Multiomics Analysis Identifies Candidate Molecular,Microbiota,and Metabolite Features Associated with Acute Myocardial Infarction Combined with Chronic Renal Failure

Abstract

Chronic renal failure (CRF) is a major risk factor for acute myocardial infarction (AMI) progression and poor outcomes, but the molecular, microbial, and metabolic features of AMI combined with CRF remain unclear. In this study, four Gene Expression Omnibus datasets were analyzed separately. GSE194388 and GSE180393 were used as discovery cohorts for AMI and CRF, and GSE109048 and GSE142153 were used as validation cohorts. Fecal samples from CRF patients without AMI (negative control [NC], n = 10) and patients with AMI and advanced CKD (SY, n = 10) were analyzed by 16S rRNA sequencing. Matched serum samples were analyzed by untargeted metabolomics. Four shared candidate hub genes, PTPRC, ITGAL, CD44, and SELL, were identified and showed preliminary discriminatory performance in validation datasets, although potential overfitting should be considered. These genes were positively correlated with increased immune cell subsets. Compared with NC, SY showed exploratory taxa-level microbiota differences, including increased Enterobacterales and decreased Bifidobacterium, without significant global diversity differences. Differential serum metabolites, including increased l-kynurenine and decreased glucose, were also identified and correlated with altered taxa. This integrated multiomics analysis identified candidate immune-related genes, microbiota features, and serum metabolites associated with AMI combined with CRF, providing preliminary evidence for future validation.

Keywords

multiomics analysis 16S rRNA sequencing metabolomics transcriptomics acute myocardial infarction renal failure

Introduction

The heart and kidneys interact closely, and dysfunction in one organ can aggravate injury in the other. Chronic renal failure (CRF) is an important risk factor for cardiovascular disease progression. The cardiovascular mortality of hemodialysis patients with renal failure is reported to be 10–30 times higher than that of the general population (Geng et al., 2024). In acute coronary syndrome, patients with CRF have a higher risk of recurrent acute myocardial infarction (AMI), heart failure, and atypical angina symptoms than patients without CRF (Moukarbel et al., 2014). CRF-related cardiac injury is partly associated with the accumulation of toxic metabolites, such as creatinine (Brankovic et al., 2020; van Veldhuisen et al., 2016). However, other metabolites and molecular changes may also contribute to myocardial injury in patients with CRF. Therefore, identifying molecular and metabolic features in AMI patients with CRF may help clarify the pathological basis of cardiorenal syndrome.

The gut microbiota also participates in cardiorenal interactions. Gut microorganisms regulate immunity, barrier integrity, and glucose and lipid metabolism (Adak and Khan, 2019). During CRF, gut microbiota dysbiosis, intestinal barrier impairment, and bacterial translocation may occur (Wang et al., 2020). Microbiota-derived uremic toxins, including indoxyl sulfate and p-cresyl sulfate, can activate inflammatory responses and promote renal fibrosis (Headley et al., 2025). Gut microbiota has also been linked to cardiovascular disease and AMI (Chen et al., 2025). For example, reduced probiotics may increase trimethylamine metabolism and the production of trimethylamine N-oxide, which has been associated with ventricular remodeling, reduced cardiac function, and myocardial fibrosis (de Vos et al., 2022; Witkowski et al., 2020). In addition, gut barrier injury may induce systemic inflammation and further aggravate myocardial damage (Yu et al., 2025). Thus, profiling gut microbiota may provide additional information on the pathological pathway of AMI combined with CRF.

Transcriptomics provides another approach to characterize complex disease phenotypes (Chen et al., 2024). Previous AMI studies have identified differentially expressed genes (DEGs) related to endothelial ferroptosis and macrophage apoptosis (Xie et al., 2024). Clinical transcriptomic profiling of peripheral blood mononuclear cells from patients with renal injury has also identified prognostic biomarkers and pathways related to hypoxia, metabolic adaptation, inflammation, renal protection, and oxidative stress (Normahani et al., 2022). In addition, worsening renal function after AMI is common and is associated with adverse prognosis, further supporting the close interaction between myocardial injury and renal dysfunction (Jin et al., 2023). However, the molecular features of AMI combined with CRF remain incompletely understood.

In this study, we integrated transcriptomics, 16S rRNA sequencing, and untargeted metabolomics to explore candidate molecular, microbial, and metabolic features associated with AMI combined with CRF. Public Gene Expression Omnibus (GEO) datasets were used to identify shared immune-related hub genes in AMI and CRF. Clinical fecal and serum samples were then used to assess gut microbiota and serum metabolite differences between CRF patients without AMI and patients with AMI and advanced CKD. Finally, microbiota–metabolite correlations were analyzed to generate hypotheses for future validation.

Methods

Data collection

AMI-related datasets GSE194388 and GSE109048, and CRF-related datasets GSE180393 and GSE142153 were downloaded from the GEO database. GSE194388, including 5 AMI samples and 5 normal controls, and GSE180393, including 53 CRF samples and 9 normal controls, were used as independent discovery datasets. GSE109048, including 19 AMI samples and 19 normal controls, and GSE142153, including 20 CRF samples and 10 normal controls, were used only as external validation datasets for candidate hub genes. Because these datasets came from different studies, they were analyzed separately and were not merged into one expression matrix.

DEG analysis and functional enrichment

Differential expression analysis was performed separately in GSE194388 and GSE180393 using the limma package in R. No cross-dataset normalization or batch-effect correction was applied because datasets were not combined for DEG calling. DEGs were identified within each dataset, and genes with concordant expression changes in both datasets were retained for downstream analysis. Genes with p < 0.05 and |logFC| >1 were considered candidate DEGs. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed using clusterProfiler. Pathways with adjusted p values <0.05 were considered enriched.

PPI analysis and hub gene assessment

The shared DEGs were imported into STRING 11.0, with species set as humans, to construct a protein–protein interaction (PPI) network. Cytoscape 3.9.1 and the MCODE algorithm were used for module analysis. The MCC, DMNC, DNC, Gradient, and EPC algorithms in CytoHubba were used to select the top five hub genes from each algorithm. The intersecting hub genes were retained, and GeneMANIA was used to visualize their interaction network. Candidate hub genes were further evaluated in GSE109048 and GSE142153 by receiver operating characteristic (ROC) analysis using the pROC package. The area under the curve (AUC) was calculated. This analysis was considered exploratory because of the small discovery cohort and the lack of prospective validation.

Immune infiltration analysis

Immune infiltration was estimated in GSE194388 and GSE180393 using the CIBERSORT algorithm in the IOBR package with the LM22 immune cell matrix. The relative abundance of 22 immune cell types was calculated for each sample. Group differences between AMI and controls and between CRF and controls were analyzed. Spearman correlation was used to assess relationships between hub genes and immune cell abundance, and ggplot2 was used for visualization.

Participants and sample collection

Clinical omics analysis included 20 participants: CRF patients without AMI as the negative control (NC) group (n = 10) and patients with AMI and advanced CKD as the SY group (n = 10). Fecal samples from all participants were used for 16S rRNA sequencing, and matched serum samples were used for untargeted metabolomics. Microbiota–metabolite correlation analysis was performed in paired fecal and serum samples.

The SY group met AMI diagnostic criteria, including chest pain with dynamic cardiac troponin I elevation and ST-T changes, imaging evidence of new myocardial loss or regional wall motion abnormality, or angiographically confirmed coronary thrombus. The NC group included CRF patients without AMI or myocardial injury within the previous 3 months based on cardiovascular evaluation. Exclusion criteria included type 2 myocardial infarction secondary to infection, heart failure, or anemia; severe cardiac complications; obstructive nephropathy; renal insufficiency aggravated by previous percutaneous coronary intervention; severe infection or sepsis; drug abuse; pregnancy; and lactation. Peripheral venous blood and fecal samples were collected before coronary angiography or temporary dialysis. Serum was separated by centrifugation, and all samples were stored at −80°C. The study was approved by the Ethics Committee of Hunan Aerospace Hospital (HTYY2023LLSH-068-01), and all participants provided written informed consent.

16S rRNA sequencing and microbiota analysis

Fecal DNA was extracted using the E.Z.N.A. Stool DNA Kit. PCR products were purified with AMPure XP beads, quantified using a Qubit fluorometer, and assessed with an Agilent 2100 Bioanalyzer before sequencing on the NovaSeq PE250 platform. Operational taxonomic units were clustered using UPARSE at 97% similarity. Alpha diversity was assessed using Chao1 and Shannon indices, and beta diversity was evaluated using weighted UniFrac principal coordinates analysis in QIIME. LEfSe analysis was performed using p < 0.05 and LDA score ≥4. Random forest analysis and ROC curves were used as exploratory approaches to assess taxa-level discriminatory signals. Because of the small cohort and the absence of significant global diversity differences, microbiome findings were interpreted as exploratory.

Untargeted metabolomics and data processing

Serum samples were mixed with l-2-chlorophenylalanine as an internal standard. Proteins were removed with methanol, and dried extracts were redissolved in methanol–water. Metabolites were analyzed on a Thermo Fisher UHPLC-Q Exactive HF-X system using a Waters BEH C8 reverse-phase column. Raw data were processed with Compound Discoverer 3.1. After peak screening, MetaX was used for data transformation. Principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were performed. Candidate differential metabolites were screened using p < 0.05 and VIP > 1. Metabolite identification was based on mzCloud, mzVault, and MassList databases. Pathway enrichment analysis was performed using MetaboAnalyst and KEGG. Because this was a pilot metabolomics study, findings were interpreted as candidate signals. Detailed pooled-QC performance metrics and per-feature reproducibility statistics were not available and were considered a limitation.

Correlation analysis and statistics

Pearson correlation and two-way orthogonal PLS (O2PLS) were used to examine relationships between differential taxa and serum metabolites. Features with the top 10 joint loadings were selected for visualization. Data were presented as mean ± standard deviation. SPSS 21.0 was used for statistical analysis. Independent-sample t tests were used for normally distributed data, and rank-sum tests were used for nonnormally distributed data. p < 0.05 was considered statistically significant.

Results

Identification of candidate hub genes shared by AMI and CRF

Independent DEG screening in the discovery datasets GSE194388 (AMI) and GSE180393 (CRF) identified 460 shared genes with concordant expression trends across the two disease settings (Fig. 1A–C; top DEGs listed in Table 1). Functional enrichment analysis of these shared genes revealed a predominant convergence on leukocyte activation, cell adhesion, and immune response regulation. Specifically, GO and KEGG analyses highlighted pathways involved in cytokine signaling and structural components facilitating cell migration (Fig. 1D, E). While multiple “viral infection” pathways were enriched, these signatures reflect the activation of conserved innate immune and interferon-mediated modules characteristic of sterile inflammation in AMI and CRF, rather than active viral pathogenesis.

FIG. 1.

Screening of candidate hub genes shared by the AMI and CRF discovery datasets. The volcano plot of DEGs between AMI patients and normal controls (A) as well as between CRF patients and normal controls (B). (C) The Venn diagram of DEGs in the GSE194388 and GSE180393 datasets. The bar chart of the GO (D) and KEGG (E) enrichment analysis of the intersected DEGs with the same trend. (F) The MCODE clustering diagram of the intersected DEGs with the same trend. (G) The Venn diagram of the top five hub genes screened using MCC, DMNC, MNC, Degree, and EPC algorithms. (H) The PPI diagram of the hub genes. AMI, acute myocardial infarction; CRF, chronic renal failure; DEG, differentially expressed gene; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes.

Table 1.

Top Significantly Downregulated Genes in Acute Myocardial Infarction (GSE194388) and Chronic Renal Failure (GSE180393) Datasets

Dataset	Condition	Top downregulated DEGs (LogFC)	p-Value
GSE194388	AMI vs. control	AIG1, PRR7, GP6, CABP5, RP11–290H9.2, MAN2A2, TFP1, ITGB3, F2R, BMP6	<0.05
GSE180393	CRF vs. control	MAGI2-AS2, GSTCD-AS1, SERPINI2, ADGRF5-AS1, PSMD6-AS2, NEBL, LUC7L3, NKTR, TET2	<0.05

To pinpoint the drivers of this inflammatory network, topological analysis of the PPI landscape was performed using MCODE and multiple centrality algorithms. This screening strategy prioritized four candidate hub genes, namely, PTPRC, ITGAL, CD44, and SELL (Fig. 1F–H). All four genes encode key cell surface molecules involved in leukocyte trafficking and adhesion, suggesting that dysregulated immune cell migration may represent a shared immune-related feature linking AMI and CRF.

Preliminary discriminatory performance of the candidate four-gene signature and its association with immune infiltration

Before examining the association between the candidate hub genes and immune infiltration, we performed a preliminary evaluation of the four-gene signature in the independent GEO validation datasets. In GSE109048, the four-gene signature yielded an AUC of 0.941 for distinguishing AMI samples from controls (Fig. 2A). In GSE142153, the corresponding AUC for distinguishing CRF samples from controls was 0.986 (Fig. 2B). Although these findings suggest potentially favorable discriminatory ability, they should be interpreted with caution because the discovery-stage feature selection was based on small public datasets, and no truly independent clinical external cohort, resampling-based validation, or prospective validation was available in the present study. Therefore, these ROC results should be regarded as preliminary rather than definitive evidence of diagnostic performance. After that, immune infiltration in the GSE194388 and GSE180393 datasets was analyzed. It revealed that, compared with the healthy, the estimated proportion of neutrophils and T cells-CD4 memory resting was significantly increased (p < 0.05) in AMI, and the estimated proportion of T cells-CD8 was significantly decreased (p < 0.01) in AMI (Fig. 2C). Compared with the healthy, the estimated proportion of monocytes was significantly increased (p < 0.01), and the estimated proportion of CD4 memory resting was significantly decreased (p < 0.01) in CRF (Fig. 2D). In addition, the correlation between the four candidate hub genes and immune infiltration was analyzed. As shown in Figure 2E, the four candidate hub genes, PTPRC, ITGAL, CD44, and SELL, were positively correlated with immune infiltration in the GSE194388 dataset, among which ITGAL shows the strongest correlation with immune infiltration. As presented in Figure 2F, the four candidate hub genes, PTPRC, ITGAL, CD44, and SELL, were positively correlated with immune infiltration in the GSE180393 dataset, among which PTPRC showed the strongest correlation with immune infiltration. The above results indicated that the four candidate hub genes were positively correlated with the infiltration of multiple immune cell subsets.

FIG. 2.

Preliminary discriminatory performance of the candidate four-gene signature and its association with immune infiltration. ROC curves of the four-gene signature in the GEO validation datasets for AMI (A) and CRF (B). Boxplots of immune infiltration in AMI (C) and CRF (D). Heatmaps showing correlations between the candidate hub genes and immune infiltration in AMI (E) and CRF (F). *p < 0.05, **p < 0.01. ROC, receiver operating characteristic; GEO, Gene Expression Omnibus.

Changes in the gut microbiota of AMI patients with CRF

Gut microbiota analysis was performed in fecal samples from NC (n = 10) and SY (n = 10). Chao1 and Shannon indices did not differ significantly between the two groups, and weighted UniFrac PCoA also showed no significant between-group separation (Fig. 3A–C), indicating that global richness, diversity, and overall community structure were not clearly different in this cohort. We therefore interpreted subsequent taxonomic analyses as exploratory assessments of relative compositional differences rather than evidence of broad microbiome restructuring.

FIG. 3.

Changes in the gut microbiota of AMI patients with CRF. Gut microbiota analysis was performed in fecal samples from the NC group (n = 10) and SY group (n = 10). (A and B) Chao1 and Shannon indices of the gut microbiota in the NC and SY groups. (C) The β diversity of gut microbiota was analyzed by weighted UniFrac PCoA. The OTU of gut microbiota based on the phylum (D), species (E), and genus (F) levels. (G) The histogram of the LDA score in the NC and SY groups. (H) The phylogenetic tree of the abundance of gut microbiota in the NC and SY groups. The mean decrease in Gini of the gut microbiota at species (I) and genus (J) levels. The ROC curves were used to evaluate the classification results of the random forest model at the species (K) and genus (L) levels of gut microbiota. NC, negative control.

Within this exploratory framework, relative abundance analyses suggested differences at the phylum, species, and genus levels between the NC and SY groups (Fig. 3D–F). LEfSe analysis identified 24 differentially represented taxa, with several Enterobacterales-related taxa enriched in the SY group and several taxa including Bifidobacterium-related features enriched in the NC group (Fig. 3G, H). Random forest analysis highlighted Bifidobacterium dentium JCM 1195 DSM 20436 at the species level and Enterococcus at the genus level as potentially discriminative taxa within this small cohort (Fig. 3I, J). However, the corresponding ROC results, particularly the AUC of 1.0 for Enterococcus, should be interpreted cautiously because they were derived from a limited exploratory dataset and may reflect cohort-specific effects rather than reproducible diagnostic performance (Fig. 3K, L). Overall, our microbiome results suggest possible taxa-level compositional differences despite the absence of significant global diversity differences.

Changes in the metabolites of AMI patients with CRF

Untargeted metabolomic profiling was performed in serum samples from NC (n = 10) and SY (n = 10). The integrative microbiota–metabolite correlation analysis was performed in the subset of participants with paired fecal and serum samples (NC, n = 10; SY, n = 10). Metabolic profiling via UHPLC-Q Exactive analysis revealed distinct signatures between the SY and NC groups. PCA demonstrated clear separation along the first two principal components (Fig. 4A), a finding further corroborated by OPLS-DA, which exhibited significant group discrimination (Fig. 4B). Subsequent exploratory screening identified 20 candidate differential metabolites satisfying the predefined inclusion criteria (Fig. 4C, D). Among these, 7 metabolites were significantly upregulated, while 13 were downregulated in the SY group compared with controls. The full list of differential metabolites and their regulation trends is detailed in Supplementary Table S1.

FIG. 4.

Changes in the metabolites of AMI patients with CRF. Untargeted metabolomic analysis was performed in serum samples from the NC group (n = 10) and SY group (n = 10). The PCA (A) and OPLS-DA (B) analyses based on the results of UHPLC-Q Exactive detection. The volcano plot (C) and bubble chart (D) of the differential metabolites between the SY and NC groups. (E and F) KEGG and MetPA pathway enrichment analyses of differential metabolites. (G) The gut microbiota and metabolites with the top 10 joint loadings. (H) The correlation between the gut microbiota and metabolites with the top 10 joint loadings. OPLS-DA, orthogonal partial least squares discriminant analysis; PCA, principal component analysis.

Functional enrichment analysis (KEGG and MetPA) indicated that these metabolic shifts primarily mapped to amino acid metabolism and carbohydrate/energy metabolism (Fig. 4E, F). To elucidate the functional interplay between the gut microbiome and the metabolome, an O2PLS model was employed to identify features with high joint loadings (Fig. 4G). Correlation analysis further uncovered specific host-microbe interactions (Fig. 4H). Notably, the upregulated metabolite l-kynurenine exhibited a strong positive correlation with specific bacterial taxa, including Lachnoclostridium phocaeense, Ruminococcus bicirculans, and Veillonella atypica. These findings suggest that the distinct metabolic perturbations observed in AMI patients with CRF are intricately linked to dysbiosis of the gut microbiota.

Discussion

This study used transcriptomics, 16S rRNA sequencing, and untargeted metabolomics to identify candidate multiomic features associated with AMI combined with CRF (Glorieux et al., 2023; Hou et al., 2025; Li et al., 2023). The main findings were as follows. First, four immune-related hub genes, PTPRC, ITGAL, CD44, and SELL, were prioritized from shared AMI and CRF transcriptomic signals. Second, these genes showed preliminary discriminatory performance and were positively correlated with immune cell infiltration. Third, clinical omics analysis suggested selected taxa-level microbiota differences and differential serum metabolites, including increased l-kynurenine and decreased glucose, in patients with AMI and advanced CKD compared with CRF patients without AMI.

PTPRC, also known as CD45, regulates T- and B-cell activation through antigen receptor signaling (Al Barashdi et al., 2021). PTPRC has been identified as an immune-related gene in chronic kidney disease and has been linked to renal injury and immune infiltration (Fang et al., 2024). In AMI, sustained immune activation may promote cytokine release and impair ventricular remodeling (He et al., 2013). In our study, PTPRC showed strong positive correlations with immune infiltration, especially in CRF, suggesting that it may reflect inflammatory activation in the cardiorenal context. ITGAL encodes CD11a, the α-chain of LFA-1, and participates in leukocyte adhesion and migration (Lin et al., 2024). A previous study identified ITGAL as a potential target related to myocardial fibrosis and leukocyte extravasation (Cinato et al., 2025). Our findings suggest that ITGAL may link immune cell trafficking with both myocardial and renal injury. CD44 and SELL are also involved in immune cell migration. CD44 mediates leukocyte adhesion to hyaluronan, whereas SELL regulates T-cell homing to lymph nodes (Zhuo et al., 2006; Yuan et al., 2024). CD44 has also been linked to myocardial ischemia–reperfusion injury and hyperuricemia-induced renal injury through macrophage-related mechanisms (Ma et al., 2024; Zhu et al., 2025). Therefore, co-dysregulation of these genes may indicate a shared leukocyte trafficking program in AMI and CRF. However, these genes remain candidate biomarkers and require experimental and clinical validation.

Immune infiltration analysis supported the relevance of immune dysregulation. In AMI, neutrophils and resting memory CD4 T cells increased, whereas CD8 T cells decreased. Neutrophils are early responders after myocardial infarction and can remove necrotic tissue but may also aggravate tissue injury through reactive oxygen species and protease release (Ma, 2021). CD4 T-cell changes may reflect a pre-existing inflammatory state, while reduced CD8 T cells may indicate impaired immune repair or altered immune balance (Gao et al., 2021; Niederlova et al., 2023). In CRF, increased monocytes and reduced resting memory CD4 T cells were observed. Monocytes can infiltrate renal tissue, differentiate into macrophages, and release profibrotic cytokines, including TGF-β (Tang et al., 2019). Reduced CD4 T-cell function may reflect immune exhaustion in chronic kidney disease (Mathew et al., 2016). The overlap in immune alterations suggests a shared immune vulnerability between the heart and kidney. Renal-derived inflammatory mediators may amplify myocardial injury, while cardiac injury may aggravate renal inflammation (Hundae and McCullough, 2014). Experimental studies also indicate that TLR4/NF-κB/NLRP3 and IDO-related inflammatory signaling may contribute to myocardial or renal injury (Zhuo et al., 2025; Wu et al., 2023). These findings support future mechanistic studies of the hub gene–immune axis.

Microbiome analysis did not show significant alpha- or beta-diversity differences between SY and NC. Therefore, our data do not support broad microbiome restructuring. Instead, they suggest selected taxa-level differences. Enterobacterales-, Enterobacteriaceae-, and Escherichia–Shigella-related signals were relatively enriched in SY, whereas Bifidobacterium-related taxa and Leuconostoc were relatively enriched in NC. This pattern is biologically plausible because Enterobacteriaceae have been associated with uremic toxin generation and systemic inflammation, while Bifidobacterium species are related to short-chain fatty acid production and gut barrier maintenance (Wang and Zhao, 2025; Modrego et al., 2023). Enterococcus may carry inflammatory or virulence-related properties (Daca and Jarzembowski, 2024), but its apparent complete separation in this small cohort should be viewed only as an exploratory signal. Diet, medication use, and recent antibiotic exposure were not adjusted for and may confound microbiome associations.

Metabolomic analysis identified 20 candidate differential serum metabolites. l-Kynurenine was increased in SY. As a tryptophan metabolite produced through host enzymes and microbial activity, l-kynurenine may connect inflammation, gut microbiota, and organ injury (El-Zaatari et al., 2014). Previous studies reported associations between elevated l-kynurenine and AMI risk, oxidative stress, myocardial cell apoptosis, and renal fibrosis through aryl hydrocarbon receptor activation (Zhang et al., 2022; Liu et al., 2021). In our study, l-kynurenine was positively correlated with L. phocaeense, R. bicirculans, Clostridium spiroforme, V. atypica, and Blautia sp. YL58, suggesting a possible microbiota–kynurenine link in AMI combined with CRF. Glucose and several carbohydrate metabolites, including d-turanose and starch, were decreased in SY. These changes may reflect impaired energy metabolism, including altered myocardial glucose uptake during ischemia and reduced renal gluconeogenesis during CKD (Verissimo et al., 2022). Hypoglycemia has also been described as a poor prognostic marker in cardiorenal syndrome (Ensling et al., 2011). The reduction in Bifidobacterium may be related to altered carbohydrate fermentation and glucose homeostasis (Zhang et al., 2024). In addition, increased (−)-epigallocatechin may represent a compensatory antioxidant response (Ambigaipalan et al., 2020). These findings are hypothesis-generating and should be validated using targeted metabolomics and mechanistic experiments.

This study has several limitations. First, the clinical microbiome and metabolomics cohort was small, increasing the risk of false-positive findings and limiting generalizability. Second, the transcriptomic discovery datasets were also limited, especially GSE194388 with only 5 AMI cases and 5 controls. Third, AMI and CRF datasets were analyzed separately rather than as a combined AMI-with-CRF cohort, so shared genes should be interpreted as common immune-related signals rather than disease-specific drivers. Fourth, the ROC results may be optimistic because no prospective cohort, bootstrap validation, or repeated cross-validation was performed. Fifth, the cross-sectional design does not allow causal inference. Sixth, 16S rRNA sequencing provides taxonomic rather than functional information, and microbiome confounders were not adjusted for. Seventh, untargeted metabolomics was not followed by targeted validation, multiple-testing correction was not used in exploratory screening, and detailed QC metrics were incomplete. Finally, animal or cellular validation was not performed. Future studies should include larger independent cohorts, targeted metabolomics, metagenomics, longitudinal outcome data, and experimental validation.

Conclusion

Our multiomics study provides a systematic exploratory analysis of molecular features associated with AMI combined with CRF, identifying candidate hub genes, exploratory microbiota features, and candidate differential metabolites. These findings provide exploratory insights into cardiorenal syndrome and identify candidate biomarkers and candidate pathways for future mechanistic investigation, but the present study does not establish causality or therapeutic utility.

Human Ethics and Consent to Participate Declarations

This study was approved by the Medical Ethics Committee of Hunan Aerospace Hospital (Approval No. HTYY2023LLSH-068-01). All procedures involving human participants were conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to enrollment.

Data Availability

The data supporting this study are available from the corresponding author upon reasonable request.

Authors’ Contributions

Huijing L.: Conceptualization, data curation, formal analysis, investigation, methodology, validation, writing—original draft, and writing—review and editing; Huiqiong L., R.W., and M.Z.: data curation, investigation, methodology, and validation.

Footnotes

Author Disclosure Statement

The authors have no conflicts of interest to declare.

Funding Information

The funding needs to revised Sponsored by Science Foundation of Hunan Aerospace Hospital.

Supplemental Material

Abbreviations

References

Adak

, Khan

. An insight into gut microbiota and its functionalities. Cell Mol Life Sci, 2019; 76(3):473–493; doi: 10.1007/s00018-018-2943-4

Al Barashdi

, Ali

, McMullin

, Mills

. K. Protein tyrosine phosphatase receptor type C (PTPRC or CD45). J Clin Pathol, 2021; 74(9):548–552; doi: 10.1136/jclinpath-2020-206927

Ambigaipalan

, Oh

, Shahidi

. Epigallocatechin (EGC) esters as potential sources of antioxidants. Food Chem, 2020; 309:125609; doi: 10.1016/j.foodchem.2019.125609

Brankovic

, Akkerhuis

, Hoorn

, et al. Renal tubular damage and worsening renal function in chronic heart failure: Clinical determinants and relation to prognosis (Bio-SHiFT study). Clin Cardiol, 2020; 43(6):630–638; doi: 10.1002/clc.23359

Chen

, Zhang

, Li

, Meng

, Chi

, Liu

. Gut microbiota and its metabolites: The emerging bridge between coronary artery disease and anxiety and depression? Aging Dis, 2025; 16(3):1265–1284; doi: 10.14336/AD.2024.0538

Chen

, Zhang

, Liu

, Li

, Chi

. Multi-OMICS Research on Angina Pectoris: A Novel Perspective. Aging Dis, 2024; 16(6):3381–3399; doi: 10.14336/AD.2024.1298

Cinato

, Kang

, Kramar

, et al. Transcriptome fingerprinting of aberrant fibroblast activation unlocks effective therapeutics to tackle cardiac fibrosis. Sci Adv, 2025; 11(33):eadx0968; doi: 10.1126/sciadv.adx0968

Daca

, Jarzembowski

. From the friend to the foe-enterococcus faecalis diverse impact on the human immune system. IJMS, 2024; 25(4):2422; doi: 10.3390/ijms25042422

de Vos

, Tilg

, Van Hul

, Cani

. Gut microbiome and health: Mechanistic insights. Gut, 2022; 71(5):1020–1032; doi: 10.1136/gutjnl-2021-326789

10.

El-Zaatari

, Chang

Y-M

, Zhang

, et al. Tryptophan catabolism restricts IFN-gamma-expressing neutrophils and Clostridium difficile immunopathology. J Immunol, 2014; 193(2):807–816; doi: 10.4049/jimmunol.1302913

11.

Ensling

, Steinmann

, Whaley-Connell

. Hypoglycemia: A possible link between insulin resistance, metabolic dyslipidemia, and heart and kidney disease (the cardiorenal syndrome). Cardiorenal Med, 2011; 1(1):67–74; doi: 10.1159/000322886

12.

Fang

, Chen

, Qiu

, Huang

, Ke

. Identification and characterization of two immune-related subtypes in human chronic kidney disease. Transpl Immunol, 2024; 82:101983; doi: 10.1016/j.trim.2023.101983

13.

Gao

, Shi

, Gu

, et al. Difference of immune cell infiltration between stable and unstable carotid artery atherosclerosis. J Cell Mol Med, 2021; 25(23):10973–10979; doi: 10.1111/jcmm.17018

14.

Geng

, Xu

, Gao

, et al. Relationship between control of cardiovascular risk factors and chronic kidney disease progression, cardiovascular disease events, and mortality in chinese adults. J Am Coll Cardiol, 2024; 84(14):1313–1324; doi: 10.1016/j.jacc.2024.06.041

15.

Glorieux

, Nigam

, Vanholder

, Verbeke

. Role of the microbiome in gut-heart-kidney cross talk. Circ Res, 2023; 132(8):1064–1083; doi: 10.1161/CIRCRESAHA.123.321763

16.

, Teng

, Yu

, et al. Injection of Sca-1+/CD45+/CD31+ mouse bone mesenchymal stromal-like cells improves cardiac function in a mouse myocardial infarct model. Differentiation, 2013; 86(1–2):57–64; doi: 10.1016/j.diff.2013.07.002

17.

Headley

, Chapman

, Germain

, et al. Effects of high amylose-resistant starch on gut microbiota and uremic toxin levels in patients with stage-g3a-g4 chronic kidney disease: A randomized trial. J Ren Nutr, 2025; 35(2):248–258; doi: 10.1053/j.jrn.2024.09.005

18.

Hou

, Li

, Wu

, Wang

, Liao

, Dong

. Unraveling the gut-immune-kidney axis in kidney stone disease: A two-step Mendelian randomization investigation. Urolithiasis, 2025; 53(1):160; doi: 10.1007/s00240-025-01830-0

19.

Hundae

, McCullough

. Cardiac and renal fibrosis in chronic cardiorenal syndromes. Nephron Clin Pract, 2014; 127(1–4):106–112; doi: 10.1159/000363705

20.

Jin

, Seong

S-W

, Kim

, et al.; KAMIR-NIH Investigators. Prevalence and prognostic implications of worsening renal function after acute myocardial infarction. Am J Cardiol, 2023; 200:40–46; doi: 10.1016/j.amjcard.2023.05.011

21.

, Wang

, Li

, et al. The correlation between gut microbiome and atrial fibrillation: Pathophysiology and therapeutic perspectives. Mil Med Res, 2023; 10(1):51; doi: 10.1186/s40779-023-00489-1

22.

Lin

, Yang

, Huang

, et al. Magnesium-related gene ITGAL: A key immunotherapy predictor and prognostic biomarker in pan-cancer. Front Pharmacol, 2024; 15:1464830; doi: 10.3389/fphar.2024.1464830

23.

Liu

, Miao

, Deng

, Vaziri

, Li

, Zhao

. Gut microbiota-derived tryptophan metabolism mediates renal fibrosis by aryl hydrocarbon receptor signaling activation. Cell Mol Life Sci, 2021; 78(3):909–922; doi: 10.1007/s00018-020-03645-1

24.

, Xie

, Li

, et al. Weighted gene co-expression network analysis and single-cell sequence analysis uncover immune landscape and reveal hub genes of necroptosis in macrophages in myocardial ischaemia-reperfusion injury. Int Immunopharmacol, 2024; 140:112761; doi: 10.1016/j.intimp.2024.112761

25.

. Role of Neutrophils in Cardiac Injury and Repair Following Myocardial Infarction. Cells, 2021; 10(7):1676; doi: 10.3390/cells10071676

26.

Mathew

, Mason

, Song

, Tryniszewski

, Kennedy

. Role of T-regulatory cells in the response to hepatitis B vaccine in hemodialysis patients. Hemodial Int, 2016; 20(2):242–252; doi: 10.1111/hdi.12326

27.

Modrego

, Ortega-Hernández

, Goirigolzarri

, et al. Gut microbiota and derived short-chain fatty acids are linked to evolution of heart failure patients. Int J Mol Sci, 2023; 24(18):13892; doi: 10.3390/ijms241813892

28.

Moukarbel

, Yu

Z-F

, Dickstein

, et al. The impact of kidney function on outcomes following high risk myocardial infarction: Findings from 27 610 patients. Eur J Heart Fail, 2014; 16(3):289–299; doi: 10.1002/ejhf.11

29.

Niederlova

, Tsyklauri

, Kovar

, Stepanek

. IL-2-driven CD8(+) T cell phenotypes: Implications for immunotherapy. Trends Immunol, 2023; 44(11):890–901; doi: 10.1016/j.it.2023.09.003

30.

Normahani

, Boyle

, Cave

, Brookes

, Woollard

, Jaffer

. Peripheral blood mononuclear cell gene expression and cytokine profiling in patients with intermittent claudication who exhibit exercise induced acute renal injury. PLoS One, 2022; 17(3):e0265393; doi: 10.1371/journal.pone.0265393

31.

Tang

, Nikolic-Paterson

, Lan

. Macrophages: Versatile players in renal inflammation and fibrosis. Nat Rev Nephrol, 2019; 15(3):144–158; doi: 10.1038/s41581-019-0110-2

32.

van Veldhuisen

, Ruilope

, Maisel

, Damman

. Biomarkers of renal injury and function: Diagnostic, prognostic and therapeutic implications in heart failure. Eur Heart J, 2016; 37(33):2577–2585; doi: 10.1093/eurheartj/ehv588

33.

Verissimo

, Faivre

, Rinaldi

, et al. Decreased renal gluconeogenesis is a hallmark of chronic kidney disease. J Am Soc Nephrol, 2022; 33(4):810–827; doi: 10.1681/ASN.2021050680

34.

Wang

, Yang

, Li

, et al. Aberrant gut microbiota alters host metabolome and impacts renal failure in humans and rodents. Gut, 2020; 69(12):2131–2142; doi: 10.1136/gutjnl-2019-319766

35.

Wang

, Zhao

. Analysis of risk factors for poor prognosis in patients with renal insufficiency combined with Enterobacteriaceae bloodstream infection: A retrospective study. PeerJ, 2025; 13:e19993; doi: 10.7717/peerj.19993

36.

Witkowski

, Weeks

, Hazen

. Gut Microbiota and Cardiovascular Disease. Circ Res, 2020; 127(4):553–570; doi: 10.1161/CIRCRESAHA.120.316242

37.

, Zhong

, Lin

, Yan

. Blockade of Indoleamine 2,3-Dioxygenase attenuates lipopolysaccharide-induced kidney injury by inhibiting TLR4/NF-kappaB signaling. Clin Exp Nephrol, 2023; 27(6):495–505; doi: 10.1007/s10157-023-02332-2

38.

Xie

, Xie

, et al. A combined analysis of bulk RNA-seq and scRNA-seq was performed to investigate the molecular mechanisms associated with the occurrence of myocardial infarction. BMC Genomics, 2024; 25(1):921; doi: 10.1186/s12864-024-10813-1

39.

, Zhou

, Li

, et al. Targeting gut-immune-heart modulate cardiac remodeling after acute myocardial infarction. Life Sci, 2025; 371:123606; doi: 10.1016/j.lfs.2025.123606

40.

Yuan

, Su

, Gao

, et al. Med1 controls thymic T-cell migration into lymph node through enhancer-based Foxo1-Klf2 transcription program. Eur J Immunol, 2024; 54(10):e2350887; doi: 10.1002/eji.202350887

41.

Zhang

, Fang

, Zhang

, et al. The Effect of Bifidobacterium animalis subsp. Lactis MN-gup on glucose metabolism, gut microbiota, and their metabolites in type 2 diabetic mice. Nutrients, 2024; 16(11):1691; doi: 10.3390/nu16111691

42.

Zhang

, Cai

, Su

, et al. Untargeted metabolomics identified kynurenine as a predictive prognostic biomarker in acute myocardial infarction. Front Immunol, 2022; 13:950441; doi: 10.3389/fimmu.2022.950441

43.

Zhu

, Mu

, Ni

, et al. Xin-Zi-Sheng-Wan decoction alleviates hyperuricemia-induced renal injury by modulating SPP1-CD44-mediated macrophage infiltration and SRC/FAK/beta-catenin signaling pathway. J Ethnopharmacol, 2025; 353(Pt A):120305; doi: 10.1016/j.jep.2025.120305

44.

Zhuo

, Kanamori

, Kannagi

, et al. SHAP potentiates the CD44-mediated leukocyte adhesion to the hyaluronan substratum. J Biol Chem, 2006; 281(29):20303–20314; doi: 10.1074/jbc.M506703200

45.

Zhuo

, Ma

, Zhang

, et al. Qifu decoction alleviates lipopolysaccharide-induced myocardial dysfunction by inhibiting TLR4/NF-kappaB/NLRP3 inflammatory pathway and activating PPARalpha/CPT Pathway. Pharmaceuticals (Basel), 2025; 18(8):1109; doi: 10.3390/ph18081109

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB