Viral Network Perturbations
Genetic differences between individuals influence susceptibility to disease. The Human Genome Project, together with the HapMap project and the Human Cancer Genome Project, have greatly accelerated the ability to find genetic variants and associated disease genes. Despite these advances, linking individual genes and their variations to disease remains daunting (Vucic et al, Genome Res 2012). Even where a causal variant has been identified, biological insights have been slow in coming. Phenotypic effects of functional sequence variants are mediated through dynamic networks of gene products and metabolites. Making sense of genotype-to-phenotype relationships thus requires that phenotypes be viewed as manifestations of network properties, rather than simply the result of genomic variations considered individually (Vidal et al, Cell 2011).
Our central hypothesis is that human genetic variations and pathogens such as viruses similarly influence local and global properties of networks to induce disease states (Figure). Our approach to understanding cellular networks is to observe perturbations of network structure by viral pathogen proteins, measuring the effects using interactome mapping, proteomic analysis, and transcriptional profiling.
Systematic virhostome mapping
We examined the global landscape of host perturbations by tumor virus proteins, using a systematic integrated pipeline to investigate at genome-scale perturbations of host interactome and transcriptome networks induced by individual gene products encoded by members of four families of human DNA tumor viruses: polyomaviruses, papillomaviruses, adenovirus, and Epstein-Barr virus (Rozenblatt-Rosen et al, Nature 2012). For two of these human DNA tumor viruses, Epstein-Barr virus and human papillomavirus 16, we modeled the phenotypic consequences of the viral-host interactions through a network-based framework (Gulbahce et al, PLoS Comput Biol 2012). Comparative interactomics, or homology by network comparison rather than by sequence similarity, was pioneered by us for the viral-host interactome networks of E6 and E7 oncoproteins from 11 distinct human papillomavirus genotypes (Neveu et al, Methods 2012).
Systematic to individual cases
One substantial benefit of systematic large-scale systems biology investigations is that the data can be drilled down to improve the modeling of specific aspects of infection and immunity. Exemplifying this benefit is our characterization of the cellular FAM111A protein as a previously unidentified host range restriction factor specifically targeted by SV40 large T antigen (Fine et al, PLoS Pathog 2012). Another example is our finding that the gamma-specific EBV gene BLRF2 interacts specifically with SRPK2 (serine/arginine-rich protein kinase 2) via a RS motif in BRLF2, and that mutation of this motif abrogates viral replication (Duarte et al, PLoS ONE 2013). Ongoing data mining of our virhostome perturbation data will undoubtedly discover more such examples.
In a remarkable validation of our hypothesis that viral-protein interactions can act as surrogates to identify proteins perturbed in human disease, a set of de novo arising heterozygous point mutations in FAM111A have been demonstrated to be behind two phenotypically related developmental disorders of previously unknown molecular etiology, Kenny-Caffey syndrome (KCS) and osteocraniostenosis (OCS) (Unger et al, Am J Hum Genet, 2013). These mutations all map to the same minimal region of FAM111A that is required for large T Antigen binding of Simian Virus 40 (Fine et al, PLoS Pathog 2012), predicting that pathogenesis is due to these mutations disrupting the binding of FAM111A to as yet undiscovered cellular proteins.
Focused Path-Hostome mapping
This group has collaborated with other laboratories for elucidation of pathogen-host interactome networks for Epstein-Barr virus (Calderwood et al, PNAS 2007), hepatitis C virus (de Chassey et al, Mol Syst Biol 2008), influenza virus (Shapira et al, Cell 2009), plant microbial pathogens (Mukhtar et al, Science 2011), human T-cell lymphotropic viruses (Simonis et al, Retrovirology 2012) and Mycobacterium tuberculosis (Mehra et al, PLoS Pathog 2013). Collaborative work on other pathogens, predominantly human viral pathogens, is ongoing.
Our central hypothesis is that human genetic variations and pathogens such as viruses similarly influence local and global properties of networks to induce disease states (Figure). Our approach to understanding cellular networks is to observe perturbations of network structure by viral pathogen proteins, measuring the effects using interactome mapping, proteomic analysis, and transcriptional profiling.
Systematic virhostome mapping
We examined the global landscape of host perturbations by tumor virus proteins, using a systematic integrated pipeline to investigate at genome-scale perturbations of host interactome and transcriptome networks induced by individual gene products encoded by members of four families of human DNA tumor viruses: polyomaviruses, papillomaviruses, adenovirus, and Epstein-Barr virus (Rozenblatt-Rosen et al, Nature 2012). For two of these human DNA tumor viruses, Epstein-Barr virus and human papillomavirus 16, we modeled the phenotypic consequences of the viral-host interactions through a network-based framework (Gulbahce et al, PLoS Comput Biol 2012). Comparative interactomics, or homology by network comparison rather than by sequence similarity, was pioneered by us for the viral-host interactome networks of E6 and E7 oncoproteins from 11 distinct human papillomavirus genotypes (Neveu et al, Methods 2012).
Systematic to individual cases
One substantial benefit of systematic large-scale systems biology investigations is that the data can be drilled down to improve the modeling of specific aspects of infection and immunity. Exemplifying this benefit is our characterization of the cellular FAM111A protein as a previously unidentified host range restriction factor specifically targeted by SV40 large T antigen (Fine et al, PLoS Pathog 2012). Another example is our finding that the gamma-specific EBV gene BLRF2 interacts specifically with SRPK2 (serine/arginine-rich protein kinase 2) via a RS motif in BRLF2, and that mutation of this motif abrogates viral replication (Duarte et al, PLoS ONE 2013). Ongoing data mining of our virhostome perturbation data will undoubtedly discover more such examples.
In a remarkable validation of our hypothesis that viral-protein interactions can act as surrogates to identify proteins perturbed in human disease, a set of de novo arising heterozygous point mutations in FAM111A have been demonstrated to be behind two phenotypically related developmental disorders of previously unknown molecular etiology, Kenny-Caffey syndrome (KCS) and osteocraniostenosis (OCS) (Unger et al, Am J Hum Genet, 2013). These mutations all map to the same minimal region of FAM111A that is required for large T Antigen binding of Simian Virus 40 (Fine et al, PLoS Pathog 2012), predicting that pathogenesis is due to these mutations disrupting the binding of FAM111A to as yet undiscovered cellular proteins.
Focused Path-Hostome mapping
This group has collaborated with other laboratories for elucidation of pathogen-host interactome networks for Epstein-Barr virus (Calderwood et al, PNAS 2007), hepatitis C virus (de Chassey et al, Mol Syst Biol 2008), influenza virus (Shapira et al, Cell 2009), plant microbial pathogens (Mukhtar et al, Science 2011), human T-cell lymphotropic viruses (Simonis et al, Retrovirology 2012) and Mycobacterium tuberculosis (Mehra et al, PLoS Pathog 2013). Collaborative work on other pathogens, predominantly human viral pathogens, is ongoing.