The ‘interactome’ of a host/pathogen triad

In order to survive and replicate within their hosts, viruses must manipulate those pathways and systems in which their host relies upon for its own survival. However, this model gets more complicated with those viruses successfully infecting multiple host species. For example, Dengue virus (DENV) – an emerging pathogen which causes over 50 million cases a year of a mild to deadly disease – infects both humans and mosquito species of the Aedes genus. Thus to accomplish survival, DENV must interact with proteins from these two distantly related hosts. Given this complexity, understanding this dual-host/pathogen system is considerably difficult yet as Doolittle and Gomez (2011) show, computational approaches based on structural predictions of viral and host proteins may allow for the accurate prediction of the complex in vivo ‘interactome’.

Transmission of DENV - the principle mode is direct mosquito to human

The group set out to understand the interactions between both DENV encoded proteins and those of its hosts – humans and Aedes mosquitos. Using previously determined structural information for human and fly (relatively closely related insect to Aedes) and how these proteins interact with each other, they were able to map these back on to host infection. They searched databases for structural similarities between dengue proteins and those from its host (human or fly) – these ‘dengue similar host proteins’ were used to search for host-host protein interactions as a surrogate for host-dengue interactions. 

 “The computational methodology employed to generate this map assumes that proteins with comparable structures will share interaction partners. Therefore, we predict that DENV2 proteins may merge into the host protein interactome at the points normally occupied by structurally similar host proteins, creating an interface for the manipulation of downstream host processes.”

Using this approach, they built up a network of possible host/pathogen interactions, assuming that DENV proteins can participate in the same interactions as host proteins. Of course, this method over estimates interactions so to counter this, they prioritised particular interactions for further study based on previously published, validated in vivo work and those interactions still left hopefully were functionally accurate and important. This approach had previously been used to study human-HIV-1 protein interactions.

DENV capsid structure

Following significantly limiting their map down to those that had been previously validated the biological functions of host target proteins and dengue-similar proteins were analysed to determine whether the predicted functions matched those that would be important for viral infection in both humans and mosquitoes. As shown above, DENV-like proteins participate in interactions involved in diverse processes – importantly including cell death, signalling cascades, immune response and metabolism. They focus the investigation into DENV manipulation of host apoptosis and innate immune signalling and also those proteins which are shared between both insect and human hosts.

They suggest that due to the disparity in the known molecular biology of dengue/host interactions this computational methodology has its limitations in this system yet these data should be used as a springboard for future investigations and hypotheses. This study highlights the importance of global computational analysis in determining basic host/pathogen biology especially in a system which has been poorly studied like DENV.

Doolittle, J., & Gomez, S. (2011). Mapping Protein Interactions between Dengue Virus and Its Human and Insect Hosts PLoS Neglected Tropical Diseases, 5 (2) DOI: 10.1371/journal.pntd.0000954

Dyer MD, Murali TM, & Sobral BW (2007). Computational prediction of host-pathogen protein-protein interactions. Bioinformatics (Oxford, England), 23 (13) PMID: 17646292