Integrated human-virus metabolic stoichiometric modelling predicts host-based antiviral targets against Chikungunya, Dengue and Zika viruses

Abstract

Current and reoccurring viral epidemic outbreaks such as those caused by the Zika virus illustrate the need for rapid development of antivirals. Such development would be facilitated by computational approaches that can provide experimentally testable predictions for possible antiviral strategies. To this end, we focus here on the fact that viruses are directly dependent on their host metabolism for reproduction. We develop a stoichiometric, genome-scale metabolic model that integrates human macrophage cell metabolism with the biochemical demands arising from virus production and use it to determine the virus impact on host metabolism and vice versa. While this approach applies to any host-virus pair, we first apply it to currently epidemic viruses Chikungunya, Dengue and Zika in this study. We find that each of these viruses causes specific alterations in the host metabolic flux towards fulfilling their biochemical demands as predicted by their genome and capsid structure. Subsequent analysis of this integrated model allows us to predict a set of host reactions, which, when constrained, inhibit virus production. We show that this prediction recovers known targets of existing antiviral drugs, specifically those targeting nucleotide production, while highlighting a set of hitherto unexplored reactions involving both amino acid and nucleotide metabolic pathways, with either broad or virus-specific antiviral potential. Thus, this computational approach allows rapid generation of experimentally testable hypotheses for novel antiviral targets within a host.

Keywords: antiviral targets; emerging viruses; flux balance analysis; host–virus interactions; metabolic modelling.

Figure 1.

Comparison of host macrophage and viral biomass compositions, and metabolic. (a) Comparison of host macrophage, CHIKV, DENV and ZIKV biomass compositions, as described from their respective biomass objective functions, using five different macromolecular classes (amino acids, RNA, DNA, sugar and lipids. Full breakdown of biomass is available in electronic supplementary material, file S1). (b) Bipartite graph visualization of the macrophage metabolic network, where nodes are metabolites (white fill) or reactions (grey fill), edges are connections between them and indicate directionality. CHIKV, Chikungunya virus; DENV, Dengue virus; ZIKV, Zika virus.

Figure 2.

Fold-change difference in usage of amino acids and nucleotides between host and CHIKV, DENV and ZIKV. (a,b) The usage of amino acids (a) and nucleotides (b) between the host and virus biomass objective functions. The differential usage was calculated against all biomass precursors. Comparison was conducted for all 20 amino acids, and four RNA nucleotides (the x-axes are labelled with the standard short notations for these). All calculations and biomass formulations are as described in the Methods, and all biomass stoichiometric values are provided as electronic supplementary material, file S1.

Figure 3.

Comparison of model fluxes between host optima and CHIKV, DENV and ZIKV optima. (a) Comparisons are visualized as the sum of fluxes over aggregated subsystems using values from host- and virus-optimal states. Abbreviations used in the subsystem classification are: FAS, fatty acid synthesis; ROS, reactive oxygen species; UFAS, unsaturated fatty acid synthesis; Misc., miscellaneous. The y-axis represents differential usage of aggregate subsystems, while the colours of the bars indicate different viruses and host (see colour coding on the panel). Positive and negative values reflect a higher or lower total flux for that subsystem in the virus- compared to host-optimal state. Pie charts over each bar provide a summary of changes on individual reactions within a subsystem. The complete set of flux values for all reactions in the model and all optimal states are provided as electronic supplementary material, file S1. (b) Simplified schematic showing reactions involved in the glycolysis pathway. (c) Corresponding flux ranges of individual reactions in the glycolysis pathway that allow attainment of host and virus optima. The flux ranges allowing optima for individual viruses, as well as the host, are shown in differentially coloured bars, with the x-axis showing flux values. The colour coding is as shown in panel a.

Figure 4.

Prediction of antiviral targets [reactions] from single-reaction perturbations in a virus-optimized system for CHIKV, DENV and ZIKV. (a) Simplified schematic for single-reaction knockouts, where flux ranges for a desired reaction flux v are shown for both host- and virus-optimized systems. Square brackets, associated with specific flux vector indices 1–7, denote the lower (lb) and upper (up) flux bounds ([lb,ub]). Under knockout conditions both lb and ub are set to zero. (b) Simplified schematic for single-reaction enforcements, where flux ranges for a desired reaction flux v are shown for both host- and virus-optimized systems. Square brackets, associated with specific flux vector indices 1–7, denote the lower (lb) and upper (up) flux bounds ([lb,ub]). Under host-derived enforcement conditions, lb and ub are set to e− and e+, respectively (see Methods). (c) Reaction pathway schematic showing the top 29 reactions from host-derived flux enforcement analysis and their associated antiviral drugs and inhibitors. Key reactions inhibiting virus optima when flux ranges derived from host and virus flux variability analysis are enforced (the same reactions listed in electronic supplementary material, file S4). Abbreviations used for the compounds and reactions are as in electronic supplementary material, tables S2 and S3, respectively. Some of the identified reactions are interconnected, forming pathways. Coloured reaction arrows indicate pathways associated with subsystems: orange, pyrimidine synthesis; purple, pentose phosphate pathway; blue, purine synthesis; green, nucleotide biosynthesis. The starting metabolites into these pathways, glutamine and d-ribose 5-phosphate, are derived from glutamine biosynthesis and pentose phosphate pathways. Reactions targeted by a known antiviral or inhibitor are marked by white and red filled stars and circles, respectively. Complete list of antiviral compounds from which the matches were obtained is provided as electronic supplementary material, table S1. A complete list of inhibitors and the associated reactions is provided as electronic supplementary material, file S4. A complete list of enforcement results for all reactions is provided as electronic supplementary material, files S2 and S5.