Motivation Hepatitis C virus infection is one of the most common and chronic in the world, and hepatitis associated with HCV infection is a major risk factor for the development of cirrhosis and hepatocellular carcinoma (HCC). Known data on virus-host and host protein interactions enable network-based analyses of viral infection supported by omics data. The study of molecular interaction networks helps to elucidate the mechanistic pathways linking HCV molecular activities and the host response that modulates the stepwise hepatocarcinogenic process from preneoplastic lesions (cirrhosis and dysplasia) to HCC. Several studies have shown that network-based approaches lead to the identification of more robust markers and better stratifications of samples. However, it is still not clear whether acute and chronic effects of HCV activity can be explained according to a local impact hypothesis, i.e. in network proximity to host proteins targeted by viral proteins (HCV targets). We developed a computational approach to examine the relation between HCV targets, host protein-protein interaction (PPI) topology, pathways that regulate HCV response and the expression variations observed in preneoplastic and neoplastic lesions of liver cell samples from HCV-infected patients [1]. Methods We used the PPIs available in the STRING database v9.0 with score greater than 0.7, designated as ''high confidence''.We obtained a PPI network composed of a total of 14,116 unique human proteins involved in 223,088 PPIs. We used the viral-host interactions collected from De Chassey et al. (2008) [2], Kwofie et al. (2011) [3], Dolan et al. (2013) [4], HPIDB [5], Intact [6], VirHostNet [7] and defined 591 unique human proteins that interact directly with HCV proteins, 517 of which appear in the PPI network. We used network propagation algorithm [8] to smooth the HCV interaction information over the PPI network and to rank all the other proteins in relation to their network proximity to HCV targets in the PPI network. This method simulates a random walk on a graph with restarts. Two gene expression datasets collected from the Gene Expression Omnibus (GEO) database that are comparable in terms of histological characteristics (normal, cirrhotic and neoplastic tissues), viral infections (HCV) and microarray platform were used. The dataset GSE6764 [9] includes 75 samples from cirrhotic and neoplastic livers of 38 HCV-infected patients and healthy livers of 10 patients. The dataset GSE14323 [10] includes 108 samples from cirrhotic, neoplastic and normal tissues from 88 HCV-infected patients and 19 HCV seronegative patients. The raw data of the two datasets were separately processed and analyzed using the statistical software R. We considered as differentially expressed the genes with p-value < 0.05. The statistical significance of the overlap between each pair of protein lists was calculated using the hypergeometric distribution. The GSEA (Gene Set Enrichment Analysis) of HCV targets and differentially expressed genes was carried out using the R package HTSanalyzeR. The search of PPIs subnetworks in (i) network proximity to HCV targets and (ii) differentially expressed was formulated as the multi-objective optimization problem of minimizing two objective functions. We solved this problem using an evolutionary algorithm that creates a population of subnetworks extracted from the whole PPI network and, then, iteration by iteration, modifies the subnetworks (adding and removing vertexes) in order to minimize simultaneously the objective functions. Results In this work we ranked the host proteins in relation to their network proximity to viral targets by simulating the impact of HCV-host molecular interactions throughout the host protein-protein interaction (PPI) network. We observed that the set of proteins in the neighborhood of HCV targets in the host interactome is enriched in key players of the host response to HCV infection. We described the use of network propagation for predicting the host proteins that are in a relevant position of the PPI network on the basis of HCV-host interactions and show that network propagation successfully prioritizes proteins that are involved in the host response to HCV. Subsequently, we showed that networks of proteins "guilt-by-association" with HCV are significantly enriched in genes differentially expressed in cirrhotic samples compared to normal samples and in hepatocellular carcinoma compared to cirrhosis. These subnetworks contain established, recently proposed and novel candidate proteins for the regulation of the mechanisms of host response to acute and chronic HCV infection. REFS [1] Mosca E, Alfieri R, Milanesi L (2014) Diffusion of Information throughout the Host Interactome Reveals Gene Expression Variations in Network Proximity to Target Proteins of Hepatitis C Virus. PLoS ONE 9(12): e113660. [2] de Chassey B, Navratil V, Tafforeau L, Hiet MS, Aublin-Gex A, et al. (2008) Hepatitis C virus infection protein network. Mol Syst Biol 4: 230. [3] Dolan PT, Zhang C, Khadka S, Arumugaswami V, Vangeloff AD, et al. (2013) Identification and comparative analysis of hepatitis C virus-host cell protein interactions.t. Mol Biosys 9: 3199-3209. [4] Kwofie SK, Schaefer U, Sundararajan VS, Bajic VB, Christoffels A (2011) HCVpro: hepatitis C virus protein interaction database. Infect Genet Evol 11: 1971-1977. [5] Kumar R, Nanduri B (2010) HPIDB-a unified resource for host-pathogen interactions. BMC Bioinformatics. Suppl 6: S16. [6] Hermjakob H, Montecchi-Palazzi L, Lewington C, Mudali S, Kerrien S, et al. (2004) IntAct: an open source molecular interaction database. Nucleic Acids Res 32: D452-D455. [7] Navratil V, de Chassey B, Meyniel L, Delmotte S, Gautier C, et al. (2009) VirHostNet: a knowledge base for the management and the analysis of proteome-wide virus-host interaction networks. Nucleic Acids Res 37: D661-D668. [8] Vanunu O, Magger O, Ruppin E, Shlomi T, Sharan R (2010) Associating genes and protein complexes with disease via network propagation. PLoS Comput Biol 6: e1000641 [9] Wurmbach E, Chen YB, Khitrov G, Zhang W, Roayaie S, et al. (2007) Genome-wide molecular profiles of HCV-induced dysplasia and hepatocellular carcinoma. Hepatology 45: 938-947. [10] Mas VR, Maluf DG, Archer KJ, Yanek K, Kong X, et al. (2009) Genes involved in viral carcinogenesis and tumor initiation in hepatitis C virus-induced hepatocellular carcinoma. Mol Med 15: 85-94.
Propagation of hepatitis C virus effects throughout the host interactome reveals liver expression variations in the network proximity to Hepatitis C virus targets.
Roberta Alfieri;Ettore Mosca;Luciano Milanesi
2015
Abstract
Motivation Hepatitis C virus infection is one of the most common and chronic in the world, and hepatitis associated with HCV infection is a major risk factor for the development of cirrhosis and hepatocellular carcinoma (HCC). Known data on virus-host and host protein interactions enable network-based analyses of viral infection supported by omics data. The study of molecular interaction networks helps to elucidate the mechanistic pathways linking HCV molecular activities and the host response that modulates the stepwise hepatocarcinogenic process from preneoplastic lesions (cirrhosis and dysplasia) to HCC. Several studies have shown that network-based approaches lead to the identification of more robust markers and better stratifications of samples. However, it is still not clear whether acute and chronic effects of HCV activity can be explained according to a local impact hypothesis, i.e. in network proximity to host proteins targeted by viral proteins (HCV targets). We developed a computational approach to examine the relation between HCV targets, host protein-protein interaction (PPI) topology, pathways that regulate HCV response and the expression variations observed in preneoplastic and neoplastic lesions of liver cell samples from HCV-infected patients [1]. Methods We used the PPIs available in the STRING database v9.0 with score greater than 0.7, designated as ''high confidence''.We obtained a PPI network composed of a total of 14,116 unique human proteins involved in 223,088 PPIs. We used the viral-host interactions collected from De Chassey et al. (2008) [2], Kwofie et al. (2011) [3], Dolan et al. (2013) [4], HPIDB [5], Intact [6], VirHostNet [7] and defined 591 unique human proteins that interact directly with HCV proteins, 517 of which appear in the PPI network. We used network propagation algorithm [8] to smooth the HCV interaction information over the PPI network and to rank all the other proteins in relation to their network proximity to HCV targets in the PPI network. This method simulates a random walk on a graph with restarts. Two gene expression datasets collected from the Gene Expression Omnibus (GEO) database that are comparable in terms of histological characteristics (normal, cirrhotic and neoplastic tissues), viral infections (HCV) and microarray platform were used. The dataset GSE6764 [9] includes 75 samples from cirrhotic and neoplastic livers of 38 HCV-infected patients and healthy livers of 10 patients. The dataset GSE14323 [10] includes 108 samples from cirrhotic, neoplastic and normal tissues from 88 HCV-infected patients and 19 HCV seronegative patients. The raw data of the two datasets were separately processed and analyzed using the statistical software R. We considered as differentially expressed the genes with p-value < 0.05. The statistical significance of the overlap between each pair of protein lists was calculated using the hypergeometric distribution. The GSEA (Gene Set Enrichment Analysis) of HCV targets and differentially expressed genes was carried out using the R package HTSanalyzeR. The search of PPIs subnetworks in (i) network proximity to HCV targets and (ii) differentially expressed was formulated as the multi-objective optimization problem of minimizing two objective functions. We solved this problem using an evolutionary algorithm that creates a population of subnetworks extracted from the whole PPI network and, then, iteration by iteration, modifies the subnetworks (adding and removing vertexes) in order to minimize simultaneously the objective functions. Results In this work we ranked the host proteins in relation to their network proximity to viral targets by simulating the impact of HCV-host molecular interactions throughout the host protein-protein interaction (PPI) network. We observed that the set of proteins in the neighborhood of HCV targets in the host interactome is enriched in key players of the host response to HCV infection. We described the use of network propagation for predicting the host proteins that are in a relevant position of the PPI network on the basis of HCV-host interactions and show that network propagation successfully prioritizes proteins that are involved in the host response to HCV. Subsequently, we showed that networks of proteins "guilt-by-association" with HCV are significantly enriched in genes differentially expressed in cirrhotic samples compared to normal samples and in hepatocellular carcinoma compared to cirrhosis. These subnetworks contain established, recently proposed and novel candidate proteins for the regulation of the mechanisms of host response to acute and chronic HCV infection. REFS [1] Mosca E, Alfieri R, Milanesi L (2014) Diffusion of Information throughout the Host Interactome Reveals Gene Expression Variations in Network Proximity to Target Proteins of Hepatitis C Virus. PLoS ONE 9(12): e113660. [2] de Chassey B, Navratil V, Tafforeau L, Hiet MS, Aublin-Gex A, et al. (2008) Hepatitis C virus infection protein network. Mol Syst Biol 4: 230. [3] Dolan PT, Zhang C, Khadka S, Arumugaswami V, Vangeloff AD, et al. (2013) Identification and comparative analysis of hepatitis C virus-host cell protein interactions.t. Mol Biosys 9: 3199-3209. [4] Kwofie SK, Schaefer U, Sundararajan VS, Bajic VB, Christoffels A (2011) HCVpro: hepatitis C virus protein interaction database. Infect Genet Evol 11: 1971-1977. [5] Kumar R, Nanduri B (2010) HPIDB-a unified resource for host-pathogen interactions. BMC Bioinformatics. Suppl 6: S16. [6] Hermjakob H, Montecchi-Palazzi L, Lewington C, Mudali S, Kerrien S, et al. (2004) IntAct: an open source molecular interaction database. Nucleic Acids Res 32: D452-D455. [7] Navratil V, de Chassey B, Meyniel L, Delmotte S, Gautier C, et al. (2009) VirHostNet: a knowledge base for the management and the analysis of proteome-wide virus-host interaction networks. Nucleic Acids Res 37: D661-D668. [8] Vanunu O, Magger O, Ruppin E, Shlomi T, Sharan R (2010) Associating genes and protein complexes with disease via network propagation. PLoS Comput Biol 6: e1000641 [9] Wurmbach E, Chen YB, Khitrov G, Zhang W, Roayaie S, et al. (2007) Genome-wide molecular profiles of HCV-induced dysplasia and hepatocellular carcinoma. Hepatology 45: 938-947. [10] Mas VR, Maluf DG, Archer KJ, Yanek K, Kong X, et al. (2009) Genes involved in viral carcinogenesis and tumor initiation in hepatitis C virus-induced hepatocellular carcinoma. Mol Med 15: 85-94.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
