeted metabolomics analysis The raw data files were imported into the Compound Discoverer 3.1 software (Thermo Fisher Scientific) for data pre-processing, including data filtering, peak alignment, peak extraction, and peak area quantification for metabolites [32]. Then, peaks were matched with the mzCloud, mzVault, and MassList databases to obtain metabolite identification and relative quantitative results. Principal component analysis (PCA) and orthogonal projection to latent structures-discriminant analysis (OPLS-DA) were performed using R package models (http://www.r-project.org/) to visualize the metabolic alterations between the two groups [33]. The OPLS-DA model was further validated by a seven-fold cross-validation method with a 200 times permutation test. A variable importance in the projection (VIP) score for each metabolite was calculated for ranking based on the established OPLS-DA model. Meanwhile, the univariate analysis t-test was also applied for screening differential metabolites. Those with a p-value of t-test < 0.05 and VIP ≥ 1.0 were considered differential metabolites (DMs) between the two groups. For visualization, the volcano and heatmap of DMs were plotted using the ggplots2 (v.3.3.1) and pheatmap (v1.0.12) packages in R software. Finally, the function and metabolic pathways of these DMs were annotated using the KEGG database and Metabolomics Pathway Analysis (MetPA, http://metpa.metabolomics.ca/) [34]. Statistical analysis Data were analyzed using GraphPad Prism 8.0.1 (GraphPad Software Inc., La Jolla, USA). The Mann-Whitney U test was used to evaluate differences in cardiac pathology scores between the two groups. Comparison of cardiac CVB3 RNA levels between the two groups was performed using an unpaired two-tailed t-test. The microflora-metabolite correlation significance was assessed by plotting the Pearson correlation heatmap and network using R language and Cytoscape (v.3.9.1) with the threshold values of |r| > 0.80 and p < 0.05. A p < 0.05 was considered statistically significant. Results Validation of the AVMC mouse model Mice in the AVMC group began to exhibit typical signs of illness from day 2 post-CVB3 infection, including lethargy, irritability, back arching, anorexia, and weight loss. By day 7, 4 of 13 mice (30.8%) in the AVMC group had died compared to 0 of 7 mice (0%) in the Control group. Under light microscopy, massive inflammatory infiltration and cardiomyocyte necrosis were observed in the hearts of AVMC mice, while the hearts of Control mice showed no significant pathological changes (HE staining). This suggests that the cardiac pathology scores of the AVMC mice were significantly elevated compared to those of the Control mice (Fig. 2A). In addition, the cardiac CVB3 RNA levels in AVMC mice were significantly higher than those in Control mice (Fig. 2B), confirming the successful establishment of an AVMC model in CVB3-infected mice. Establishment of an AVMC mouse model and analysis of gut microflora structure. (A) Representative HE staining images of hearts from the Control (n = 7) and AVMC (n = 9) groups (scale bar = 200 μm, 50 μm) and comparison of cardiac pathological scores between the two groups. (B) Comparison of cardiac CVB3 RNA levels between the two groups (n = 5/group). (C-D) Alpha diversity indexes of the gut microflora at the OTU level, including community richness (C) and diversity (D). An unpaired two-tailed Welch’s t-test was used for statistical analysis between the two groups. E-F Beta diversity analysis of PCoA (E) and NMDS (F), based on the Bray-Curtis and Weighted UniFrac distances between all samples. Green dots indicate the Control group, and red dots indicate the AVMC group, where each dot represents an individual sample. **p < 0.01 CVB3 infection reduced the richness and diversity of gut microflora Next, genomic DNA was collected from colonic content samples of the Control (C-1 ∼ 7) and AVMC (A-1 ∼ 9) groups and analyzed using 16S rRNA gene sequencing. Following taxonomic assignment, 2374 OTUs were obtained throughout the samples (Supplementary Table S1). A Venn diagram showed that the two groups shared 393 OTUs, whereas 293 and 143 OTUs were unique to the Control and AVMC mice, respectively (Fig. S1). Next, bacterial community richness and diversity were measured at the OTU level using the Alpha diversity metrics. As shown in Fig. 2C-D, significantly lower bacterial richness (as measured by the ACE and Chao1 indexes) and diversity (as measured by the Shannon and Simpson indexes) were observed in the CVB3-infected mice compared to the Control mice. To display the differences in overall bacterial community composition between samples, Beta diversity analysis was performed through PCoA and NMDS plots based on the Bray-Curtis and Weighted UniFrac indexes at the OTU level. The results showed that the samples in the AVMC group significantly differed from those in the Control group, implying a distinct bacterial community composition between the two groups (Fig. 2E-F). Alterations in the composition and function of the gut microflora in AVMC The relative proportions of preponderant taxa were determined by microbial taxonomic assignment. The top 10 most abundant bacterial taxa for the two groups of samples at the phylum and genus levels are shown in Fig. 3A-B. At the phylum level, Bacteroidota, Firmicutes, Proteobacteria, and Patescibacteria dominated the gut bacterial community in both groups. Compared to the Control group, the AVMC mice exhibited significantly higher levels of Proteobacteria, but significantly lower levels of Cyanobacteria and Desulfobacterota (Supplementary Table S2 and Fig. S2A). At the genus level, the top 3 predominant bacterial genera in the AVMC group were Escherichia-Shigella, Lactobacillus, and Alistipes, while those in the Control group were Alistipes, Prevotellaceae_UCG-001, and Lachnospiraceae_NK4A136_group. The proportions of Prevotellaceae_UCG-001, Odoribacter, Muribaculum, Ruminococcus, Colidextribacter, and other genera were significantly lower in the AVMC group than in the Control group, and only the proportions of Escherichia-Shigella, Lactobacillus, and Clostridium_sensu_stricto_1 were significantly higher in the AVMC group (Supplementary Table S3 and Fig. S2B). Alterations in the composition and function of the gut microflora in AVMC. (A-B) Composition and relative abundance of bacterial phyla (A) and genera (B) in the Control (C-1 ∼ 7) and AVMC (A-1 ∼ 9) groups. The top 10 most abundant taxa across all samples are shown, with other known taxa categorized as “Other” and unknown taxa as “Unclassified”. (C) Cladogram representation of different gut microflora taxa using LEfSe analysis, from the phylum level down to the genus level. Green indicates enriched taxa in the Control group, and red indicates enriched taxa in the AVMC group. Each circle’s diameter is proportional to the taxon’s relative abundance. (D) Bar graph showing LDA scores. Taxa enriched in the Control group are indicated with a positive LDA score (green), and taxa enriched in the AVMC gourp have a negative score (red). Only taxa with an LDA score > 4.0 are shown. (E) Differences in bacterial function at KEGG level 2 using PICRUSt2 To further identify bacterial biomarkers between groups, LEfSe analysis was performed from the phylum to the genus level. Cladograms of the taxa with LDA score > 4.0 were depicted in Fig. 3C. A total of 20 bacterial taxa with rich differences were identified. Of these, the Control group had 8 (1 class, 2 orders, 3 families, and 2 genera) and the AVMC group had 12 (1 phylum, 2 classes, 3 orders, 3 families, and 3 genera) (Fig. 3D and Supplementary Table S4). At the genus level, the biomarkers for the AVMC group were Escherichia-Shigella, Lactobacillus, and Clostridium_sensu_stricto_1, and for the Control group were Prevotellaceae_UCG-001 and Odoribacter. Finally, the functional differences in gut microflora between the two groups were predicted using PICRUSt2. At KEGG level 2 (Fig. 3E), 5 differential KEGG pathways (including Lipid metabolism, Xenobiotics biodegradation and metabolism, Membrane transport, Signal transduction, and Infectious diseases) were identified in the gut microbiome of AVMC mice (Supplementary Table S5). At KEGG level 3 (Fig. S3), we found a total of 44 statistically different functional KEGG pathways between the two groups (Supplementary Table S6). Most of the pathways were enriched in the AVMC group, mainly involving metabolism, biosynthesis, and degradation. By contrast, only 3 pathways (Non-homologous end-joining, Toxoplasmosis, and mRNA surveillance pathway) were significantly enriched in the Control group. Taken together, these results suggest that CVB3 infection can significantly alter the composition and function of the gut microflora. Altered metabolic profile of colonic contents in AVMC mice To determine the effects of CVB3 infection on the gut metabolome, we analyzed colonic contents using UHPLC-MS/MS-based untargeted metabolomics. After data preprocessing and filtering, a total of 2924 metabolites were obtained, of which 2138 and 786 metabolites were annotated according to MS1 and MS2, respectively. The 3D PCA plot showed that the metabolites in the Control and AVMC groups were well separated, with 57.9%, 22.9%, and 6.3% variation attributed to the principal components PC1, PC2, and PC3, respectively (Fig. 4A). The OPLS-DA score showed that the Control and AVMC groups were scattered in different areas. The goodness of fit values (R2) and predictive ability values (Q2) indicated that the model possessed a satisfactory fit with good predictive power (Fig. 4B). In the permutation test, both R2’ and Q2’ were smaller than R2 and Q2 of the OPLS-DA model, which demonstrated the model was robust and not overfitted (Fig. 4C). Altered metabolic profile of colonic contents in AVMC mice. (A) 3D PCA scatter plot for the colonic metabolite profiling. (B) OPLS-DA analysis showing good discrimination between the Control and AVMC groups (R2X = 0.478, R2Y = 0.969, Q2Y = 0.911). (C) The permutation test (n = 200) for the OPLS-DA model. (D) Volcano plot showing changes in DMs between the two groups. DMs were screened based on the criterion of VIP ≥ 1.0 and p-value < 0.05. Red dots indicate upregulated metabolites, and blue dots indicate downregulated metabolites. (E) Hierarchical clustering heatmap analysis of DMs at the MS2 level between the two groups The candidate metabolites with VIP ≥ 1.0 and p-value < 0.05 were considered as potential biomarkers. In total, 198 DMs were identified between the Control and AVMC groups, including 79 upregulated and 119 downregulated metabolites (Supplementary Table S7 and Fig. 4D). Specifically, a nonredundant number of 5 DMs in positive ion mode and 102 DMs in negative ion mode were assigned at the MS1 level. At the MS2 level, 10 positive ion mode DMs and 81 negative ion mode DMs were assigned, which were then selected to construct a heatmap for better visualization (Fig. 4E). Among them, 26 DMs belonged to the superclass of lipids and lipid-like molecules, accounting for the largest proportion. Additionally, Pearson’s correlation analysis was performed to examine the correlation between the identified DMs. The top 50 DMs with the most significant VIP values were selected for visual analysis (Fig. S4). A majority of the DMs exhibited high correlations with other DMs. Pathways associated with the altered colonic metabolites To identify the pathways associated with these DMs, KEGG pathway enrichment analysis was performed. Several key pathways were significantly enriched (p < 0.05), such as Degradation of aromatic compounds, GABAergic synapse, ABC transporters, Taurine and hypotaurine metabolism, 2-Oxocarboxylic acid metabolism, Alpha-Linolenic acid metabolism, and Biosynthesis of unsaturated fatty acids (Supplementary Table S8 and Fig. 5A). Of these, Alpha-Linolenic acid metabolism and Biosynthesis of unsaturated fatty acids involving Traumatic acid (TA), Alpha-Linolenic acid (ALA), Eicosapentaenoic acid (EPA), and Docosahexaenoic acid (DHA), belonged to the Lipid metabolism category according to the KEGG_B_class annotation. Meanwhile, the MetPA module of MetaboAnalyst was used to screen the relevant metabolic pathway in which the DMs were involved. A total of 36 pathways were found to be associated with DMs, of which 2 pathways were significant (p < 0.05 and impact > 0.10), including Arginine biosynthesis and D-Glutamine and D-glutamate metabolism (Supplementary Table S9 and Fig. 5B). These results suggest that CVB3 infection may alter the metabolic function of the gut microflora in mice, which deserves further investigation. Pathways associated with the altered colonic metabolites. (A) KEGG pathway enrichment analysis of DMs. The significantly enriched pathways are displayed by a bubble plot (p < 0.05), with the dot size indicating the number of enriched DMs, the dot color indicating the -log10(p-value), and the dot position indicating the rich factor. (B) Summary of pathway analysis using MetPA (top 25). The color and size of each circle are based on -log10(p-value) and pathway impact value, respectively. Statistically significant pathways are annotated with their names in blue font (p < 0.05 and impact > 0.10) Association between the gut microbiome and its metabolome in AVMC To explore the relationship between the altered gut bacterial taxa and metabolites in AVMC, we performed a Pearson’s correlation analysis between 16 differential bacterial genera and 198 DMs (Supplementary Table S10). At the MS2 level, we identified 65 statistically significant interactions (|r| > 0.80 and p < 0.05) between them (Fig. 6A-B). Among them, Prevotellaceae_UCG-001 displayed the most connections with the DMs. Based on the KEGG database, Lipid metabolism was a highly enriched pathway. Pearson’s correlation analysis revealed that the TA levels positively correlated with the Anaerotruncus and Bilophila abundance. The EPA levels positively correlated with the Clostridium_sensu_stricto_1 abundance. The DHA levels positively correlated with the Escherichia-Shigella abundance. This suggests that the observed alterations in colonic metabolites upon CVB3 infection are likely derived from the gut microflora, which may be one of the underlying pathogenic mechanisms of AVMC. Association between the gut microbiome and its metabolome in AVMC. (A) Pearson correlation heatmap of differential gut bacterial genera and DMs at the MS2 level. The horizontal axis is the differential bacterial genera, and the vertical axis is the DMs at the MS2 level. Red means positive correlation, and blue means negative correlation (* represents |r| > 0.80 and p < 0.05). (B) Network diagram of correlation analysis. Circular nodes represent correlated differential bacterial genera, and diamond-shaped nodes represent correlated DMs. Node size represents the relative abundance/expression, with red indicating upregulation and blue indicating downregulation. Lines represent the interactions between the two nodes (red for positive and blue for negative correlation) Discussion The gut is the first biological barrier against enterovirus CVB3 infection, and its microenvironment may influence the disease outcome. In this study, we identified the gut microbiome and its metabolome in mice with CVB3-induced AVMC, a widely used model for exploring the pathogenic mechanisms of human AVMC. The results showed that the gut microflora composition and metabolic profile were significantly altered in AVMC mice compared to the Controls, and these alterations were correlated. They may be involved in the pathogenesis of AVMC through multiple pathways, most notably lipid metabolism, suggesting an essential link between gut microecology and AVMC. The present study helps bridge the gap in existing research to some extent and lays a foundation for future investigations in this field. The gut microflora is a diverse and dynamic community of microorganisms colonizing the human body. Among them, Bacteroidota and Firmicutes are the dominant phyla, accounting for more than 90% of the total microbiome [35]. The richness and diversity of gut microflora are key factors in maintaining gut barrier function and integrity, and their loss has been associated with various cardiovascular diseases, such as myocardial infarction, heart failure, and hypertension [36–38]. Similarly, based on the Alpha and Beta diversity analyses, we determined significant differences in gut microflora composition between the two groups, and that CVB3 infection significantly reduced gut microflora richness and diversity in mice. The abundance of Proteobacteria, a microbial indicator of gut microflora imbalance, was also found to be significantly elevated in AVMC mice [39]. However, we did not observe a significantly decreased abundance of Bacteroidetes, which contrasts with the findings of the study by Kong et al. [21]. This phenomenon may be related to the different time points of the assay. Furthermore, unlike a previous study that found significantly elevated Firmicutes/Bacteroidota ratios in EAM [20], our study did not find a significant difference in Firmicutes/Bacteroidota ratios between the two groups. Thus, AVMC and EAM may exhibit distinct bacterial community distributions, as the mechanisms underlying alpha-cardiac myosin heavy chain-induced EAM differ from those of AVMC. At the genus level, the dominant genera Prevotellaceae_UCG-001 and Odoribacter (belonging to the phylum Bacteroidota) decreased significantly after CVB3 infection. Downregulation of the abundance of these two genera has been shown to reduce anti-inflammatory metabolite production, disrupt homeostasis in the gut, and promote host inflammation [40]. Here, we noted a significant positive correlation between the abundance of Prevotellaceae_UCG-001 and the levels of organic acids and derivatives, especially Ethyl acetoacetate, Alpha-ketoisovaleric acid, N-Acetylserine, and Acetylglycine. These organic acids and derivatives can improve gut microflora dysbiosis by regulating intestinal pH and promoting the growth of probiotics [41]. In the future, it would be an interesting direction to further explore the interaction patterns of Prevotellaceae_UCG-001 with these organic acids and derivatives in AVMC. Additionally, the gut bacterial community in AVMC mice was dominated by Escherichia-Shigella, Lactobacillus, and Clostridium_sensu_stricto_1. Escherichia-Shigella is generally considered a non-pathogenic bacterium, but when stimulated by stress, it can become pathogenic and participate in disease progression [42]. Several studies have shown that the abundance of Escherichia-Shigella is positively correlated with levels of proinflammatory molecules, such as IL-6, IL-1β, and NLRP3 [43, 44]. Lactobacillus is a type of probiotic with anti-inflammatory properties [45]. It is important to note that Lactobacillus can regulate cellular immune responses mediated by CD4 + T cells and CD8 + T cells, which are essential for controlling viral infections and preventing their spread [45, 46]. This suggests that Lactobacillus may be beneficial for the treatment of AVMC. Clostridium_sensu_stricto_1 is another opportunistic pathogen that affects gut inflammation and barrier function [47]. The relative abundance of Clostridium_sensu_stricto_1 has been shown to be negatively correlated with the expression of intestinal tight junction proteins [48]. Overall, these findings support the idea that CVB3 infection leads to significant gut microflora dysbiosis. The gut microflora is a virtual endocrine system that influences the physiology and function of distal organs through its metabolites. In this study, alterations in the gut microflora of AVMC mice were accompanied by considerable changes in the gut metabolome. Based on the results of untargeted metabolomics analysis, a total of 198 DMs were identified, mainly involving Lipids and lipid-like molecules, Organic acids and derivatives, Organoheterocyclic compounds, Benzenoids, Phenylpropanoids and polyketides, and Organic oxygen compounds. Although we identified the roles of several DMs (e.g., Succinic acid, Phthalic acid, L-Glutamic acid, Oxoglutaric acid, and Taurine, etc.) in metabolism, synthesis, and degradation through pathway enrichment analysis, there are still many DMs whose functions remain unclear. For example, Petroselinic acid (PSA), an 18-carbon monounsaturated fatty acid with the highest VIP value, was highly expressed in the gut of AVMC mice. Previous studies have shown that PSA exerts anti-inflammatory activity by impeding NO production, biofilm formation, and virulence gene expression [49, 50]. In a recent report, Huang et al. [51] found that PSA could alleviate intestinal ischemia/reperfusion injury by activating the AMPK-mTOR-P70S6K pathway. 7-Ketodeoxycholic acid (7-KDCA), the only secondary bile acid identified in the DMs, is the major intermediate byproduct of the gut bacterial conversion of chenodeoxycholic acid into ursodeoxycholic acid [52]. Elevated levels of 7-KDCA are thought to help alleviate inflammatory damage in colitis [53, 54]. The potential role of these two DMs in maintaining gut homeostasis of AVMC deserves further exploration. Furthermore, we performed PICRUSt2 to predict the potential functions of bacterial community based on 16S rRNA gene sequencing data. Interestingly, both altered gut bacterial taxa and metabolites were significantly enriched in the Lipid metabolism pathway, with TA, ALA, EPA, and DHA identified as key DMs. ω-3 polyunsaturated fatty acids (ω-3 PUFAs), including ALA, EPA, and DHA, are essential fatty acids to mammals with known immunomodulatory properties [55]. They alleviate the inflammatory process by regulating dendritic cells, neutrophils, macrophages, and T/B cells, which may contribute to early viral clearance of AVMC [56]. Thus, high levels of ω-3 PUFAs may be a response to the inflammatory environment of AVMC in vivo. The beneficial action of ω-3 PUFAs supplementation can also occur through gut microflora. Several studies have demonstrated that ω-3 PUFAs dietary supplementation increases the abundance of the beneficial bacterial genera in the gut [57, 58]. In this study, we found certain correlations between the gut microflora and ω-3 PUFAs, which will help to explore the impact of ω-3 PUFAs supplementation on gut microflora dysbiosis in AVMC. TA is a beneficial compound secreted by injured plants with antioxidant, anti-inflammatory, and tissue repair potential [59]. Zhao et al. [60] found that reduced levels of microbial-derived TA promote the development of psoriasis, suggesting that low levels of TA may be detrimental to AVMC. Importantly, TA displayed strong positive correlations with the relative abundance of Bilophila and Anaerotruncus, indicating that the supplementation of these microbes in the gut may help to increase TA and benefit AVMC. Our study highlights that targeting these microbes or their metabolites can be a promising new therapy for AVMC. However, some limitations of this study should be discussed. First, this study was conducted during the acute phase of CVB3 infection (day 7). It would be interesting to include additional time points to observe the dynamic changes in the gut microbiome and its metabolome, such as by retesting on day 14, day 28, and day 56 post-infection to evaluate their potential roles in cardiac repair following myocardial injury. Second, there are several limitations associated with the functional prediction of PICRUSt2, including the bias caused by the reference database and the inability to distinguish strain-specific functions. Hence, these predictions must be considered exploratory. Furthermore, we only performed untargeted metabolomics to identify DMs in the gut. Further targeted metabolomics analysis is necessary to validate our preliminary results and increase confidence in our findings. Finally, FMT helps to re-establish a normal gut microflora composition and is increasingly used to treat various cardiovascular diseases, such as doxorubicin-induced cardiotoxicity [61], atrial fibrillation [62], and atherosclerosis [63]. Targeting the gut-heart axis using FMT to alleviate AVMC has broad application prospects in future development. Conclusions CVB3 infection disrupts the gut microflora and its metabolic homeostasis, which may serve as a new entry point for mechanistic studies of AVMC and a target for early treatment. Our findings shed new light on the understanding of disease pathogenesis and the development of novel therapeutic strategies for AVMC. Electronic Supplementary Material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Acknowledgements We are grateful for the sequencing platform and bioinformation analysis of Gene Denovo Biotechnology Co., Ltd (Guangzhou, China). The schematic diagram in this article was created by Figdraw (www.figdraw.com). Author contributions Yimin Xue and Feng Chen conceived and designed the study. Yimin Xue, Shirong Lin, Mingguang Chen, Jun Ke, Jiuyun Zhang, Qiaolian Fan, and Yimei Chen conducted experimental studies and collected data. Yimin Xue, Shirong Lin, and Mingguang Chen performed data analysis, drafted and revised the manuscript. Feng Chen coordinated the study and reviewed the manuscript. All authors read and approved the final manuscript. Funding This work was supported by Joint Funds for the Innovation of Science and Technology, Fujian Province (No. 2023Y9287), Fujian Provincial Health Technology Project (No. 2023GGA009, 2022ZD01008, and 2019-2-1), and Fujian Provincial Natural Science Foundation (No. 2022J01402). Data availability The raw datasets are available from the corresponding author upon reasonable request. Declarations Ethics statement All animal protocols were evaluated and approved by the Institutional Animal Care and Use Committee of Fujian Provincial Hospital (Permit No. IACUC-FPH-SL-20231020[0219]). Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Yimin Xue, Shirong Lin, and Mingguang Chen contributed equally to this work. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement The raw datasets are available from the corresponding author upon reasonable request.