Peter Shengyang Wu, 2009, Chemistry

“Current antiviral drugs work through one of the five following strategies 1) by blocking viral attachment to the host cell, 2) by disrupting replication of viral DNA/RNA, 3) by halting transcription of viral DNA or translation of viral mRNA, 4) or by preventing viral release from the host cell. However no antiviral drug yet takes advantage of the fact that viruses are almost completely dependent on host cell enzymes to produce key precursors for their growth, such as lipids for viral envelopes or amino acids for viral proteins.”

“Drugs that inhibit specific human metabolic enzymes, such as statin drugs and methotrexate, have long been used as effective therapies for high cholesterol and leukemia respectively. Their success demonstrates that regardless of the cause of an illness, many diseases can be effectively managed at the level of metabolism. We assert that viral infections too may be viewed and treated, in part, as metabolic disorders. Inhibiting host-cell enzymes critical to virus replication is a promising new way to approach antiviral drug treatment. The difficulty however is to determine which specific enzymes in the set of over three-thousand known reactions in the human metabolism are critical to virus replication. However, not all the metabolic fluxes can be quantified through direct measurement, and those that cannot must be estimated in other ways. Flux balance analysis (FBA) is a computational technique used to predict the optimal capabilities of large-scale networks at steady state. It has been used, for example, to accurately predict the growth rate of a E. coli culture and the viability of E. coli gene knockouts.

“Flux balance analysis on the entire human metabolism is a new, unexplored, area. In 2006, Duarte et al. presented a global reconstruction of the human metabolism based on information from genomic data and published literature. To date it is the most comprehensive data set of its kind.
In my summer work, I apply FBA to the human model in an attempt to quantify the effect of human cytomegalovirus (HCMV) infection on host cell metabolic fluxes. Ichoose to study HCMV because of the large amount of transcriptional and metabolomic data available to our group that can be used to verify predictions made by the flux balance model. However the method described may be applied to any virus given an accurate description of its chemical composition.”

“I use FBA to answer the following questions: Given a set of experimentally observed fluxes into and out of a host cell, what is the maximum number virus particles that can be produced? What single and double-knockouts of the host-cell enzymes will halt viral replication? What specific reactions are upregulated during viral replication compared to a mock infected cell? Enzymes catalyzing reactions that are both critical and upregulated comprise a promising set of antiviral drug targets. Thanks, huge thanks, to the Grand Challenges in Health for funding my summer work!”