Limits...
Layered functional network analysis of gene expression in human heart failure.

Zhu W, Yang L, Du Z - PLoS ONE (2009)

Bottom Line: The characteristics of the gene expression pattern of the four layers were compared.In the extracellular and plasma membrane layers, there were more proteins encoded by down-regulated genes than by up-regulated genes, but in the other two layers, the opposite trend was found.This study illustrated that by incorporating subcellular localization information into a PPI network based analysis, one can derive greater insights into the mechanisms underlying ICM.

View Article: PubMed Central - PubMed

Affiliation: Institute of Clinical Pharmacology, The Second Affiliated Hospital of Harbin Medical University, The University Key laboratory of Hei Long Jiang Province, Heilongjiang, China.

ABSTRACT

Background: Although dilated cardiomyopathy (DCM) is a leading cause of heart failure (HF), the mechanism underlying DCM is not well understood. Previously, it has been demonstrated that an integrative analysis of gene expression and protein-protein interaction (PPI) networks can provide insights into the molecular mechanisms of various diseases. In this study we develop a systems approach by linking public available gene expression data on ischemic dilated cardiomyopathy (ICM), a main pathological form of DCM, with data from a layered PPI network. We propose that the use of a layered PPI network, as opposed to a traditional PPI network, provides unique insights into the mechanism of DCM.

Methods: Four Cytoscape plugins including BionetBuilder, NetworkAnalyzer, Cerebral and GenePro were used to establish the layered PPI network, which was based upon validated subcellular protein localization data retrieved from the HRPD and Entrez Gene databases. The DAVID function annotation clustering tool was used for gene ontology (GO) analysis.

Results: The assembled layered PPI network was divided into four layers: extracellular, plasma membrane, cytoplasm and nucleus. The characteristics of the gene expression pattern of the four layers were compared. In the extracellular and plasma membrane layers, there were more proteins encoded by down-regulated genes than by up-regulated genes, but in the other two layers, the opposite trend was found. GO analysis established that proteins encoded by up-regulated genes, reflecting significantly over-represented biological processes, were mainly located in the nucleus and cytoplasm layers, while proteins encoded by down-regulated genes were mainly located in the extracellular and plasma membrane layers. The PPI network analysis revealed that the Janus family tyrosine kinase-signal transducer and activator of transcription (Jak-STAT) signaling pathway might play an important role in the development of ICM and could be exploited as a therapeutic target of ICM. In addition, glycogen synthase kinase 3 beta (GSK3B) may also be a potential candidate target, but more evidence is required.

Conclusion: This study illustrated that by incorporating subcellular localization information into a PPI network based analysis, one can derive greater insights into the mechanisms underlying ICM.

Show MeSH

Related in: MedlinePlus

Heatmaps of the number of PPIs.The shade of red color in each grid represents the number of PPIs between the two crossed clusters. a, The number of PPIs between proteins encoded by genes of significantly over-represented GO biological processes. b, The number of PPIs between proteins in differential layers encoded by up and down regulated genes.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2712681&req=5

pone-0006288-g003: Heatmaps of the number of PPIs.The shade of red color in each grid represents the number of PPIs between the two crossed clusters. a, The number of PPIs between proteins encoded by genes of significantly over-represented GO biological processes. b, The number of PPIs between proteins in differential layers encoded by up and down regulated genes.

Mentions: To assess the relationship between the significantly over-represented biological processes, the 254 encoded proteins were analyzed by using the Cytoscape plugin GenePro [15] (Annex S2). A heatmap was generated to display the number of PPIs in each cluster and between every two clusters (Figure 3a). The heatmap revealed that the greatest number of PPIs was found between clusters in which proteins were encoded by up-regulated genes. However, there were a few PPIs that were shared between the up-regulated and down-regulated biological processes. Notably, three PPIs were found that were shared between biological processes involving protein import into nucleus and immune response activation that included JAK2 (Janus kinase 2): STAT6 (Signal transducer and activator of transcription 6), JAK2:STAT3 (Signal transducer and activator of transcription 3) and GSK3B (Glycogen systhase kinase 3 beta):BCL3 (B-cell cll/lymphoma 3). The 1182 proteins encoded by up and down-regulated genes were also analyzed by utilizing GenePro (Annex S3). A heatmap was generated by distributing these proteins into 8 clusters according to their subcellular localization and gene expression pattern (Figure 3b). Once again we found that the number of PPIs identified between clusters was associated with gene expression pattern rather than protein subcellular localization. For example, there were 142 PPIs identified between proteins encoded by up-regulated genes in the nucleus and cytoplasm but only 16 PPIs identified between the proteins encoded by up and down-regulated genes in the nucleus. There were only five PPIs identified between proteins encoded by up-regulated genes in the nucleus and down-regulated genes in the cytoplasm.


Layered functional network analysis of gene expression in human heart failure.

Zhu W, Yang L, Du Z - PLoS ONE (2009)

Heatmaps of the number of PPIs.The shade of red color in each grid represents the number of PPIs between the two crossed clusters. a, The number of PPIs between proteins encoded by genes of significantly over-represented GO biological processes. b, The number of PPIs between proteins in differential layers encoded by up and down regulated genes.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2712681&req=5

pone-0006288-g003: Heatmaps of the number of PPIs.The shade of red color in each grid represents the number of PPIs between the two crossed clusters. a, The number of PPIs between proteins encoded by genes of significantly over-represented GO biological processes. b, The number of PPIs between proteins in differential layers encoded by up and down regulated genes.
Mentions: To assess the relationship between the significantly over-represented biological processes, the 254 encoded proteins were analyzed by using the Cytoscape plugin GenePro [15] (Annex S2). A heatmap was generated to display the number of PPIs in each cluster and between every two clusters (Figure 3a). The heatmap revealed that the greatest number of PPIs was found between clusters in which proteins were encoded by up-regulated genes. However, there were a few PPIs that were shared between the up-regulated and down-regulated biological processes. Notably, three PPIs were found that were shared between biological processes involving protein import into nucleus and immune response activation that included JAK2 (Janus kinase 2): STAT6 (Signal transducer and activator of transcription 6), JAK2:STAT3 (Signal transducer and activator of transcription 3) and GSK3B (Glycogen systhase kinase 3 beta):BCL3 (B-cell cll/lymphoma 3). The 1182 proteins encoded by up and down-regulated genes were also analyzed by utilizing GenePro (Annex S3). A heatmap was generated by distributing these proteins into 8 clusters according to their subcellular localization and gene expression pattern (Figure 3b). Once again we found that the number of PPIs identified between clusters was associated with gene expression pattern rather than protein subcellular localization. For example, there were 142 PPIs identified between proteins encoded by up-regulated genes in the nucleus and cytoplasm but only 16 PPIs identified between the proteins encoded by up and down-regulated genes in the nucleus. There were only five PPIs identified between proteins encoded by up-regulated genes in the nucleus and down-regulated genes in the cytoplasm.

Bottom Line: The characteristics of the gene expression pattern of the four layers were compared.In the extracellular and plasma membrane layers, there were more proteins encoded by down-regulated genes than by up-regulated genes, but in the other two layers, the opposite trend was found.This study illustrated that by incorporating subcellular localization information into a PPI network based analysis, one can derive greater insights into the mechanisms underlying ICM.

View Article: PubMed Central - PubMed

Affiliation: Institute of Clinical Pharmacology, The Second Affiliated Hospital of Harbin Medical University, The University Key laboratory of Hei Long Jiang Province, Heilongjiang, China.

ABSTRACT

Background: Although dilated cardiomyopathy (DCM) is a leading cause of heart failure (HF), the mechanism underlying DCM is not well understood. Previously, it has been demonstrated that an integrative analysis of gene expression and protein-protein interaction (PPI) networks can provide insights into the molecular mechanisms of various diseases. In this study we develop a systems approach by linking public available gene expression data on ischemic dilated cardiomyopathy (ICM), a main pathological form of DCM, with data from a layered PPI network. We propose that the use of a layered PPI network, as opposed to a traditional PPI network, provides unique insights into the mechanism of DCM.

Methods: Four Cytoscape plugins including BionetBuilder, NetworkAnalyzer, Cerebral and GenePro were used to establish the layered PPI network, which was based upon validated subcellular protein localization data retrieved from the HRPD and Entrez Gene databases. The DAVID function annotation clustering tool was used for gene ontology (GO) analysis.

Results: The assembled layered PPI network was divided into four layers: extracellular, plasma membrane, cytoplasm and nucleus. The characteristics of the gene expression pattern of the four layers were compared. In the extracellular and plasma membrane layers, there were more proteins encoded by down-regulated genes than by up-regulated genes, but in the other two layers, the opposite trend was found. GO analysis established that proteins encoded by up-regulated genes, reflecting significantly over-represented biological processes, were mainly located in the nucleus and cytoplasm layers, while proteins encoded by down-regulated genes were mainly located in the extracellular and plasma membrane layers. The PPI network analysis revealed that the Janus family tyrosine kinase-signal transducer and activator of transcription (Jak-STAT) signaling pathway might play an important role in the development of ICM and could be exploited as a therapeutic target of ICM. In addition, glycogen synthase kinase 3 beta (GSK3B) may also be a potential candidate target, but more evidence is required.

Conclusion: This study illustrated that by incorporating subcellular localization information into a PPI network based analysis, one can derive greater insights into the mechanisms underlying ICM.

Show MeSH
Related in: MedlinePlus