Compositional alterations of gut microbiota in children with primary nephrotic syndrome after initial therapy

The study was approved by the ethical committee of Shanghai Children's Hospital (#R037-F01). We recruited 20 children with PNS who were hospitalized between September 2016 and October 2017. Fecal samples were firstly collected from children at onset of PNS before treatment (Group A). Then, fecal samples were collected when these patients achieved complete remission after 4-week initial therapy (Group B). These patients have taken prednisone (2 mg/kg/day, maximum dosage was 60 mg/day) and compound of vitamin D3 and calcium carbonate (one tablet/day, containing 300 mg calcium and 100 Unit vitamin D3) orally for 4 weeks.

Inclusion criteria: Children were diagnosed with PNS and achieved complete remission after 4-week initial therapy. Complete remission was defined as urine protein: creatinine ratio (uPCR) < 200 mg/g for 3 consecutive days in accordance with 2012 KDIGO Clinical Practice Guideline for Glomerulonephritis [14].We excluded PNS patients who had concomitant diseases, estimated glomerular filtration rate (eGFR) < 90 ml/min/1.73m², a history of gastrointestinal surgery, received antibiotic, probiotic and immunosuppressant treatment in the previous 2 months. Meanwhile, we collected clinical data including age, gender, delivery type, history of feeding types, eGFR, serum albumin, time to proteinuria resolution and the ratio of CD4+ to CD8 + T cells.

Fresh fecal samples were collected and stored in − 80 °C refrigerator until use. Total bacteria DNA were extracted by using QIAamp DNA stool Mini Kit (Qiagen, Hilden, Germany) as previously described [18]. The DNA concentration and purity were measured with a NanoDrop2000 spectrophotometer (Thermo Scientific, USA), and the integrity was evaluated by agarose gel electrophoresis.

The V3-V4 regions of bacteria 16S ribosomal RNA gene was amplified with the forward primer illumina adapter sequence1+ (5′-CCTACGGGNGGCWGCAG-3′) and reverse primer illumina adapter sequence2+ (5′-GACTACHVGGGTATCTAATCC-3′) as reported before [19]. PCR was performed in triplicate using a Gene Amp PCR-System 9700 (Applied Biosystems, Foster City, CA, USA) in a total volume of 25 μl, which contained 2.5 μl of 10× PCR buffer II, 0.5 unit of HerclueaseII DNA Polymerase High Fidelity (Agilent, USA), 0.4 μM of each primer, and 10 ng of template DNA. Thermal cycling conditions were as follows: an initial denaturation at 95 °C for 2 min, followed by 25 cycles at 95 °C for 20 s, 55 °C for 20 s, 72 °C for 45 s, and a final extension at 72 °C for 3 min. The quality of amplification products were checked using gel electrophoresis and were purified using the Agencourt AMPure XP Kit (Beckman Coulter, CA, USA). The sequencing of 16S rRNA gene amplicon was performed with the 2 × 250 bp paired-end method by using the Illumina MiSeq Bench top Sequencer [20]. The V3-V4 regions of 16S rRNA gene of fecal samples from 20 patients were sequenced. However, one sample from Group B was excluded as it failed to meet with the standard of sequencing experiment. A median depth of sequencing of 118828 reads per sample (80913–351500) was performed. All samples were sequenced by Genesky Biotechnologies Inc. (Shanghai, China).

Sample size was estimated by analyzing Species Accumulation Curve, and indicated that the 20 subjects met with the sample size calculation. Pairs of reads from the original DNA fragments were merged by using FLASH software (v1.2.11). The raw reads were checked with the QIIME quality filters under the default settings for Illumina processing. The qualified reads were chimera checked compared with the gold.fa database (http://​drive5.​com/​uchime/​gold.​fa) and clustered into operational taxonomic units (OTUs) by UPARSE pipeline with 97% similarity cutoff value. The OTUs were classified based on Ribosomal Database Project (RDP) Release 9,201,203 and alpha diversity including Chao1, ACE, Shannon, Simpson, InvSimpson and coverage index were analyzed by using Mothur software. Principal Coordinate Analysis (PCoA) using Bray-Curtis distance, Jaccard, unweighted and weighted UniFrac metric was performed with R Project (Vegan package, V3.3.1). Compositional changes of gut microbiota at different taxonomic levels were analyzed by using Metastats method. Microbial metagenome functional information was inferred from 16S rRNA gene data by the PICRUSt (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States) software using an extended ancestral-state reconstruction algorithm. The workflow of PICRUSt consists of gene content inference and metagenome inference. In brief, the OUTs were normalized upon 16 s rRNA gene copy number. Then, microbial community metagenomes were inferred and categorized into Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways after the normalized OUT-table was input [21, 22]. The bioinformatics analysis was performed by Genesky Biotechnologies Inc.(Shanghai, China).

Twenty children (male: female = 15:5) with PNS were enrolled in this study. All of them were Han Chinese ethnicity. Age of onset was 3.5 ± 2.1 years old. Nine patients were born by vaginal delivery and 11 by caesarean section. As for the patterns of infant feeding, the number of patients receiving formula feeding, breast feeding, formula and breast feeding were 3,12 and 5 respectively. Serum albumin at onset was 16.2 ± 4.1 g/L and the ratio of CD4+/CD8+ T cells was 1.8 ± 0.6. Patients achieved remission in 13.6 ± 5.3 days after initial therapy. All had normal renal function (eGFR 207.0 ± 52.9 ml/1.73m².min).

A total of 577 distinct OTUs were observed. The observed OTUs, Chao1 and ACE indexes were used for evaluating microbial richness, while the Shannon, Simpson, InvSimpson and Coverage index were the indicators of microbial diversity. It showed that the richness and diversity of gut microbiota were similar between Group A and B (Fig. 1). Since feeding type may influence microbiota composition directly [23], we analyzed the differences in composition of gut microbiota in three different groups (formula feeding, breasting feeding, formula and breast feeding group) before initial therapy. As showed in Additional file 1, the richness and diversity of gut microbiota among these three groups before initial therapy were similar. The results of the beta diversity included Bray-Curtis, Jaccard, unweighted and weighted UniFrac distances. Principal Coordinate Analysis (PCoA) created a scatter plot to show the phylogenetic tree-based distances between gut microbiota of individuals. Unweighted UniFrac qualitatively measures the inter-individual differences with or without each taxon, while weighted UniFrac quantitatively analyses inter-individual differences in the relative abundance of each taxon. The data on Fig. 2 revealed that fecal microbial community did not differ significantly in patients with PNS before and after initial therapy.

To explore the fecal microbial changes after initial therapy, the composition of gut microbiota at taxonomic levels was analyzed. At the phylum level, Firmicutes, Proteobacteria, Bacteroidetes and Actinobacteria dominated in the gut microbiota, but the increased relative abundance after initial therapy was observed in Deinococcus-Thermus and Acidobacteria (Additional file 2 and Additional file 3). As showed in Fig. 3, increased relative abundance of microbe at the genus level were listed as follows: Romboutsia, Stomatobaculum, Cloacibacillus,Howardella,Mobilitalea, Deinococcus, Paracoccus, Stenotrophomonas, Gp1, Kocuria, Pseudomonas, Acinetobacter, Brevundimonas and Lactobacillusbacteria. However, lower relative abundance of Finegoldia and Corynebacterium was found after initial therapy. The raw data are available in the Additional file 4.

To explore the possible microbial metabolic pathways, the functional profile of gut microbiota was analyzed by the PICRUSt method. As showed in Fig. 4, microbial metabolic pathways including selenocompound metabolism, isoflavonoid biosynthesis and phosphatidylinositol signaling system significantly weakened after initial therapy.