Study population. Comparison of general data between the HC and ACS group. There were no statistically significant differences in age, sex, previous history of basic diabetes between the HC and the ACS groups (P>0.05), and the data were comparable (Table 1). A total of 66 consecutive cardiology patients were enrolled according to the exclusion criteria, the remaining 66 were divided in three groups for analysis: an ACS group, including 31AMI patients, 29 UAP patients and 6AS patients; and a control group comprising 46 healthy volunteers, no evidence of enrollment bias was found.
Characteristics of study participants. The baseline characteristics of UAP patients, AMI patients, AS patients and healthy controls were shown in Table 1. Compared to healthy controls, parameters in ACS including body weights (BWs), low-density lipoprotein (LDL)(Fig1), triglyceride (TG), total cholesterol (TC), C-reactive protein (CRP) and high homocysteine (HCY) were elevated; whereas high-density lipoprotein (HDL) were decreased.
Gut microbial profile in patients with ACS and HC
diversity index analysis
Shown in Figure2.A and B, with the increase of sample sequencing depth, the observed species index curve tended to be flat, indicating that the current sequencing depth is sufficient to detect predominant species contained in each sample. The abundance was reflected by the length of the curve on the horizontal axis and the evenness was reflected by the shape of the curve. After analyzing the rank-abundance curve about OTU of the samples, we found a smooth curve, indicting high evenness among samples. The gut microbiota of HCs and patients with ACS showed consistent values in the analysis of observed species and Shannon diversity index. Figure2C represented species diversity and indicated significant differences in the number of species among AS-HC (p=0.0000), HC-UAP (p=0.0000) and AMI-HC (p=0.0000). There was no significant difference in the number of species among AMI-AS (p=0.7204) AMI-UAP (p=0.792) and AS-UAP (p=0.8776). It was suggested that there were significant differences between the disease group (AMI, UAP, AS) and the healthy group. There was no difference in comparative analysis between disease groups. Figure.2D representing the Shannon index found the significant differences in HC-UAP (p=0.0000), AMI-HC (p=0.0001) and AS-HC (p=0.0291). There was no significant difference in Shannon index between AS-UAP (p=0.5457), AMI-AS (p=0.7100) and AMI-UAP (p=0.7232). Principle Coordinate Analysis (PCoA) (weighted UniFrac distance) between the ACS and control groups showed distinguished in total (Fig. 2C), top 40 abundant stool microbial taxa, and overlapping clustering (Fig.2D) was observed in AMI, UAP and AS groups.
β diversity analysis
Based on unweighted UniFrac distance and Bray-Curtis distance matrices of the 16S rRNA sequence, samples contribution rates of the first pCoA(PC1), second pCoA(PC2) were 21.45% and 9.25%, respectively, which highlighted a clear clustering of the microbial populations of the ACS patients away from that of the healthy controls.(A) NMDS(B) showed that the ACS and control groups were distinguished by total (Fig. 3A) or top 40 (Fig. 3B) abundant stool microbial taxa (stress=0.16, Stress<0.2), and overlapping clustering was observed in AMI, UAP and AS groups.
OUT analysis
A Venn diagram (Fig.4) showed that 789 OTUs were commonly detected between HC and AMI, common OTUs in the two groups of HC and UAP were 788, 632 OTUs were generally recognized between HC and AS, in addition, 596 OTUs were commonly detected in the four groups, while 450, 91, 144 and 20 OTUs were unique in the HC, AMI, UAP, or AS, respectively.
The ACS and HC samples were significantly different in multivariate analyses. At the phylum level (Fig. 5), compared to healthy controls, the dominant stool microbes in the ACS group were Firmicutes, Bacteroidetes, Proteobacteria. The relative abundances of Bacteroidetes, Proteobacteria, and Verrucomicrobia were increased, while Firmicutes was decreased in the ACS group compared to the HC group(p<0.05). At the genus level (Fig. 6), compared to HC group, the levels of Bacteroides, Parabacteroide, Unidentified Enterobacteriaceae, Subdoligranulum, Akkermansia, Alistipes, Streptococcus, Paraprevotella and Paraprevotella were significantly increased, whereas Subdoligranulum, Roseburia, Faecalibacterium, Blautia, Agathobacter, Bifidobacterium and Anaerostipes were significantly reduced.
Further analysis of the vegetation of the AMI, UAP, AS and HC healthy groups using the network diagram showed that there are significant differences in four groups (Fig.7). Compared with the entire HC group, the dominant ACS bacteria (AMI, UAP) in Bacteroides, Parabacteroide, Unidentified_Enterobacteriaceae and Subdoligranulum were increased significantly. Akkermansia, Anaerostipes Blautia, Agathbacter were decreased significantly. The abundances of Alistipes, Streptococcus and Paraprevotella were also higher in ACS than those in control samples.
Analysis of inflammatory factors
Compared with the HC healthy group (Fig. 8), the inflammatory factors including TNF-α (P=0.049), MCP-1 (P =0.046), IL-6 (P =0.047), IL-1β (P =0.035), IL-10 (P =0.049) and LPS (P =0.0001) in ACS group (AMI group and UAP group) were significantly increased, suggesting that there were significant changes in inflammatory factors in ACS.
Correlations of changes in clinical indexes, inflammatory factors with alterations in intestinal flora
According to the correlation between clinical acute coronary syndrome and clinical correlation index and flora inflammation, we further analyzed the correlation between clinical related indexes such as age (Age), body mass index (BMI), alanine aminotransferase (ALT), hyperhomocysteine (HCY) and differential bacteria (Fig. 9A). The analysis of the relationship between age and differential bacteria found that age was negatively correlated with Firmicutes, but positively correlated with Bacteroidetes, Proteobacteria and Verrucomicrobia, suggesting that under disease conditions, the increase of age had obvious influence on the disorder of microflora. BMI was negatively correlated with Firmicutes and positively correlated with Bacteroidetes. In the disease condition, changes in body weight affected microbial changes. ALT was negatively correlated with Firmicutes and positively correlated with Bacteroidetes, suggesting that liver function damage was closely related to the bacterial community. HCY was positively correlated with Proteobacteria and Bacteroidetes, and negatively correlated with Firmicutes. There was a correlation between HDL-C and different bacteria. HDL-C is positively correlated with Firmicutes, negatively correlated with Bacteroidetes, and positively correlated with F/B. There was no statistical difference in the correlation analysis between LDL-C and bacteria. It may be caused by the small number of cases, drug intervention and other related interfering factors. TG was negatively correlated with Firmicutes and positively correlated with Bacteroidetes. It was suggested that age, weight, liver function damage and HCY sepsis in AMI and UAP patients can affect the distribution of microflora.
The relative abundance of Firmicutes was positively correlated with plasma inflammatory factors IL-1β, TNF-α, MCP-1 and LPS (Fig. 9B). The abundance of Bacteroides was negatively correlated with IL-1β, TNF-α, MCP-1 and LPS; Akkermansia was positively correlated with IL-10, and there was no significant difference in the correlation between IL-6 and Firmicutes and Streptococcus, suggesting that intestinal flora and inflammatory indicators interfered with each other and were closely related in ACS.