Elevated levels of circulating DNA in cardiovascular disease patients: metagenomic profiling of microbiome in the circulation

Abstract

Cardiovascular diseases (CVDs) are the leading cause of death worldwide. An expanding body of evidence supports the role of human microbiome in the establishment of CVDs and, this has gained much attention recently. This work was aimed to study the circulating human microbiome in CVD patients and healthy subjects. The levels of circulating cell free DNA (circDNA) was higher in CVD patients (n = 80) than in healthy controls (n = 40). More specifically, the relative levels of circulating bacterial DNA and the ratio of 16S rRNA/β-globin gene copy numbers were higher in the circulation of CVD patients than healthy individuals. In addition, we found a higher circulating microbial diversity in CVD patients (n = 3) in comparison to healthy individuals (n = 3) by deep shotgun sequencing. At the phylum level, we observed a dominance of Actinobacteria in CVD patients, followed by Proteobacteria, in contrast to that in healthy controls, where Proteobacteria was predominantly enriched, followed by Actinobacteria. The circulating virome in CVD patients was enriched with bacteriophages with a preponderance of Propionibacterium phages, followed by Pseudomonas phages and Rhizobium phages in contrast to that in healthy individuals, where a relatively greater abundance of eukaryotic viruses dominated by Lymphocystis virus (LCV) and Torque Teno viruses (TTV) was observed. Thus, the release of bacterial and viral DNA elements in the circulation could play a major role leading to elevated circDNA levels in CVD patients. The increased circDNA levels could be either the cause or consequence of CVD incidence, which needs to be explored further.

Figure 1. Levels of circDNA in the plasma of CVD patients and healthy control samples (in Log10 scale).

Figure 2. Number of copies of 16S rRNA gene (A), number of copies of β-globin gene (B) and the ratio of number of copies of 16S rRNA to β-globin (C) in the plasma of CVD patients and healthy control samples (in Log10 Scale).

Figure 3. Receiver operating characteristic (ROC) curve for plasma 16S rRNA gene levels, β-globin gene levels and circDNA levels of (A) Valvular Heart disease (VHD), (B) Ischemic Heart disease (IHD) and (C) Congenital Heart disease (CHD) patients.

Figure 4. Relative abundance of bacterial, human, viral and archeal signatures in human plasma as profiled by metagenome shotgun sequencing.

Figure 5. Comparison of bacterial compositional summary of control and CVD samples by deep shotgun sequencing (A) and amplicon sequencing (earlier study) (B).

(A) Major compositional differences include (i) higher frequency of phylum actinobacteria than phylum proteobacteria among CVD samples (ii) higher frequency of Proteobacteria than Actinobacteria among healthy control samples (iii) negligible presence/complete absence of the Phylum actinobacteria in the control sample CON029. (B) Relative distribution of bacterial phyla obseveved by amplicon sequencing. Proteobacteria and Firmicutes were dominant in control samples and only Proteobacteria was dominant among CVD samples .

Figure 6. The ratio of Actinobacteria: Proteobacteria in CVD patients and healthy controls.

(A) Relative occurrence of Actinobacteria and Proteobacteria in the circulating microbiome of CVD patients and healthy controls. An increase in Actinobacterial population coupled with a parallel reduction in Proteobacterial population in CVD patients is discernible. (B) Increase in the ratio of Actinobacteria / Proteobacteria (p = 0.035) in CVD patients compared to controls. (C) Increase in the ratio of Proteobacteria / Actinobacteria (p = 0.028) in controls compared to CVD patients.

Figure 7. OTU heatmap displaying the distribution of OTUs from all 6 samples.

The heatmap highlighting the distribution of CVD and healthy sample specific taxonomic lineage are indicated. Phyla Proteobacteria and Actinobacteria are dominant both in test and control samples. Phylum Actinobacteria is dominant among test samples while Phylum Proteobacteria is dominant among control samples.

Figure 8. Comparison of the viral compositional summary of control and CVD samples by deep shotgun sequencing.

(A) Major compositional differences include (i) relatively higher abundance of families Siphoviridae, Myoviridae, Podoviridae and Inoviridae among CVD samples than healthy individuals (ii) relatively higher abundance of families Mimiviridae, Poxviridae, Phycodnaviridae, Iridoviridae, Anelloviridae and Geminiviridae among healthy control samples (iii) negligible presence of the family Retroviridae in all the samples. (B) Relative occurrence of Bacteriophages and Eukaryotic viruses in the circulating microbiome of CVD patients and healthy controls. An increase in the population of Bacteriophages coupled with a parallel reduction in the population of Eukaryotic viruses in CVD patients is discernible.