Aberrant Gut Microbiome Contributes to Barrier Dysfunction, Inflammation, and Local Immune Responses in IgA Nephropathy

IgAN is considered the most common primary glomerular disease and the leading cause of end-stage renal disease. IgAN can only be diagnosed by histopathological examination of renal biopsy specimens showing predominant or co-predominant mesangial IgA deposits [1]. The disease causes a wide spectrum of clinical manifestations. Some patients suffer no symptoms or just mild ones and require little or no treatment, while others progress [2, 3].

A number of mechanisms are thought to be responsible for IgA deposition in mesangial cells, including increased Gal-deficient (Gd)-IgA1 production, altered IgA receptors, and reduced IgA clearance [4, 6]. Various studies have shown that the percentage of the serum Gd-IgA1 levels of IgAN patients is increased, and these levels are a risk factor for IgAN [7, 9]. The relationship between IgAN and the mucosa was already known in the 1970s. In particular, the connection of IgAN with diseases in which the mucosa plays a role, especially coeliac disease, has been considered evidence of a gut-kidney axis [10]. A growing body of evidence suggests that disrupting the immune response of the gut lining is the root cause of IgAN. However, gut floras that grow on mucosal surfaces are exposed to intestinal epithelial cells and modulate intestinal immune responses by altering intestinal permeability and interacting with toll-like receptors (TLRs) expressed by gut mucosal cells [11]. Therefore, we conducted this study to investigate the role of barrier dysfunction, inflammation, and immune responses in IgAN by intestinal microbiome dysbiosis.

The ELISA kits for tumor necrosis factor-α (TNF-α; cat. no. SDH0001), lipopolysaccharide (LPS; cat. no. SDH0133), soluble intercellular adhesion molecule-1 (sICAM-1; cat. no. SDH0131), d-lactate (D-LAC; cat. no. SDH0134), and interleukin-6 (IL-6; cat. no. SDH0021) were purchased from Shanghai SIMUWUBio Co., Ltd. (Shanghai, China). A sandwich ELISA for Gd-IgA1 was constructed using KM55, which was purchased from Immuno-Biological Laboratories Co., Ltd. (Japan; cat. no. 1L-803). The primary antibodies of Western blot (TLR4: cat. no. Ab13556; MyD88: cat. no. Ab219413; NF-κB: cat. no. Ab32536; B-cell-activating factor [BAFF]: cat. no. Ab5965; a proliferation-inducing ligand [APRIL]: cat. no. Ab200836) were obtained from Abcam (UK). The glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody was purchased from Cell Signaling Technology (CST, USA; cat. no. 5174). The primary antibodies of immunohistochemistry were purchased from Abcam and CST (UK) (TLR4: cat. no. Ab22048; NF-κB: cat. no. 8242). Antibiotics (ampicillin, neomycin sulfate, and metronidazole) were purchased from Sangon Biotech Co., Ltd. (Shanghai, China). Vancomycin was obtained from VIANEX S.A. (PLANT C, Greece). The creatinine (Scr), blood urea nitrogen (BUN), and 24 h urinary protein activities were measured with an ARCHITECT Automatic Biochemistry Analyzer. A glycogen periodic acid Schiff (PAS) staining kit was purchased from Beijing Solarbio Science & Technology Co., Ltd. (Beijing, China).

The experiments and procedures were conducted in accordance with the Declaration of Helsinki of 1975 and were approved by the Ethics Committee of the Minhang Hospital, Fudan University (Shanghai, China) (No. 2021-056-01K), and written informed consent has been obtained from each subject. A total of 25 patients with IgA nephropathy (IgAN) who were treated at the Department of Nephrology of Fudan University Affiliated Minhang Hospital between October 2021 and December 2021 were included in the study. To avoid the effects of drugs and other hospital factors on intestinal microorganisms, we included only newly diagnosed IgAN patients who had not received any prior treatment and collected in patient samples before the initiation of any medication. The inclusion criteria were as follows: diagnosis of IgAN as confirmed by renal biopsy; presence (or absence) of microscopic hematuria; and estimated glomerular filtration rate ≥60 mL/min/1.73 m². There were 20 controls in the normal health group who visited the physical examination center of our hospital, and they were enrolled in the control group. All the patients were of Shanghai Han nationality, and their geographic area and eating habits were similar. The exclusion criteria were as follows: use of glucocorticoids or immunosuppressants or receipt of kidney transplant; use of oral antibiotics, probiotics, prebiotics or synbiotics within 2 months; or history of tumor, blood system disease, inflammatory bowel disease, irritable bowel syndrome, or other major digestive system diseases. Fecal, serum, and urine samples were collected from these patients and the controls. Stool samples were collected in sterilized 2-mL tubes containing bacterial cryopreservation fluid and frozen at −80°C until DNA extraction. For all patients with chronic nephritis, 5 mL of blood and urine were collected on an empty stomach on the early day of biopsy, then centrifuged, and in sterilized 2-mL tubes, and put in −80°C refrigerator for storage. Patient information was registered, and specimens from patients with IgAN confirmed by renal biopsy were removed for subsequent trials.

DNA amplification targeted the V3-V4 regions using the 338F (5'-ACT​CCT​ACG​GGA​GGC​AGC​A-3') and 806R (5'-GGACTACHVGGGTWTCTAAT-3') primers. The 5' ends of the primers were tagged with specific barcodes per sample and universal sequencing primers. PCR amplification was performed in a reaction mixture with a total volume of 25 μL containing 25 ng of template DNA, 12.5 μL PCR PreMix, 2.5 μL of each primer, and PCR-grade water to adjust the volume. The PCR conditions for the amplification of the prokaryotic 16S fragments consisted of an initial denaturation at 98°C for 30 s; 32 cycles of denaturation at 98°C for 10 s; annealing at 54°C for 30 s; and extension at 72°C for 45 s; then, a final extension at 72°C for 10 min. The PCR products were confirmed with 2% agarose gel electrophoresis. Throughout the DNA extraction process, ultrapure water, instead of a sample solution, was used to exclude the possibility of false-positive PCR results as a negative control. The PCR products were purified by AMPure XT beads (Beckman Coulter Genomics, Danvers, MA, USA) and quantified by Qubit (Invitrogen, USA). The amplicon pools were prepared for sequencing, and the size and quantity of the amplicon library were assessed on an Agilent 2100 Bioanalyzer (Agilent, USA) and with the Library Quantification Kit for Illumina (Kapa Biosciences, Woburn, MA, USA), respectively. The libraries were sequenced on NovaSeq PE250 platform [11, 12].

The serum concentrations of TNF-α and IL-6 were measured using enzyme-linked immunosorbent assay (ELISA) kits following the manufacturer’s protocols. The D-LAC, ICAM-1, and LPS levels in the serum are commonly considered indirect indicators of intestinal permeability. The serum concentrations of D-LAC, ICAM-1, and LPS were determined using ELISA kits. A sandwich ELISA for Gd-IgA1 was established using KM55 following the manufacturer’s protocols. The levels of Gd-IgA1 in each serum and urine sample were extrapolated by referring to a standard curve (4-parameter logistic curve fitting) of optical density (450 nm) and are expressed as units. Enzymatically generated Gd-IgA1 from human plasma and urine IgA1 was serially diluted (1.37–1,000 ng/mL) and used as a standard.

All the animal-related protocols were approved by the Institutional Animal Care and Use Committee of Fudan University (Shanghai, China) (No. 2022-JS-028). The mice were cared for and treated in accordance with the guidelines established by the Shanghai Public Health Service Policy on the Humane Care and Use of Laboratory Animals. Female C57BL/6J mice (3–5 weeks, 14–16 g) were obtained from the Shanghai Experimental Animal Centre of the Chinese Academy of Sciences (Shanghai, China) and maintained under specific pathogen-free conditions in a temperature-controlled colony room with a 12-h light/dark cycle. The mice were given access to rodent chow and water ad libitum. After a 1-week adaptation, the mice were used for experiments. The mice were mainly divided into two groups. The mice used for the antibiotic experiments were called the ABX mice, and the control mice were administered PBS instead of antibiotics and called the CBX mice. ABX mice were fed a mixture of antibiotics (1 g/L ampicillin, 1 g/L neomycin sulfate, 500 mg/L vancomycin, and 1 g/L metronidazole) in sterile water, which was sterile filtered through a 0.22-μm filter, for 4 weeks. The water bottles were changed once per week. In a pseudosterile mouse model, weight changes in the mice were observed, and a high-throughput sequencing method was used to detect fecal bacteria by 16S RNA sequencing to evaluate the effect of the model. The control mice were administered PBS instead of antibiotics.

The model of experimental IgAN was established by the immunity combination method of oral bovine serum albumin (BSA) (MedChemExpress, USA) administration and staphylococcal enterotoxin B (MedChemExpress, USA) injection [12, 14]. In brief, pseudo-germfree mice (ABX mice) were randomly divided into four groups: the blank group (ABX₀), model group (half-dose immunity combination (ABX_1/2), 75% immunity combination (ABX_3/4), and full-dose immunity combination (ABX₁) group; each group included 6 animals. C57BL/6J mice (CBX mice) were divided into two groups: the blank group (CBX₀) and model group (full-dose immunity combination (CBX₁); each group included 6 animals. The specific implementation plan was as follows: gavage with 0.1% BSA acidified water (0.4 mL/mouse) once every other day for 5 consecutive weeks; beginning in the 6th week, 1% BSA buffer (0.4 mL/mouse) was injected into the tail vein once a day for 3 days in the 7th and 8th weeks; and beginning in the 9th week, staphylococcal enterotoxin B (0.4 mg/kg) was injected into the tail vein once a week for 3 continuous weeks. The experiment ended in the 12th week. In the modeling group, the doses described above were administered. The blank control group was simultaneously administered equal amounts of 6 mmol/L acidified water by gavage and 0.01 mmol/L PBS by tail vein injection.

Intestinal and renal samples from the sacrificed mice were fixed in formaldehyde and embedded in paraffin. Deparaffinized 3–5-μm-thick sections were stained by hematoxylin-eosin staining. For histological analysis, deparaffinized 3–5-μm-thick sections of intestinal and renal tissues were treated with 0.3% H2O2 in methanol for 10 min to eliminate endogenous peroxidase activity, rinsed with distilled water, and immersed in PBS for 5 min. After the sections were blocked with normal goat serum for 30 min at room temperature, rabbit anti-TLR4 (diluted 1:200) and anti-NF-κB polyclonal antibodies (diluted 1:200) were added and incubated overnight at 4°C and then washed with PBS three times. The sections were then incubated for 30 min with biotinylated secondary antibodies (Boster, China) at 37°C and then for 30 min with streptavidin-biotin-peroxidase complex before being visualized by diaminobenzidine and counterstained with hematoxylin [15, 16].

Immediately after the mice were sacrificed, tissues were collected from the left kidney for immunofluorescence analysis. Frozen kidney sections of 3–5 μm thickness were air-dried on glass slides. The sections were stained with fluorescein isothiocyanate-labeled goat anti-mouse IgA (Sigma Chemical Co., St. Louis, MO, USA) (diluted 1:10) and allowed to stand overnight in a humidified box at 4°C. The slices were removed the next day and rewarmed at room temperature for 30 min. The corresponding immunofluorescence secondary antibodies were chosen and then incubated at 37°C for 30 min in the dark. The nuclei were stained with DAPI in the dark. Anti-fluorescence quencher reagent was added for mounting. Finally, a fluorescence microscope was used to observe the sections and capture pictures [15, 16].

Immediately after the mice were sacrificed, tissue from the left kidney was collected for the PAS technique. The sections were fixed in 10% formalin fixative, and then, the sections were conventionally dehydrated and embedded. Paraffin sections were dewaxed in distilled water. Frozen section was directly immersed in distilled water and restored to room temperature. After washing, the sections were placed in oxidant for 5–8 min at room temperature and then washed with distilled water twice. The section was incubated in Schiff Reagent in the dark at room temperature and stained for 10–20 min. The sections were washed in tap water for 10 min and stained with Mayer hematoxylin solution for 1–2 min. Then, the staining was differentiated by acidic differentiation solution for 2–5 s. The samples were washed with tap water for 10–15 min until they became blue, and then, the samples were conventionally dehydrated by series of ethanol. Finally, the samples were made transparent with xylene and sealed with resin.

Immediately after the mice were sacrificed, tissues were collected from the ileum and kidney for total RNA isolation according to the protocol for the TRIzol Reagent (Invitrogen, CA). cDNA was generated using the PrimeScript 1st Strand cDNA Synthesis Kit (TaKaRa, Japan). Relative quantitative real-time PCR was performed using SYBRs Premix Ex Taq^TM reagents (TaKaRa, Japan) on a Light Cycler (Roche Diagnostic). GAPDH was used as an internal reference. The PCR primers used were as follows: TLR4 forward, 5’-CAG​TCG​GTC​AGC​AAA​CGC-3’, and reverse, 5’-TAG​CCA​GGA​GCC​AGG​GAG-3’; MyD88 forward, 5’-ACT​CCA​CAG​GCG​AGC​GTA​C-3’, and reverse, 5’-CCA​GCA​AGG​TCC​AGT​CGG-3’; NF-κB forward, 5’-CGC​CCT​TTT​CGA​CTA​CGC-3’, and reverse, 5’-TCA​GGT​CAG​CCC​CAA​CCC-3’; BAFF forward, 5’-CCT​GGT​GAC​CCT​GTT​CCG-3’, and reverse, 5’-TCT​CCG​TTG​CGT​GAA​ATC​TG-3’; APRIL forward, 5’-TGG​TAT​CTC​GGG​AAG​GAC​AAG-3’, and reverse, 5’-CCA​TGC​GGA​GAA​AGG​CTA​AG-3’; and GAPDH forward, 5’-AAC​AGC​AAC​TCC​CAC​TCT​TC-3’, and reverse, 5’-TGG​TCC​AGG​GTT​TCT​TAC​TC-3’.

After the animals were executed, under sterile conditions, the ileum and kidney tissues were gently dispersed into a single-cell suspension, cultured in 1,640 culture medium (1 × 10⁶ cells/plate), and homogenized using RIPA lysis buffer (Beyotime Institute of Biotechnology, Jiangsu, China). The protein concentrations were determined using the Pierce BCA Protein Assay Reagent Kit (Rockford, IL, USA). The homogenates were diluted to the desired protein concentrations with 2× SDS-PAGE loading buffer (Invitrogen). The samples were boiled and loaded onto polyacrylamide minigels (Invitrogen) for electrophoresis. The proteins were transferred from the gels to Immobilon PVDF membranes (Millipore, Bedford, MA, USA) using a semidry apparatus (Bio-Rad, Hercules, CA, USA). Rabbit anti-BAFF (diluted 1:500), anti-APRIL (diluted 1:1,000), anti-TLR4 (diluted 1:500), and anti-NF-Κb (diluted 1:1,000) PcAb and Mouse anti-MyD88 McAb (diluted 1:1,000) (Abcam, UK) were used as the primary antibodies, and a horseradish peroxidase-conjugated goat anti-rabbit/mouse immunoglobulin-G antibody (diluted 1:1,000) was used as the secondary antibody.

The data are expressed as the mean ± standard error of the mean and were analyzed with either R (version 3.5.1) or GraphPad Prism 8 (GraphPad, San Diego, CA, USA). One-way ANOVA and post hoc least significant difference tests were used to determine the statistical significance of differences compared to the control. The differences between specific taxa at the phylum and genus levels were determined using the Kruskal-Wallis test for intergroup difference testing. Operating characteristic curves were constructed, and the area under the curve (AUC) was calculated to determine the discriminatory ability of the Lasso regression model. p values <0.05 were considered to indicate statistically significant differences.

The individual demographics, biochemical characteristics, and pathological features of the studied subjects are summarized in Table 1. The frequency of the observation of pathological features (percentages) in 25 biopsies scored according to the Oxford Classification are summarized in Table 2. There were no significant differences between the IgAN patients and controls in terms of age, sex, and body mass index as well as the glucose, total cholestrol, triglyceride, and blood C-reactive protein levels. However, the serum cystatin C, BUN, Scr, and 24 h urinary protein levels and microscopic hematuria were significantly increased, but the estimated glomerular filtration rate was significantly decreased in IgAN patients compared to healthy controls (HCs).

In our present microbiome investigation, PCR amplification of the V3-V4 region of the 16S rRNA gene was successful for all the collected samples. The Venn diagram visually presents the number of common and unique features of each group according to the results of the feature table and feature sequence (Fig. 1a). The two groups shared 2,196 common features. The IgAN group had 8,135 features, while the HC group had 8,606 features. To evaluate the differences in bacterial diversity between the two groups, the sequences were aligned to estimate the alpha diversity and beta diversity. There were statistically significant differences in Chao1 index and Good’s coverage, and there were no differences between the Shannon index and observed species (Fig. 1b), indicating that the richness of the microbial community in the gut microbiota was significantly lower in IgAN patients than in HCs. Based on the principal coordinates analysis plot, we observed a separation trend in the β diversity between the IgAN patients and HCs (Fig. 1c, d), suggesting that the fecal microbial community structure in the IgAN patients was significantly different from that in the HCs according to the presence of fecal OTUs.

Taxonomic analysis identified Firmicutes (52.68 vs. 72.48%, p < 0.05), Bacteroidetes (35.94 vs. 11.83%, p < 0.05), Proteobacteria (7.16 vs. 2.83, p < 0.05), Fusobacteria (0.87 vs. 0, p < 0.05), and Actinobacteria (3.17 vs. 12.59%, p < 0.05) were the most predominant taxa at the phylum level between the IgAN and HC groups (Fig. 2a). At the genus level, the most abundant taxa were Bacteroides (28.65 vs. 8.34%, p < 0.001), Faecalibacterium (10.29 vs. 7.75%, p < 0.05), Ruminococcus (6.7 vs. 3.52%, p < 0.05), Shigella (4.27 vs. 1.43%, p < 0.001), Bifidobacterium (2.05 vs. 9.67%, p < 0.05), Blautia (5.5 vs. 17.59%, p < 0.001), Roseburia (2.25 vs. 3.53%, p < 0.05), Coprococcus (1.95% vs. 7.23%, p < 0.01) between the IgAN and HCs groups (Fig. 2b). These data suggest that the taxonomic abundances of the microbial communities in IgAN patients and HCs are different.

To screen the diagnostic markers of the disease, we performed a taxonomic assignment of the sequences and analyzed each patient’s taxonomic profile using the LEfSe algorithm, representing significant differences between groups in the exploration cohort at all taxonomic levels (Fig. 2c, d). According to random forest analysis of the top 10 most abundant genera in the IgAN and HC groups, 10 candidate biomarkers were selected to predict the risk of IgAN in patients: Coprococcus, Bacteroides, Dorea, Bifidobacterium, Blautia, Parabacteroides, Gemmiger, Streptococcus, Clostridium, Lactococcus (Fig. 2e).

TNF-α and IL-6 in patients with IgAN were assayed. Higher levels of serum TNF-α (242.6 ± 25.18 pg/mL) and IL-6 (44.13 ± 5.22 pg/mL) were detected in IgAN patients than in healthy subjects (Fig. 3a, b) (*p < 0.05, **p < 0.01). The serum Gd-IgA1 level (5,711 ± 297.7 ng/mL) was significantly higher in the IgAN patients than in HCs. Compared with HCs, urine levels of Gd-IgA1 (48.16 ± 4.94 ng/mL) were higher in IgAN patients (p < 0.01) (Fig. 4a, b). Moreover, the serum and urine levels of Gd-IgA1 had good diagnostic efficacy with an AUC > 0.75. Notably, the urine level of Gd-IgA1 could best distinguish IgAN patients from HCs, with an AUC of 0.9952 (95% CI, 98.7–100; p < 0.0001) (Fig. 4c, d). And urine Gd-IgA1 levels were positively correlated with serum Gd-IgA1 levels. Interestingly, Coprococcus, Dorea, Bifidobacterium, Blautia, Lactococcus from 10 candidate biomarkers to predict risk in IgAN patients described above were strongly inversely correlated with urine Gd-IgA1 levels, Blautia was mildly inversely correlated with serum Gd-IgA1 levels. Bacteroides and Parabacteroides were mildly positively correlated with serum Gd-IgA1 levels (**p < 0.01, *p < 0.05) (Fig. 4e).

To investigate the establishment of pseudosterile mouse models after antibiotic lavage, we conducted 16S rRNA gene sequencing of feces. These data suggest that the taxonomic abundances of the microbial communities in the pseudosterile mouse model and HCs are different (Fig. 5a, b). There were statistically significant differences in the Chao1, goods_coverage, Shannon and observed_species indices (Fig. 5c), and compared with the HCs, the α diversity had significantly reduced in the pseudosterile mouse model. Based on the principal coordinates analysis plot, we found a trend of separation in the β diversity between these two groups (Fig. 5d, e). The results indicated that the microflora in the mouse intestine was basically suppressed, and the pseudo-germfree mouse model was successfully established. Then, we established a model of IgAN in pseudosterile mice. The results showed that the 24 h proteinuria concentration (24 h pro) was 1,902.7 ± 390.1 mg/L in the ABX₁ model group and 1,111 ± 164.3 mg/L in the CBX₁ model group, which was much higher than that in the ABX₀ control group (634.4 ± 40.67 mg/L) and the CBX₀ control group (658.3 ± 95.92 mg/L). Moreover, the levels of urea nitrogen (BUN) and blood Scr in the IgAN model group increased significantly (*p < 0.05, **p < 0.01) (Fig. 5f–h).

The results of immunofluorescence staining showed that the IgA was distributed in the mesangial area or in the capillary wall in the experimental model group, while the IgA level was negligible in each section of the control group (Fig. 5i–l). This result is consistent with the clinical presentation of IgAN. PAS staining technology was used to analyze the characteristics of pathological damage to kidney tissue, and the results are shown in Fig. 5m–p. The degree of kidney damage from high to low was as follows: ABX₁ group > CBX₁ group > ABX₀ group > CBX₀ group.

D-LAC, sICAM-1, and LPS levels were measured as markers of intestinal permeability [17, 19]. The results demonstrated that compared with the control group, the serum levels of D-LAC, sICAM-1, and LPS in the model group were significantly increased (*p < 0.05, **p < 0.01; Fig. 6a–c). Moreover, the IgA level in the IgAN model group was higher than that in the control group, and the level of IgA in the ABX₁ model group was higher than that in the CBX₁ model group (*p < 0.05, **p < 0.01; Fig. 6d). Moreover, the levels of serum TNF-α and IL-6 were elevated in pseudosterile mice with IgAN (**p < 0.01; Fig. 6e–f).

To analyze the activity of the inflammatory markers in pseudosterile mice with IgAN, immunohistological analysis of the intestines and kidneys from the various treatment groups described above was performed. We observed increases in inflammatory responses (increased TLR4 and NF-κB levels) (stained brownish yellow) in the kidneys in the model group than in the control group (Fig. 7a–d and Fig 7i–l). We also observed a similar response in the intestinal tissue (Fig. 7e–h and Fig. 7m–p). Furthermore, hematoxylin-eosin staining technology was used to analyze the characteristics of pathological damage to the intestinal tissues, and the results are shown in Figure 7a–d. The degree of intestinal damage from high to low was as follows: ABX₁ group > CBX₁ group > ABX₀ group > CBX₀ group.

To further confirm the role of the Inflammatory and local immune responses in IgAN, Western blotting and real-time PCR were used to analyze the expression of components of the inflammatory markers and B-cell-activating factor in renal and intestinal tissues. The results revealed that the expression of TLR4, MyD88, and NF-κB was significantly upregulated in the IgAN model group of renal (Fig. 8a1, a2) and intestinal (Fig. 8b1, b2) tissues compared with the control group. Moreover, the expression of BAFF and APRIL in intestinal tissue were elevated in the IgAN model group compared with the control group (Fig. 8c1, c2). The above correspond to the original films of the Western blotting experiments provided in online supplementary Figures S1, S2, and S3.

IgAN is diagnosed by the presence of bright IgA deposits within the mesangium. It is regarded as an immune-mediated disease featured by the deposition of polymeric and Gd-IgA1 in the mesangium [20]. Zachova et al. [21] reported that peripheral blood of IgAN patients is enriched with migratory λ⁺ mb-Gd-IgA1⁺ B cells, with the potential to home to mucosal sites where Gd-IgA1 could be produced during local respiratory or digestive tract infections. Zhang et al. [22] reported that the level of plasma Gd-IgA1 positively correlated with levels of UA, total IgA levels, and complement activation products. On the basis of this mechanism, we analyzed the Gd-IgA1 levels in the serum and urine of IgAN patients. Serum and urine Gd-IgA1 levels in the IgAN group were significantly higher than in HCs. Notably, the level of urine Gd-IgA1 could best distinguish IgAN patients from HCs. Gd-IgA1 plays a key role in the pathogenesis of IgAN. However, the exact mechanism of Gd-IgA1 production remains unelucidated [23]. Moreover, increasing evidence suggests that gut-renal junction plays a key role in Gd-IgA1 production [24]. Recently, a genome-wide association study revealed that most loci related with IgAN are also related with maintenance of the intestinal barrier [25]. Tan et al. [26] demonstrated that supplementation with probiotics mainly containing Bifidobacterium could markedly improve gut dysbiosis in IgAN.

Gut microbial composition and diversity differ in Italian Caucasian IgAN patients and are closely related to the clinical phenotype [25]. Our results showed that the intergroup diversity of the gut flora in the two groups, i.e., β diversity, could be completely separated. In our results, the α diversity was lower in the IgAN patients than in HCs. In further studies, we observed that the bacteria were very different between the two groups at the genus and phylum level. According to random forest analysis of the top 10 most abundant genera in the IgAN and HCs, 10 candidate biomarkers were selected to predict the risk of IgAN in patients: Coprococcus, Bacteroides, Dorea, Bifidobacterium, Blautia, Parabacteroides, Gemmiger, Streptococcus, Clostridium, and Lactococcus. We then analyzed the association of those 10 differential bacteria with serum and urinary Gd-IgA1 levels. Interestingly, Coprococcus, Dorea, Bifidobacterium, Blautia, Lactococcus were strongly inversely correlated with urine Gd-IgA1 levels. However, Blautia was mildly inversely and Bacteroides and Parabacteroides were mildly positively correlated with serum Gd-IgA1 levels. This suggests that Gd-IgA1 levels in urine have a greater value for IgAN diagnosis than Gd-IgA1 levels in serum and disturbed flora induced Gd-IgA1 production.

Collectively, our research and growing research indicate that the gut-kidney axis plays a role in the development of IgAN. However, to date, how disorders of the intestinal flora led to increased Gd-IgA1 production and thus to IgAN remains unclear. Prolonged and hyperactive activation of TLRs can ultimately lead to overproduction of IgA1/Gd-IgA1 [27]. Zheng et al. [28] reported that high TLR7 expression in B cells has dual roles in the development and progression of IgAN, by facilitating renal inflammation and Gd-IgA1 antibody synthesis. Makita et al. [29] demonstrated that TLR9 activation enhanced synthesis of aberrantly glycosylated IgA that, in a mouse model of IgAN, further enhanced kidney injury. Hence, APRIL and IL-6 synergistically, as well as independently, enhance synthesis of Gd-IgA1. Based on these findings, this study established a pseudosterile model to verify the vital role of gut microbiota in IgAN. After the end of antibiotic intervention, the diversity and abundance of the intestinal flora were decreased in the pseudosterile model mice compared with the normal specific pathogen-free mice, and at all the levels of bacterial classification, the proportion of intestinal bacteria was very different in the two groups, suggesting that antibiotic treatment can remove a large number of intestinal bacteria and indicating that the pseudosterile model was successfully established. Then, we generated IgAN models using the pseudosterile mice described above. The results of immunofluorescence staining showed that the IgA was distributed in the mesangial area or in the capillary wall in the pseudosterile IgAN mouse, while the IgA level was negligible in each section of the control group. The results revealed that the proteinuria concentration in the pseudosterile mice with IgAN was much higher than that in the control mice. Kidney function in pseudosterile mice with IgAN was much lower than that in the control mice. Several researchers report that a diverse and stable microbiome can promote human health, while an unbalanced gut microenvironment contributes to a variety of surrounding diseases [30]. Therefore, our results indicate that the normal flora exerts a protective effect on IgAN.

To further analyze the role of disordered gut microbiota in IgAN, we evaluated the levels of molecules related to the inflammation responses and gut barrier function in different groups. Peng SN demonstrated that rhein reduces intestinal permeability by protecting intestinal epithelial tight junction proteins ZO-1 and occludin, which alleviates the damage to the intestinal mucosa in IgAN [31]. Impairments in the intestinal barrier were assessed by intestinal permeability. Serum D-LAC, sICAM-1, and LPS levels were measured as markers of intestinal permeability [32, 33]. Our results demonstrated that the intestinal permeability (D-LAC, sICAM-1, and LPS) in pseudosterile mice with IgAN was significantly increased than that in control mice. Moreover, higher serum levels of TNF-α and IL-6 were detected in pseudosterile mice with IgAN than in the control groups. Meanwhile, immunohistological analysis was performed in the kidneys and intestines from the groups described above. Compared with the control, the removal of the gut microbiota with an antibiotic cocktail exerted a more potent stimulating effect on the inflammatory markers (TLR4 and NF-κB). Moreover, the mRNA expression of selected genes and protein related to inflammation markers was detected by RT-PCR and Western blotting analysis of intestinal and renal tissues. The results showed that the expression of components of the related molecules (TLR4, MyD88, and NF-κB) were significantly upregulated in pseudosterile mice with IgAN compared with the control mice. BAFF (also known as B-lymphocyte stimulator) and APRIL are cytokines expressed by antigen-presenting cells that play a crucial role in the development of B lymphocytes. Several pieces of evidence revealed that BAFF, APRIL, and BAFF-R are molecules also involved in autoimmunity [34]. We observed increases in local immune responses (increased BAFF and APRIL levels) in intestinal tissue were elevated in the IgAN model group compared with the control group.

In summary, our results show that there are significant differences in gut microbiota composition between IgAN and HCs. In addition, the normal flora exerts a protective effect on IgAN. Moreover, the serum and urine Gd-IgA1 levels were significantly increased in IgAN patients than in HCs. Furthermore, Gd-IgA1 in urine has a greater value for IgAN diagnosis than Gd-IgA1 in serum. Additionally, aberrant gut microbiome contributes to barrier dysfunction, inflammation, and local immune responses in IgAN. Therefore, this may be a potential therapeutic strategy for IgAN patients.

The experiments and procedures were conducted in accordance with the Declaration of Helsinki of 1975 and were approved by the Ethics Committee of the Minhang Hospital, Fudan University (Shanghai, China) (No. 2021-056-01K), and written informed consent has been obtained from each subject.

Results presented in this paper have not been published previously in whole or part, except in abstract format. The authors declare they have no conflicts of interest regarding the publication of this article.

This research was supported by the grants from the Suzhou Science and Technology Bureau of the application of the basic research project (No. SYS2020119), Jiangsu Province Traditional Chinese Medicine Science and Technology Development Plan Project (No. MS2021098), the Ministry of Education Industry-University Cooperation Collaborative Education Project (No. 202102242003), and Shanghai Minhang District High-Level Specialty Backbone Physician Training Program Funding Project (No. 2020MZYS19).

YuYan Tang, MingGang Wei, YiFan Zhu, and Yong Xiao conceived the study and designed the study protocol. YuYan Tang, YiFan Zhu, HaiDong He, Zhaowei Yan, and XuDong Xu performed the experiments and wrote the paper. Ping Hu, Zhen Liu, and WeiQian Sun analyzed the data. YuYan Tang prepared the manuscript, with editing and revision by all authors.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. The raw data of 16S sequencing for this study can be found in the PRJNA837902 in NCBI. Supplementary data to this article can be found online at https://www.ncbi.nlm.nih.gov/sra/PRJNA837902. Further inquiries can be directed to the corresponding author.

Yuyan Tang, Yong Xiao, and Haidong He contributed equally to this work.