Study population. We enrolled 66 children with diarrhea and 138 children with constipation (age < 3 years) from June 2017 to September 2021 at Beijing Dr. Cui Yutao Children's Health Management Center. Patients aged 0–3 years were included into the disease groups if they: (1) had bowel movement frequency of 3 times/week over the preceding 3 months; or (2) had an increased frequency of defecation (≥ 5 times/day with loose stool) accompanied by a change in stool consistency (i.e., significant increase in fecal water content), with a duration of diarrhea lasting ≥ 3 days. Patients were excluded based on any of the following: (1) had metabolic or neuropsychiatric conditions, cancer, congenital heart disease, liver and kidney dysfunction, or other severe diseases; (2) were taking antibiotics, probiotics, prebiotics, nonsteroidal anti-inflammatory drugs (NSAIDs), opioids, traditional Chinese medicine (TCM), proton pump inhibitors (PPIs), or histamine receptor antagonists during the preceding month before sample collection; and (3) were on a restricted diet, including low-fat diet and vegetarian diet. A total of 414 healthy children of the same age range who participated in health checkups during the same period were selected as healthy controls. An independent cohort comprised 390 healthy children, 191 children with constipation, and 73 children with diarrhea were recruited by the same criteria and used for fecal Ruminococcus qPCR validation. This study was performed with the approval of the Ethical Committees of Beijing Institute of Microbiology and Epidemiology, and written informed consent from the guardians of all the participants was obtained.
Sample collection and DNA extraction. Sterile fecal sampling tubes (SARSTEDT AG & Co. KG, Nümbrecht, Germany) were used to collect approximately 5 mL of feces per participants, minimizing the risk of bias. The samples were stored at − 80°C until genomic DNA was extracted using the TIANamp Stool DNA Kit (Tiangen Angen Biotech (Beijing) Co., Ltd., Beijing, China) according to the manufacturer's instructions. Before subjecting to sequencing, we assessed the quality and purity of DNA using a Qubit R3.0 Fluorometer (Thermo Fisher Scientific Inc., Waltham, Massachusetts, USA), with pure DNA having an OD260/OD280 ratio between 1.8 and 2.0 and all DNA concentrations being higher than 2.5 ng/µL.
16S rRNA gene sequencing. All of the recruited 618 participants underwent 16S rRNA sequencing, and one-step PCR was used to prepare the PCR Illumina sequencing libraries with the forward and reverse primers for the V3–V4 region (333 nmol each) and KAPA Hi-Fi PCR master mix (Kapa Biosystems, Boston, MA, USA). The forward and reverse primers used were 5’-CCTAYGGGRBGCASCAG-3’ and 5’-GGACTACNNGGGTATCTAAT-3’, respectively. The PCR conditions consisted of an enzyme activation step at 95°C for 3 minutes, followed by 20 cycles of 15 s at 98°C, 30 s at 50°C, 40 s at 72°C, and 10 min at 72°C, with a final hold at 10°C. The cDNA was purified using Clean Beads (Beckman Coulter Inc., Brea, California, USA) and sequenced on an Illumina HiSeq2500 platform (Illumina, Inc., San Diego, California, USA) to generate approximately 4.5 million reads of 16S rRNA V3–V4 amplicons, including the partial C3 region (341F, 17 base pair (bp)), full V3 region (57 bp), full V4 region (62 bp), and partial C5 region (806R, 20 bp).
Raw data filtering, classification, and annotation. Adaptors and PCR primers were eliminated from the reads, and paired-end reads were merged using FLASH version 1.2. Reads were truncated if they had three consecutive base calls with a quality score below 20, and only high-quality reads accounting for over 75% of the input read length (per single-end read) were included for further analysis. Chimeric reads were identified and removed with USEARCH version 6.1. The reads were clustered into OTUs with QIIME software version 1.9 at an identity threshold of 0.97 (32). OTUs with a count below four were excluded from the analysis. OTUs were annotated to their closest taxonomic neighbors using QIIME version 1.9 (33) based on the Greengenes database version 13.8 (34).
Analysis of diversity and microbiota differences. Four α-diversity indices including Chao1, Accumulated Cyclone Energy (ACE), Shannon, and Simpson were utilized to compare the microbial community richness and evenness among the diarrhea and constipation groups of children and healthy controls. The QIIME diversity alpha plugin was used to calculate alpha diversity measures for different groups. To compute beta diversity, unweighted Unifrac distance was used, which was defined based on the profiling table (35, 36). Visual methods (box plot) and statistical methods (Multiple Response Permutation Procedure (MRPP), Adonis, Amova) were employed to examine the significance of inter- and intra-group differences. A t test was used to detect biomarkers with significant differences, while Picrust (37) was used to annotate the metabolic pathways of the genera. Differential metabolic pathways between the disease groups and healthy controls were identified using the Wilcox rank-sum test. Finally, the correlation between genus and pathway was analyzed using Spearman correlation coefficient.
Quantification of Ruminococcus by qPCR. The independent cohort of 654 participants underwent qPCR test to validate the absolute quantification of Ruminococcus levels. qPCR was conducted using Direct Detect Seven Genus/Species Gut Microbes Detection Kit (PCR-Fluorescence Probe, Coyote Bioscience Inc., China). The PCR program was as follows: 10 cycles at 50°C for 5 s, 95°C for 5 s, and 40 cycles at 95°C for 50 s and 60°C for 30 s. The PCR amplification was performed on an ABI 7500 Real-Time PCR System (Applied Biosystems, Foster City, USA). A PCR standard curve were then used for absolute quantification for Ruminococcus in the validation groups, including 390 healthy children, 191 children with constipation, and 73 children with diarrhea.