Association Between Fecal Microbiota Composition, Metabolite Concentrations, and Indoxyl Sulfate Levels

Overview

Gut dysbiosis is frequently characterized by decreased microbial diversity and alterations in the abundance of certain microbial species. In individuals with chronic kidney disease (CKD), dysbiosis and metabolic imbalances are prevalent, contributing to the buildup of gut-derived retention solutes and metabolites in the bloodstream. Research has consistently shown that CKD patients exhibit lower levels of beneficial gut bacteria. However, the specific functional changes in gut microbiota and their interactions with levels of uremic toxins in hemodialysis (HD) patients remain incompletely understood. This study seeks to explore the association of fecal metagenomics and targeted metabolomics in a cohort of 60 patients with different levels of to characterize the complex interplay between the gut microbiome and fecal and serum metabolites.

Patients and Study Design Sixty HD patients will be enrolled, from dialysis unit, Renal Division of The Tungs' Taichung Metroharbour Hospital. The written informed consent to take part in the study is obtained from each participant after being informed on study design and aims. The experimental protocol will send to the Ethical Committees of IRB for approval. Inclusion criteria are age 18-80 years and diagnosed with CKD stage V (currently receiving hemodialysis treatment\>3 months). Exclusion criteria included pregnant or nursing women; patients with kidney transplant, severe infections, severe cardiac diseases and liver diseases, malignancy, autoimmune disorders, severe malnutrition; consumed any type of pre-or probiotics or had antibiotic therapy within 1 month of study commencement; diagnosed irritable bowel syndrome, Crohn's disease, or ulcerative colitis; receiving or have received bowel radiation or had large bowel resection The participants will receive dietary instructions and are encouraged to maintain stable dietary intake, which is quantified by nutritional questionnaires during patient interviews. The questionnaires gathered information about the weekly amount and frequency of food and beverage consumption. Data regarding the nutrient composition of the different foods are obtained using the tables of the Institute of Nutrition and Food Safety, National Taiwan University. Sample Collection, Preparation and Analysis Blood samples are collected separately from each participant. Blood samples are collected after an 8 to 12-hour fasting period, using sodium fluoride tubes, and kept on ice until transfer to the research laboratory. Once there, the samples are centrifuged at 3,000×g for 15 minutes at 4°C. The serum is then transferred to sterile tubes and stored at -20°C for future analysis. investigators measure serum levels of acetate, propionate, butyrate, and valerate, along with branched-chain SCFAs such as isobutyrate and isovalerate. These metabolite profiles are analyzed using ultraperformance liquid chromatography-mass spectrometry (UPLC-MS). In brief, 100 μL of each serum sample is mixed in 1.5-mL microtubes with 20 mg of NaCl, 10 mg of citric acid, 20 μL of 1 M HCl, and 100 μL of butanol. The mixture is vortexed for 2 minutes and then centrifuged at 18,000×g for 15 minutes. The resulting supernatant is transferred to fresh microtubes for analysis. Measurement of uremic toxins Investigators measure total uremic toxin levels rather than free levels. The concentrations of total indoxyl sulfate, p-cresyl sulfate, indole-3-acetic acid (IAA), and hippuric acid are centrally quantified using a previously described method \[19\]. Briefly, for binding competition, 200μl serum to which investigators added 20μl 0.50mM 1-naphthalenesulfonic acid (internal standard) was vortex-mixed with 250μl 0.24M sodium octanoate (binding competitor).After incubation at room temperature for 5min, investigators added 2ml cold acetone to precipitate proteins. Following vortex-mixing and centrifuging at 4 ◦C, 1860×g for 20 min, the supernatant was transferred to 12mm×100mm, GL 14 glass test tubes and 2ml dichloromethane was added. After vortex-mixing and centrifuging at 4 ◦C, 1860×g for 10min, 200μl of the upper layer was transferred to glass autosampler vials, followed by addition of 20μl 1M HCl and 15μl was injected onto the HPLC. The HPLC analysis was performed on an Agilent 1100 series LC (Santa Clara, CA),and Agilent ChemStations software were used for the chromatographic analysis. The separation was carried out on a ZORBAX SB-C18 Solv Saver Plus HPLC column (5 μm, 3.0 mm×150 mm).at a flow rate of 0.6 ml/min. Mobile phase A is 0.2% trifluoroacetic acid in Milli-Q water and mobile phase B is 0.2% trifluoroacetic acid in acetonitrile. The analytical method consists of an isocratic run with 92% mobile phase A for 23 min.. Each analytical run was followed by a 1.3 min washout gradient to 100% B. Column temperature was 25 ◦C, and autosampler tray temperature was 6 ◦C. Investigators quantified the analytes by using the analyte to standard peak area ratio on a Agilent 1100 High Performance Fluorescence detector G1321A and Agilent 1100 Series UV-Visible detectors G1314A. Detector settings were λex 260 nm/λem288nm for p-cresyl sulfate and λex 280 nm/λem 390nm for indoxyl sulfate, indole-3-acetic acid and internal standard. Hippuric acid was monitored by UV-Vis detector at 254 nm. Quantitative results are obtained and calculated in terms of their concentrations (mg/L). Fecal microbiota profile Sample Collection Qualified stool samples are self-collected in sterile frozen tubes by subjects and were transported immediately to the laboratory where they were stored at - 80 ◦C for further testing. DNA Extraction and Sequencing Genomic DNA from fecal samples is extracted using the QIAGEN DNA Stool Fast Kit (QIAGEN) according to the manufacturer's guidelines. The quality of the extracted metagenomic DNA is assessed through 1.0% agarose gel electrophoresis and spectrophotometric analysis (measuring the optical density at a 260/280 nm ratio). For the extracted DNA to be considered suitable, the concentration has to exceed 20 ng/μl, with a 260/280 nm ratio within the range of 1.8 to 2.0. The variable region 4 (V4) of the bacterial 16S rRNA gene is amplified through polymerase chain reaction (PCR) with bacterial/archaeal primers 515F/806R, including barcodes for sample identification. The resulting amplicons are purified using the GeneJET Gel Extraction Kit (Thermo Scientific) and quantified with a Qubit dsDNA HS Assay Kit on a Qubit 2.0 Fluorometer (Qubit). Sequencing libraries are prepared using the NEBNext® Ultra™ DNA Library Prep Kit for Illumina (NEB), following the manufacturer's protocol. The purified libraries are quantified, normalized, pooled, and then use for cluster generation and sequencing on an Illumina HiSeq 2500 platform to produce 250 bp paired-end reads. Paired-end reads are merged using FLASH v1.2.7, and quality filtering is performed using the QIIME 1.7 pipeline with Python scripts. Chimeric sequences were removed by UCHIME. The processed reads (effective tags) are clustered into operational taxonomic units (OTUs) at 97% sequence identity using UPARSE, and taxonomic classification is assigned based on the SILVA database. To assess the phylogenetic relationships of different OTUs, multiple sequence alignments are performed using PyNAST v1.2 against the SILVA database, and a phylogenetic tree was constructed using FastTree. Alpha diversity is estimated by species richness using the Chao1 index at the OTU level. A rarefaction curve is generated by randomly selecting a subset of sequencing data from each sample to represent the number of observed species, and a species accumulation curve is plotted to show the occurrence of new OTUs (species) with continuous sampling. For beta diversity, Bray-Curtis dissimilarities at the OTU level are calculated and analyzed using the vegan package. Principal coordinate analysis (PCoA) is conducted based on Bray-Curtis distances, and both weighted and unweighted UniFrac parameters are computed through the QIIME pipeline. Non-metric dimensional scaling (NMDS) is performed using the weighted correlation network analysis (WGCNA), stat, and ggplot2 packages in R by transforming a distance matrix of weighted and unweighted UniFrac parameters into a new set of orthogonal axes. All analyses are conducted using in-house R scripts, unless otherwise specified. Functional composition of metagenomes is predicted from 16S rRNA data using the PICRUSt software with Python scripts. A precomputed table of gene copy numbers for each gene family from sequenced bacterial and archaeal genomes, based on the IMG database, along with a phylogenetic tree from the Greengenes database, is used for gene content prediction.