The research scheme was approved by the Ethics Review Committee of Peking University People’s Hospital of China (Approval No:2019PHB110-01). All participants and their parents received oral and written information about the study and written consent was obtained before recruitment. All experiments were conducted in accordance with approved guidelines and regulations (Ren et al., 2022).

As described in our previous study (Ren et al., 2022), 15 children with DS were randomly recruited from integrated boarding schools in Beijing and its surrounding areas, and 15 age- and diet structure-matched children were recruited as the healthy group. Fifteen pairs of samples were collected for amplicon sequencing analysis. There were no differences in age and gender between DS group and healthy group. All the study participants had the same diet offered by school, which greatly reduces the impacts of diet, lifestyle and geography on gut microbiota profile. Besides, we excluded children who had received antibiotics within 3 months. Each participant’s height and weight were measured, and body mass index (BMI) was calculated by dividing weight by the square of height. The DS group had a lower average height (P = 0.037), a higher BMI (P = 0.009) than the healthy group. There was no difference in the weight between the two groups. Due to the limitation of the Chinese version of the Wechsler Intelligence Scale for Children (WISC-IV), only eleven pairs of participants over the age of 6 were subjected to the cognition test.

To assess the participants’ cognitive function, we calculated the Full Scale IQ (FSIQ) test and measured the other four factor indices: the perceptual reasoning index (PRI), which was used to examine nonverbal fluid reasoning skills; the verbal comprehension index (VCI), which was used to measure verbal acquired knowledge and verbal reasoning; the working memory index (WMI), which was used to measure a person’s ability to manipulate linguistic information in short-term instant memory and then formulate responses; and the processing speed index (PSI), which was used to evaluate children’s ability to efficiently scan and understand visual information and use information to complete the work. Cognitive tests were conducted by two well-trained pediatricians in a hospital conference room with no interference. As described in our previous study (Ren et al., 2022), the scores of FSIQ, PRI, VCI, WMI and PSI were all decreased in the DS group (all P < 0.001). The correlation coefficient of this type is approximately 0.857.

Fecal bacterial DNA storage tubes (Longsee Biological Company, Guangzhou, China) were used to collect fecal samples. Before processing, samples are transferred and stored at − 80°C. The average moving time between sample collection and storage is 19 minutes at -80°C, and the maximum is 23 minutes. Total fecal DNA was extracted using a QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany). The concentration of genomic DNA in each fecal sample was quantified using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, MA, USA). DNA integrity and sizes were assessed by 1% agarose gel electrophoresis. The DNA was resuspended in H₂O and stored at −80°C prior to use.

The concentration detection results of the collected and extracted genomic DNA were all in line with the standards ( Table S1 ). The purity, fragment size, and concentration of the amplified products of all samples met the sequencing requirements ( Figure S1 ).

Quantitative PCR (qPCR) amplifications were performed in a 20-μL reaction solution containing SYBR^® Green Real-time PCR Master Mix (Toyobo, Code No. QPK-201, Osaka, Japan). The specific primers for Blautia sp. covers all species of Blautia ( Figure S2 ). The cycling conditions were as follows: 95°C for 1 min and then 33 cycles of 95°C for 15 s, 56°C for 15 s, and 72°C for 45 s.

The relative abundance of Blautia was evaluated using the following formulae. First, the relative value (RV) of Blautia was determined by the 2^−ΔΔct method with normalization to the total bacterial level. Then, the relative abundance (RA) of Blautia was scaled to a reasonable range with the following equation: RA = log₂ (RV × 10000 + 1).

DNA extraction was performed for Blautia-specific PCR amplicon sequencing. Standard protocols were used to generate amplicons using a genus-specific primer from Blautia. Briefly, the hypervariable region of the 16S rRNA gene was amplified using a primer pair (477F: 5′-CGGTACCTGACTAAGAAGC-3′ and 719R: 5′-GTTCCTCCTAATATCTACGC-3′) with barcodes (Luu et al., 2017). All PCRs were performed with Phusion^® High-Fidelity PCR Master Mix (New England Biolabs). After the purification of a Blautia-specific PCR amplicon, the libraries were generated with an Ion Plus Fragment Library kit, and paired-end sequencing (2 × 125 bp) was performed with an IonS5™XL platform (Thermo Fisher) from Novogene (Beijing, China).

These reads were analyzed with QIIME ver. 2.2020.6 (Quantitative Insights into Microbial Ecology) software package. Single-end FASTQ reads were imported into QIIME using a q2 manifest file import method and quality filtered using the q2-dada2 denoising method. The frequencies of amplicon sequence variants (ASVs) were identified at the 100% similarity level. The representative sequences of ASVs were aligned with the Silva database. The taxonomic composition of each sample at the species level was calculated.

Gas chromatography−mass spectrometry (GC−MS) analysis was conducted using a SHIMADZU GC2030-QP2020 NX GC−MS system with an HP-FFAP capillary column. A 1-μL aliquot of the analyte was injected in split mode (5:1). Helium was used as the carrier gas, the front inlet purge flow was 3 mL min⁻¹, and the gas flow rate through the column was 1 mL min⁻¹. The initial temperature was kept at 80°C for 1 min, then raised to 200°C at a rate of 10°C min⁻¹ for 5 min, and then kept for 1 min at 240°C at a rate of 40°C min⁻¹. The injection, transfer line, quad, and ion source temperatures were 240°C, 240°C, 200°C, and 150°C, respectively. The energy was −70 eV in electron impact mode. The mass spectrometry data were acquired in Scan mode with a m/z range of 33–150 after a solvent delay of 3.5 min.

The core metric phylogenetic method in the q2 diversity plug-in of QIME was used for diversity analysis. To estimate α-diversity (within sample), different metrics were calculated: Shannon (a quantitative measure of the number of ASVs present in a given sample and their relative abundance or community richness) and Simpson metrics. Alpha diversity indices were compared using the Mann–Whitney test. Principal coordinate analysis (PCoA) and Principal component analysis (PCA) based on the weighted UniFrac measures was used to visualize beta diversity relationships.

The microbial comparisons between the DS and healthy groups were performed using the Mann–Whitney test. Associations between clinical indices and Blautia were evaluated by the MaAsLin analysis (Stewart et al., 2018). Statistical analysis of the clinical data was performed using SPSS (Statistical Package for Social Sciences) 22.0 software (SPSS Inc., Chicago, IL, USA). P < 0.05 was considered statistically significant.

A total of 2,400,845 high-quality combined sequence reads were obtained after data processing (average, 80,028.2 reads). All sequences were clustered into 150 representative ASVs, and 10 of the 150 ASVs were identified as Blautia taxa. The representative sequence of each ASV was assigned to the BLASTN dataset to estimate the diversity of Blautia species in the DS and healthy groups. Six published Blautia species with the best hits are shown in Table 1 , namely, B. massiliensis, B. wexlerae, B. faecis, B. stercoris, B. obeum, and B. argi. Of these, ASV1, assigned to B. massiliensis, was the dominant species, representing 49.75% of Blautia sequences. In addition, the proportions of B. wexlerae, B. faecis, B. stercoris, B. obeum, and B. argi (ASV34 and ASV53) in the Blautia sequences were 41.85%, 2.98%, 2.54%, 2.10%, and 0.62%, respectively (ASV34 = 0.54%, ASV53 = 0.08%).

These findings revealed the detailed phylogenetic inventory of the Blautia bacterial group in the human gut microbiota and showed that Blautia is prevalent in fecal samples from DS and healthy children.

Microbial alpha diversity was assessed by means of the Shannon and Simpson indices. The Shannon index was significantly increased in the DS group ( Figure 1A ) compared with the healthy group (4.104 vs. 3.984, P = 0.004), and there was no significant change of Simpson diversity index between the two groups (0.908 vs. 0.904, P = 0.1192) ( Figure S3A ), suggesting greater Blautia commensal diversity in the DS group. To further assess the alteration in the Blautia composition in the DS group, we used PCA to examine the observed discrete clustering of the intestinal microbiome in the DS and healthy groups. The results showed that the characteristics of the Blautia genus in DS patients were significantly different from those in healthy individuals ( Figure S3B ). Significant differences were also found by PCoA analysis, indicating that the Blautia genus structure differed significantly between the DS and healthy groups ( Figure 1B ). In addition, 7 of the 10 total ASVs were shared among all the samples. As shown in the Venn diagram, ASV143 was exclusively present in the DS group, while ASV53 (Blautia argi) and ASV89 were exclusively present in the healthy group ( Figure 1C ).

As a comparison of the group differences at the Blautia genus level, the quantitative PCR results showed that the relative abundance (RA) of Blautia was significantly decreased in the intestinal tract of children with DS (P < 0.05, Figure 2A ). This result was consistent with our previous findings (Ren et al., 2022), which showed a significantly lower level of Blautia in children with DS. Taken together, these results indicate that the Blautia genus is disturbed in individuals with DS.

Next, we explored the RA of Blautia at the species level in DS and healthy children through Blautia-specific amplicon sequencing. The abundances of B. wexlerae and B. obeum were not significantly different between the two groups ( Figures 2C, F ).

Three species of Blautia were significantly decreased in children with DS: B. massiliensis (ASV1, P < 0.0001, Figure 2B ), B. stercoris (ASV21, P = 0.0054, Figure 2E ), and B. argi (ASV34, P = 0.0001, Figure 2G ; ASV53, P = 0.0039, Figure 2H ). Notably, B. argi (ASV53, P = 0.0039, Figure 2H ) was nearly absent in children with DS. While, B. faecis significantly increased in children with DS (ASV17, P < 0.0001). While, B. faecis significantly increased in children with DS (ASV17, P < 0.0001, Figure 2D ).

To explore the factors contributing to the Blautia variation in the gut microbiome of DS and healthy children, canonical correspondence analysis (CCA) was performed. Some clinical parameters were identified as important contributors for differentiating DS samples from healthy samples ( Figure 3A , Table S2 ). Among them cognitive indices had significant effects on healthy subject clustering, indicating that these clinical parameters are highly associated with the Blautia community variation between DS and healthy children. WMI had the largest effect on the healthy sample variation (R² = 0.798105, P = 0.005).

Furthermore, we examined correlations between the WISC-IV scores and the species level of Blautia. Significant negative correlations between B. faecis and all WISC-IV scores were observed by MaAsLin analysis ( Figures 3B–F ). B. argi (ASV53) had a significant positive association with all WISC-IV scores ( Figures 3G–K ), which is consistent with the findings of the CCA. While, B. argi (ASV34) was significant positively related to the score of WMI ( Figure 3L ). B. argi (ASV53) was indicated as an important contributor for differentiating healthy samples from DS samples. In addition, B. massiliensis had a significant positive association with BMI ( Figure 3M ). These results indicate that B. argi might be beneficial for cognitive function, while B. faecis might be a risk factor for cognitive function.

SCFAs, particularly acetic acid, propionate, and butyrate, are the end products of the intestinal microbial fermentation of dietary fibers and resistant starch. The levels of SCFAs were determined by GC−MS analysis ( Table S3 ), and the average level of acetic acid was significantly lower in the feces of the DS group than in that of the healthy samples (P < 0.05, Figure 4A ). Notably, Blautia has been reported to be an acetic acid-producing bacterium (Aoki et al., 2021), and B. massiliensis (ASV1) and acetic acid contents had a significant positive correlation (P = 0.0014, Figure 4B ), indicating its acetic acid-producing property. The relative abundance of Blautia species showed a positive correlation with a high level of acetic acid, which is consistent with previous studies (Fung et al., 2017).

To further investigate the relationship between clinical indices and acetic acid, we used MaAsLin analysis to measure the correlation between WISC-IV scores, BMI and the level of acetic acid content. The findings indicated that acetic acid and B. massiliensis was negatively correlated with BMI ( Figures 3M and 4C ), suggesting that acetic acid plays an important role in over-weight of DS children. However, our findings indicated no significant differences between the abundance of acetic acids and the WISC-IV scores.

The importance of the association between the proportion of Blautia species and a high level of acetic acid is therefore supported by the acetic acid-producing effect, implying a beneficial impact of B. massiliensis as an acetic acid producing species on cognitive (Wen et al., 2020) and metabolic function in Chinese children by the metabolites.

Tax4Fun, based on the closed-reference ASVs, was used to predict the functional categories of the KEGG Orthology (KO). The genomes of the Blautia-specific groups in DS children and healthy children were predicted, in conjunction with their physiological, cellular, and molecular functions ( Figure 5 ). The genes of Blautia species in DS children were mainly enriched in molecular pathways, that is, pyruvate metabolism and glutamate metabolism. In contrast, the genes were deficient in the DS group in molecular pathways such as starch and sucrose and glycolysis metabolism. There were significant differences in the enrichment of the molecular pathways between the two groups of genes ( Figure S4 ). Thus, functional genomics prediction suggested that Blautia may affect cognitive function by altering metabolites.