In this study, a prospective clinical cohort study was conducted to analyze the gut microbiota in the feces of IgA nephropathy patients without segmental sclerosis, and IgA nephropathy patients with segmental sclerosis were used as controls. This study started in December 2022 and lasted for 2 years. All fecal samples in this study were from the participants in Zhejiang, China.

At the beginning of the study, fecal samples were collected from 750 patients who underwent renal biopsy in our hospital. A total of 434 samples were obtained, of which 207 samples were diagnosed as IgAN. Excluding the secondary IgAN and the unclear part of the Oxford classification, the remaining 180 patients were further screened according to the Oxford classification. Twelve patients with pathological diagnosis of IgA nephropathy and Oxford classification of M1E0S0T0C0 and 19 patients with Oxford classification of M1E0S1T0C0 were selected to study the characteristics of intestinal microbial composition in the two groups (Figure 1).

Diagnostic standard were as follows:

Pathological diagnostic criteria of IgA nephropathy:

Immunofluorescence showed granular and mass-like deposition of IgA or IgA deposition in the mesangial area, with or without other immunoglobulins (Kidney Disease: Improving Global Outcomes (KDIGO) Glomerular Diseases Work Group, 2021).

Oxford classification diagnostic criteria:

M refers to mesangial cell proliferation, which is divided into M0 (no or mild mesangial cell proliferation) and M1 (moderate or severe mesangial cell proliferation) according to the degree of mesangial cell proliferation. E refers to the proliferation of capillary endothelial cells, E0 means no, E1 means yes; s refers to segmental sclerosis or adhesion, S0 is none, S1 is yes; t refers to renal tubular atrophy or renal interstitial fibrosis, T0 represents renal tubular atrophy or renal interstitial fibrosis ≤ 25%, T1 represents 26% -50%, T2 represents > 50%; c represents cell or cell fibrous crescent, C0 is absent, and C1 is present (Trimarchi et al., 2017).

The inclusion criteria were as follows:

The exclusion criteria were as follows:

A fresh fecal sample is taken through a rectal swab. The patient is instructed to thoroughly clean the anal area with water and 75% alcohol before collection, scrape the fecal sample expelled from the anus using a sterile swab, place it in a sterile tube, and store it immediately in a -80°C freezer.

Total genomic DNA of microbial communities was extracted from fecal samples following the instructions using the FastPure Stool DNA Isolation Kit. Subsequently, the integrity of the extracted genomic DNA was checked by 1% agarose gel electrophoresis, and the concentration and purity of the DNA were determined using a NanoDrop2000 spectrophotometer.

Using the extracted DNA as a template, specific primers carrying Barcode sequences, 338F (5 ‘-ACTCCTACGGGAGGCAGCAG-3’) and 806R (5 ‘-GGACTACHVGGGTWTCTAAT-3’), were used to perform PCR amplification of the V3-V4 variable region of the 16S rRNA gene. The PCR products from the same sample were mixed and then purified using 2% agarose gel electrophoresis to recover the target fragment. The size of the recovered fragment was determined by 2% agarose gel electrophoresis, and the purified product was quantified using Synergy HTX. The purified PCR products were indexed using NEXTFLEX Rapid DNA-Seq Kit and sequenced on the NextSeq 2000 PE300 platform provided by Illumina.

Fastp software was used to control the quality of the double-ended original sequencing sequence, and FLASH software was used for splicing. Used the cutadapt software to identify and remove the primer sequence to obtain Clean Reads; the DADA2 method in QIIME2 software was used to denoise the sequence to obtain amplicon sequence variants (ASVs), Merged Reads and remove the chimera sequence to obtain the final valid data (Non-chimeric Reads). The Silva 16SrRNA gene database (v138) was annotated by RDP classifier, and the confidence threshold was 70%, and the community composition of each sample was counted at different species classification levels. PICRUSt2 (version2.2.0) software was used for 16S function prediction analysis. Alpha diversity index was calculated by mothur software, and the difference of Alpha diversity between groups was statistically analyzed by Wilcoxon rank sum test. Principal component analysis based on Bray-Curtis distance algorithm was used to evaluate the similarity of microbial community structure between samples, and non-parametric test was used to analyze the significant differences of microbial community structure between groups. In addition, LEfSe (Line Discriminant Analysis Effect Size) analysis was used to identify bacterial groups with significant differences in abundance between groups from phylum to species level. In the statistical analysis, P value less than 0.05 was considered statistically significant (Figure 2).

In the comparison of general clinical data, S0 and S1 were statistically different in age (p=0.004) and not statistically different in terms of gender and course (p>0.05). In the comparison of test result, S0 and S1 did not show statistical difference in leucocyte, hemoglobin, thrombocyte, prothrombin time, thrombin time, d-dimer assay, fibrinogen, serum creatinine, blood albumin, urine red blood cell, 24-hour urine protein and glomerular filtration rate (p>0.05) (Table 1).

Based on the concept clustering of ASVs, the results showed that there were 995 ASVs in S1 group and 888 ASVs in S0 group, and the ASVs in S1 group were more than those in S0 group. According to the ASV Wayne diagram, it can be seen that there are 694 ASVs shared by S1 group and S0 group, 301 ASVs owned by S1 group alone, and 194 ASVs owned by S0 group alone. This preliminary indicates that there are some differences in the types and composition of intestinal microorganisms between the two groups (Figure 3).

In this study, LEfSe analysis, combined with Wilcoxon rank sum test and LDA (Linear discriminant analysis) effect size, was used to find intestinal biomarkers with statistical differences between S1 group and S0 group. The LDA histogram showed that there were five different species between the S1 group and S0 group. Among them, there were two dominant bacteria in the S1 group, which were Subdoligranulum and unclassified_Erysipelotrichaceae_UCG-003, suggesting that these two bacteria have the potential to be used as intestinal microbial markers for segmental glomerulosclerosis. There were three dominant bacteria in the S0 group, namely Phascolarctobacterium, Streptococcus_parasanguinis, and Proteobacteria, suggesting that these three bacteria have the potential to be used as intestinal microbial markers for non-segmental glomerulosclerosis (Figure 11).

Kyoto Encyclopedia of Genes and Genomes (KEGG) is one of the databases commonly used in pathway research. KEGG describes a series of metabolic pathways and the relationship between metabolic pathways, covering information such as metabolism, genetic information, environmental information, and cellular processes. In this study, Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) software was used to compare the 16SrRNA characteristic sequence with the KEGG microbial functional genome database, and the gene information of unknown species was predicted according to the gene type and abundance information of known species, so as to predict the pathway of intestinal microbial community in the two groups.

The results of PICRUSt2 function prediction showed that the two groups of samples showed significant differences in 18 functional pathways. Among them, the S1 group samples were significantly enriched in 7 functional pathways (P < 0.05), including Protein processing in endoplasmic reticulum, Linoleic acid metabolism, Secondary bile acid biosynthesis, Primary bile acid biosynthesis, Dilated cardiomyopathy, Arrhythmogenic right ventricular cardiomyopathy, and Pancreatic secretion. The S0 group samples were significantly enriched in 11 functional pathways (P < 0.05), including Ubiquinone and other terpenoid-quinone biosynthesis, Biosynthesis of Siderophore Group Nonribosomal Peptides, Drug metabolism - cytochrome P450, Metabolism of xenobiotics by cytochrome P450, Lipoic acid metabolism, Geraniol degradation, Ion channels, Retinol metabolism, Caprolactam degradation, African trypanosomiasis, and Ubiquitin system (Figure 12).

In IgA nephropathy (IgAN), segmental glomerulosclerosis (S lesion) defined by the Oxford classification is a critical pathological feature and an independent risk factor for disease progression, closely associated with increased proteinuria, accelerated decline in glomerular filtration rate (GFR), and poor renal prognosis (Bellur et al., 2017). Subtypes of S lesions, such as tip lesions and podocyte hypertrophy, indicate active podocytopathy and are linked to higher baseline proteinuria and more rapid renal function deterioration, particularly in patients without immunosuppressive treatment (Bellur et al., 2017). The presence of S lesions also reflects complex pathological mechanisms, including podocyte injury, complement activation, and mesangial cell dysfunction, which collectively drive glomerular scarring and tubulointerstitial fibrosis (Xie et al., 2015; Chiu et al., 2021). Our study further contextualizes these findings by identifying distinct gut microbial signatures in S lesion-positive (S1) and S lesion-negative (S0) IgAN patients. Specifically, the S1 group was characterized by enriched Subdoligranulum and unclassified_Erysipelotrichaceae_UCG-003, which may exacerbate glomerular injury through pathways like endoplasmic reticulum stress and dysregulated bile acid metabolism (Bellur et al., 2017). In contrast, the S0 group harbored bacteria associated with xenobiotic metabolism and redox balance, potentially reflecting a less aggressive inflammatory milieu (Bellur et al., 2017). These results align with prior research demonstrating that S lesions in IgAN are not merely markers of chronicity but active drivers of disease progression, underscoring the importance of integrating pathological subclassification with microbiome profiling to refine risk stratification and therapeutic strategies in IgAN.

The formation of segmental sclerosis in IgA nephropathy (IgAN) involves multiple interacting mechanisms. Podocyte injury is central, with gene mutations (e.g., NPHS2, TRPC6) disrupting the filtration barrier or calcium homeostasis to drive proteinuria and sclerotic progression (Singh et al., 2022). Soluble urokinase receptor (suPAR) exacerbates podocyte loss via integrin-mediated oxidative stress and Src kinase activation (Kim et al., 2018). Immune dysregulation plays a key role, including activation of complement pathways (elevated C3a, C5a, C5b-9) (Huang et al., 2020), anti-nephrin antibodies (Shirai et al., 2024), and B-cell abnormalities (e.g., increased memory B cells and plasmablasts) (Liu et al., 2021). In transplanted kidneys, metabolic/structural changes—such as podocyte overexpression of vascular endothelial growth factor A (VEGF-A) disrupting endothelial function and parietal epithelial cell dedifferentiation promoting collagen deposition—are critical (Gallon et al., 2025). Cytoskeletal proteins (e.g., Btg2 inducing podocyte-mesenchymal transition via the Smad3 pathway) (Dan Hu et al., 2023) and mitochondrial dysfunction (oxidative stress, dynamics defects) (Li et al., 2023) further propagate sclerosis. Genetic factors (e.g., APOL1 risk variants in African populations) (Bruno and Edwards, 2021), hemodynamic stress (obesity-related glomerular hyperfiltration) (Sandino et al., 2021), and cellular senescence (tubular p16INK4A upregulation) (Verzola et al., 2020) act as independent risk factors. Our study’s findings of distinct intestinal microbial markers between segmental sclerosis-positive (S1) and -negative (S0) groups suggest gut microbiota dysregulation may contribute to segmental sclerosis via the gut-kidney axis, potentially through metabolic or immune cross-reactivity pathways, offering new insights into its pathogenesis.

Unclassified_Erysipelotrichaceae_UCG-003, a member of the Erysipelotrichaceae family, is associated with detrimental effects in the S1 group (segmental glomerulosclerosis-positive IgAN). Erysipelotrichaceae dysbiosis leading to tryptophan metabolism imbalance promotes increased precursors such as indole and p-cresol, which are subsequently converted in the liver into protein-bound uremic toxins such as indoxyl sulfate (IXS) and p-cresyl sulfate (pCS) (Miao et al., 2025). These toxins, particularly their free forms, show a strong negative correlation with renal function (eGFR) and are reliable biomarkers for CKD progression (Corradi et al., 2024). They act as endogenous ligands activating signaling pathways such as AHR and NF-κB (Miao et al., 2025), promoting renal inflammation and fibrosis (He et al., 2025).

In the S1 group (IgAN with segmental glomerulosclerosis positive), Subdoligranulum demonstrates detrimental effects associated with metabolic inflammation and intestinal barrier dysfunction, which may further exacerbate the pathological progression in this subgroup. Changes in its abundance are linked to metabolic disturbances and may impair intestinal barrier integrity by reducing the production of short-chain fatty acids (SCFAs) (Kwiatkowska et al., 2024; Li et al., 2025). Therefore, the enrichment of Subdoligranulum and Erysipelotrichaceae UCG-003 in the S1 group may exacerbate the progression of IgA nephropathy through a synergistic and potentially additive mechanism involving damage to intestinal barrier function and increased bacterial endotoxin translocation (Cuartero et al., 2024; Li et al., 2024; Jin et al., 2026), which promotes the entry of uremic toxin precursors (such as indole and p-cresol) into systemic circulation and further activates signaling pathways including AHR, NF-κB, and PKC/PI3K (He et al., 2025).

In contrast, the S0 group (segmental glomerulosclerosis-negative IgAN) is enriched with potentially protective bacteria. Phascolarctobacterium is a known SCFA-producing bacterium, and its primary metabolite, propionate, belongs to the SCFA family (Li et al., 2025; Ni et al., 2025). Studies have indicated that SCFAs (e.g., propionate, butyrate) can inhibit the NF-κB pathway and enhance intestinal barrier integrity, thereby alleviating renal inflammation and fibrosis (He et al., 2025; Jin et al., 2026). Mendelian randomization studies support that the renal benefits of protective gut microbiota may be mediated by modulating the host’s blood metabolome (e.g., increasing levels of beneficial metabolites) (Cao et al., 2025). Additionally, research suggests that natural products (e.g., certain polysaccharides) can increase SCFA production by modulating the gut microbiota, further improving kidney disease (Zhang et al., 2022). Furthermore, Phascolarctobacterium interacts with Bacteroides thetaiotaomicron by utilizing succinate to generate propionate, promoting microbial symbiosis (Ikeyama et al., 2020; Ni et al., 2025).

Streptococcus_parasanguinis, a dominant bacterium in the S0 group, contributes to gut health via oral-gut axis interactions. As an oral commensal, it produces hydrogen peroxide to modulate microbial diversity and adheres to collagen, supporting mucosal integrity (Fultz et al., 2022; Li et al., 2025). In mouse pups, breast milk-derived S. parasanguinis upregulates anti-inflammatory Treg-related genes (e.g., Foxp3, TGF-β) in the ileum (Tsunoda et al., 2018; Fultz et al., 2022), a mechanism that may reduce glomerular immune complex deposition in IgAN. Its depletion in periodontitis patients correlates with mucosal barrier dysfunction (Chen et al., 2019; Fultz et al., 2022), highlighting its protective role.

The Proteobacteria phylum in the S0 group comprises species with diverse metabolic functions. Notably, certain members of this phylum (such as Escherichia coli) have been clearly demonstrated to metabolize tryptophan, producing key precursors like indole, which play a critical role in regulating immune balance and maintaining intestinal barrier function (Miao et al., 2025). Large-scale blood metabolome analyses indicate that specific gut microbiota are significantly associated with host circulatory levels of lipid metabolites and amino acid metabolites (Cao et al., 2025). Furthermore, this study found that specific Proteobacteria subgroups, including the model strain Escherichia coli str. K-12 substr. MG1655, are capable of elevating alpha-ketoglutarate (α-KG) levels (Cao et al., 2025). This metabolite exhibits anti-inflammatory and antioxidant properties, suggesting a potential inhibitory role in the progression of glomerulosclerosis. In this study, the Proteobacteria in the S0 group may include more beneficial species, which maintain ecological balance by promoting nutrient cycling and reducing the accumulation of toxic metabolites (Deng et al., 2021; Barron et al., 2024; Li et al., 2025).

Functional pathway analysis using PICRUSt2 revealed distinct molecular mechanisms differentiating the S1 (sclerotic) and S0 (non-sclerotic) phenotypes. The S1 group was characterized by enrichment in pathways promoting endoplasmic reticulum (ER) stress, such as “Protein processing in endoplasmic reticulum.” This involves mutant proteins (e.g., INF2, α-actinin-4) disrupting ER function, activating the unfolded protein response (UPR), and inducing podocyte apoptosis via caspase-12, supported by evidence of upregulated BIP and p-eIF2α in IgAN glomeruli (Lan et al., 2023; Tran et al., 2024). Dysregulated “Linoleic acid metabolism” in S1 promotes pro-inflammatory lipid peroxidation, generating reactive oxygen species that damage podocyte cytoskeletons (e.g., through α-actinin-4 oxidation) and link gut microbiota-driven inflammation to reduced SCFA production (Wu et al., 2022; Deng et al., 2024). Additionally, aberrant “Primary/secondary bile acid biosynthesis” in S1 exacerbates injury via FXR-TGR5 pathway activation and mitochondrial apoptosis induced by secondary bile acids (e.g., deoxycholic acid), with reduced Bifidobacterium abundance in IgAN patients connecting gut dysbiosis to bile acid toxicity (Widmeier et al., 2019; Deng et al., 2024).

In contrast, the S0 group maintains protective pathways including “Ubiquinone biosynthesis,” which preserves mitochondrial integrity via coenzyme Q10 and mitigates APOL1-associated ROS production (Wada et al., 2023; Mengarelli et al., 2025; Wang et al., 2025), and “Cytochrome P450 drug/xenobiotic metabolism,” which facilitates detoxification of nephrotoxic compounds (e.g., tacrolimus, bile acids) via enzymes such as CYP3A4 and CYP2C9 (Deng et al., 2022; Chebotareva et al., 2023). A robust “Ubiquitin system” in S0 ensures proteasomal clearance of misfolded proteins (e.g., nephrin, TRPC6), preventing aggregate formation and steroid resistance observed in S1, which is partly attributed to anti-UCH-L1 autoantibodies (Jia et al., 2025; Goeckeler-Fried et al., 2026).

Cross-pathway analysis underscores a fundamental contrast: S1 exhibits ER stress-overwhelmed proteostasis and pro-inflammatory lipid signaling, whereas S0 maintains mitochondrial-lipid metabolic balance and efficient gut-renal detoxification. This highlights the critical role of ER stress–ubiquitin–proteasome system crosstalk and microbiota–bile acid–CYP450 interactions in driving disease progression or conferring protection (Widmeier et al., 2019; Deng et al., 2024; Tran et al., 2024; Wang et al., 2025).