Friday, February 23, 2024
BestWooCommerceThemeBuilttoBoostSales-728x90

Gut microbiome signature of metabolically healthy obese individuals according to anthropometric, metabolic and inflammatory parameters – Scientific Reports


Baseline characteristics of healthy obese subjects

The clinical characteristics and blood samples of the 120 study subjects were collected for analysis (Table 1). A total of 21 men and 99 women participated in this study. The mean age of the participants was 44 years old, and 7 out of the 120 subjects were smokers. The mean body mass index (BMI), and the median waist circumference (WC), and waist-to-hip ratio (WHR) were measured to be 27.8 kg/m2, 95.1 cm and 0.92, respectively.

Table 1 Baseline characteristics (N = 120).

The mean serum triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) levels were 123.7 mg/dl, 118.5 mg/dl, and 55.7 mg/dl, respectively. The median apolipoprotein-A1, and apolipoprotein-B levels were measured to be 142.5 mg/dl and 99 mg/dl, respectively. Body fat composition was measured and calculated using CT and dual energy X-ray absorptiometry (DEXA). The visceral fat area, subcutaneous fat area, and the visceral to total body fat percentage were calculated to be 117.0 cm2, 275.9 cm2, and 30.15% on the median, respectively. Mean body fat mass measured by DEXA was 28.7 kg and mean body fat percentage was 40.6%.

We analysed the correlation between different phenotypic, inflammatory parameters (Supp. Figure 1). The BMI level shows a positive correlation with visceral fat area, subcutaneous fat area, waist circumference, and waist-to-hip ratio. The BMI also positively correlates with levels of IL-1β, IL-6, TG, and TG/HDL-C ratio. In contrast, HDL-C, and Apoprotein-A1 levels inversely correlates with apolipoprotein-B/apolipoprotein-A1 ratio, TG/HDL-C ratio, and BMI level (all p < 0.05 by Spearman’s rank correlation test).

EV-derived microbiota composition

The faecal microbiome originating from bacterial cells and extracellular vesicles (EV) was profiled using 16S rDNA sequencing. There were no significant findings when analysing the microbiome originating from bacterial cells (data not shown). In contrast, when the correlation between faecal bacterial EV-derived microbiome composition and host phenotype was analysed, there were some significant correlations as the following (Fig. 1A–C). Clinical phenotypes, including serum level of IL-1β and resistin showed, significant correlation with the abundance of different faecal EV-derived microbiota of different species. A list of significant correlations between taxa and clinical variables with p < 0.05 is provided in Supplementary Table 1. Some of them are as follows: on the phylum level, the abundance of Firmicutes-derived EVs showed positive correlations with visceral fat area, serum apolipoprotein-B/apolipoprotein-A1 ratio, apolipoprotein-B, serum LDL-C level and serum IL-1β level (Spearman’s rank correlation coefficient ρ = 0.18, 0.19, 0.22, 0.24 and 0.24 respectively, all p < 0.05). On the genus level, the abundance of Akkermansia-derived EVs showed negative correlations with BMI and subcutaneous fat area (ρ = − 0.19, − 0.23 respectively, all p < 0.05). The abundance of Akkermansia-derived EVs also showed a negative correlation with serum IL-1β, leptin, fasting insulin, HOMA-IR, and resistin level (ρ = − 0.44, − 0.21, − 0.23, − 0.24 and − 0.32 respectively, all p < 0.05). The abundance of Bacteroides-derived EVs also showed weak positive correlations with serum leptin level (ρ = − 0.23, p = 0.01). The abundance of Prevotella-derived EVs also showed weak positive correlations with serum IL-1β level (ρ = 0.19, p = 0.04). Among them, however, only negative correlations between the abundance of EV-derived Akkermansia, and serum resistin and IL-1β levels remained significant after FDR adjustment (FDR q < 0.01).

Figure 1

Correlation between clinical/laboratory parameters and microbial abundances. Correlation between different clinical parameters were analyzed by Spearman’s rank correlation analysis. Statistically significant correlation with FDR q value < 0.05 are indicated by asterisks. (A) Correlation in phylum level, (B) Correlation in family level. (C) Correlation in genus level. Dendrograms on X, Y axis were generated using complete-linkage hierarchical clustering. WC waist circumference, BMI body mass index, IL-1B interleukin-1β, WHR waist-to-hip ratio, dBP diastolic blood pressure, sBP systolic blood pressure, ALT alanine aminotransferase, AST aspartate aminotransferase, TG triglyceride, HDL-C high-density lipoprotein, LDL-C low-density lipoprotein, IL-6 interleukin-6.

Gut microbe-derived extracellular vesicles in different enterotypes

The overall study population could be classified into two distinct enterotypes based on their stool EV-derived microbiome profile (enterotype I: Prevotellaceae-predominant, enterotype II: Akkermansia/Bacteroides-predominant, Fig. 2). The Calinski-Harabasz (CH) index was used to calculate the optimal number of clusters (Fig. 2A). Out of a total of 120 subjects, 34 were classified as enterotype I, and 86 were classified as enterotype II. In contrast, the bacterial cell-derived microbial compositions failed to separate the study population into distinct subgroups of patients (Supp. Figure 2A–C).

Figure 2
figure 2

Enterotyping of the study subjects by extracellular vesicle-derived microbial compositions. (A) Calinski-Harabasz (CH) index. (B) Principal coordination analysis (PCoA) plot showing enterotype distribution of the obese population by Jensen–Shannon Divergence distance. The distance of one grid corresponds to 0.2 in Jensen–Shannon Divergence distance (d = 0.2).

We compared the species richness and evenness of the bacterial EV-derived microbiome between the two enterotype groups (Fig. 3A). The microbial diversity calculated by the Shannon index and Faith’s phylogenetic diversity were both significantly lower in enterotype I than in enterotype II (p < 0.001 and p = 0.003, respectively). The microbial composition, analysed by unweighted and weighted UniFrac distance, is depicted in Fig. 3B,C. Statistical analysis revealed a distinct distribution between the two enterotypes (PERMANOVA, all p = 0.001).

Figure 3
figure 3

Extracellular vesicle-derived microbial diversity (A), β-diversity (principal coordinates analysis plots: B,C), and relative abundances (D) according to the enterotypes. (A) Shannon index and Faith’s phylogenetic diversity (Wilcoxon rank-sum test, p < 0.001, p = 0.003 for Shannon index and Faith’s phylogenetic diversity). Asterisks are added for p value < 0.05 (ns: p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). (B) Principal coordinates analysis plots showing EV-derived bacterial distribution in two enterotypes by unweighted UniFrac distance. (PERMANOVA p = 0.463 by sex, p = 0.001 by enterotype) (C) Principal coordinates analysis plots showing EV-derived bacterial distribution in two enterotypes by weighted UniFrac distance. (PERMANOVA p = 0.022 by sex, p = 0.001 by enterotype) (D) Relative abundance of EV-derived microbiome (Phylum, Family and Genus levels).

We analysed the relative abundance of gut microbe-derived EVs at the phylum, family, and genus levels (Fig. 3D, Supp. Table 2). Enterotype I subject showed significant enrichment of Bacteroidetes-derived EVs, and depletion of Actinobacteria, Firmicutes and Proteobacteria-derived EVs in phylum level (all FDR q < 0.05). Among the phylum Bacteroidetes, subjects in enterotype I showed a higher abundance of Prevotellaceae-derived EVs at the family level and Prevotella-derived EVs at the genus level (all FDR q < 0.05). At the family level, subject in enterotype II had a higher abundance of Lachnospiraceae and Ruminococcaceae-derived EVs (all FDR q < 0.05). At the genus level, subjects in enterotype II a had significantly higher abundance of Akkermansia-derived EVs (FDR q < 0.05).

Enterotype and host metabolic and inflammatory markers

The metabolic and inflammatory markers as well as the body fat compositions according to the enterotypes are summarised in Table 2. The enterotypes were independent of patient age and sex. Subjects in enterotype I tended to have higher levels of BMI, which did not reach statistical significance (p = 0.060).

Table 2 Characteristics of the subjects according to enterotypes.

Although no significant difference was seen, serum resistin tended to be higher in subjects in enterotype I compared to subjects in enterotype II (nominal p = 0.096). Serum IL-1β levels were higher in subjects in enterotype I than in those in enterotype II (nominal p = 0.025 and FDR q = 0.050). In contrast, there was no significant difference in serum IL-6 levels between the two enterotype groups (p = 0.622).

The total body fat mass also tended to be higher in the enterotype I group than in enterotype II group (nominal p = 0.068). Both visceral fat area and subcutaneous fat area did not differ between the two enterotypes. There was no difference in the dietary intake of total calories, carbohydrates, lipids, proteins, fibers, or total cholesterols.

Further analysis on the clinical variables were performed according to sex (Supp. Table 3). The distribution of overall anthropometric measurements, metabolic parameters, and inflammatory parameters by enterotype showed generally similar trends in both men and women. In men, however, BMI and waist circumference were significantly lower in enterotype II (nominal p = 0.002 and 0.028, respectively, Supp. Table 3), and HDL-C level were significantly higher in enterotype II (nominal p = 0.024, Supp. Table 3). In women, BMI and waist circumference failed to show statistically significant difference (p > 0.05, Supp. Table 3).

The distribution of some phenotypic and inflammatory markers in faecal EV-derived microbiome is visually depicted in Fig. 4. The microbiome profile of the study participants appears to be largely divided into two clusters. Interestingly, the distribution of faecal EV-derived microbiome profiles according to serum IL-1β levels appeared to be markedly different in the two clusters (Fig. 4D). The distribution of faecal EV-derived microbiome profiles did not differ according to the BMI (Fig. 4A), waist circumference (Fig. 4B) or serum IL-6 levels in the two enterotypes (Fig. 4C).

Figure 4
figure 4

EV-derived microbiome profile and distribution of clinical parameters and inflammatory markers in MHO subjects. (A) Weighted UniFrac distance matrix showing distribution of body mass index. (B) Weighted UniFrac distance matrix showing distribution of waist circumference. (C) Weighted UniFrac distance matrix showing distribution of interleukin-6. (D) Weighted UniFrac distance matrix showing distribution of interleukin-1β.

In the case of the bacterial cell-derived microbiome, there were no significant findings based on anthropometric and inflammatory biomarkers of obesity (Supp. Figure 3A–D).



Source link

Related Articles

Leave a Reply

[td_block_social_counter facebook="beingmedicos1" twitter="being_medicos" youtube="beingmedicosgroup" style="style8 td-social-boxed td-social-font-icons" tdc_css="eyJhbGwiOnsibWFyZ2luLWJvdHRvbSI6IjM4IiwiZGlzcGxheSI6IiJ9LCJwb3J0cmFpdCI6eyJtYXJnaW4tYm90dG9tIjoiMzAiLCJkaXNwbGF5IjoiIn0sInBvcnRyYWl0X21heF93aWR0aCI6MTAxOCwicG9ydHJhaXRfbWluX3dpZHRoIjo3Njh9" custom_title="Stay Connected" block_template_id="td_block_template_8" f_header_font_family="712" f_header_font_transform="uppercase" f_header_font_weight="500" f_header_font_size="17" border_color="#dd3333"]
- Advertisement -spot_img

Latest Articles