Profiling system-wide variations and similarities between Rheumatic Heart Disease and Acute Rheumatic Fever-A pilot analysis

Abstract

Rheumatic heart disease (RHD) continues to affect developing countries with low income due to the lack of resources and effective diagnostic techniques. Understanding the genetic basis common to both the diseases and that of progression from its prequel disease state, Acute Rheumatic Fever (ARF), would aid in developing predictive biomarkers and improving patient care. To gain system-wide molecular insights into possible causes for progression, in this pilot study, we collected blood transcriptomes from ARF (5) and RHD (5) patients. Using an integrated transcriptome and network analysis approach, we identified a subnetwork comprising the most significantly differentially expressed genes and most perturbed pathways in RHD compared to ARF. For example, the chemokine signaling pathway was seen to be upregulated, while tryptophan metabolism was found to be downregulated in RHD. The subnetworks of variation between the two conditions provide unbiased molecular-level insights into the host processes that may be linked with the progression of ARF to RHD, which has the potential to inform future diagnostics and therapeutic strategies. We also found a significantly raised neutrophil/lymphocyte ratio in both ARF and RHD cohorts. Activated neutrophils and inhibited Natural Killer cell gene signatures reflected the drivers of the inflammatory process typical to both disease conditions.

Fig 1. Overview of the study.

(A) Relationship between the groups of analysis. (B) Overview of the workflow and results. Functionally significant DEGs characteristic to the RHD vs. ARF group obtained from a sensitive network mining approach were identified by excluding DEGs from sub-networks of RHD vs. Healthy, ARF vs. Healthy, and CHD vs. Healthy groups.

Fig 2. TopNet for RHD vs ARF.

The red nodes represent genes in the top active network, the blue nodes represent nodes of the top repressed network, and the purple nodes indicate nodes common to both. Upper-facing triangles represent upregulated genes, and lower-facing triangles indicate nodes that are downregulated in RHD compared to ARF. Functional enrichment of some hub nodes such as GATA2, CCR3, SMAD7, and ALOX15 are also shown.

Fig 3. TopNet for ARF vs.

Healthy. The red nodes represent genes in the top active network, the blue nodes represent nodes of the top repressed network, and the purple nodes indicate nodes common to both. The upper-facing triangles represent upregulated genes, and lower-facings ones indicate downregulated genes in each test condition compared to the reference groups. The DEGs are marked with a larger node size. Enrichment of some of the hub genes has also been shown.

Fig 4. TopNet for RHD vs Healthy.

The red nodes represent genes in the top active network, the blue nodes represent nodes of the top repressed network, and the purple nodes indicate nodes common to both. The DEGs are labeled and represented with larger node sizes. The upper-facing triangles represent genes that are upregulated, and lower-facing triangles indicate nodes that are downregulated in each of the test conditions compared to the reference groups. Enrichment of some of the hub genes has also been shown.

Fig 5. TopNet for CHD vs Healthy.

The red nodes represent genes in the top active network, the blue nodes represent nodes of the top repressed network, and the purple nodes indicate nodes common to both. The DEGs are labeled and represented with larger node sizes. The upper-facing triangles represent genes that are upregulated, and lower-facing triangles indicate nodes that are downregulated in each of the test conditions compared to the reference groups. Enrichment of some of the hub genes has also been shown.

Fig 6

A) Heatmap of the 171 DEGs. Each row denotes a gene while each column represents a sample B) Top 5 pathways that the upregulated and downregulated genes of the 171 candidate genes unique to the RHD vs. ARF group are involved in (ranked based on p-value).