|
Network-based drug design -
Optimization of drug selectivity based on different
control structures in parasite and host metabolism
Barbara Bakker
The validation of drug targets is often based on lethality of gene knock-outs.
Although essentiality is a prerequisite for a target enzyme, it is not sufficient.
Enzyme inhibitors do not block an enzyme by 100 %, but only approach total
inhibition asymptotically. Therefore, we should take into account to what
extent a putative target enzyme needs to be inhibited before a sufficient
inhibition of flux is reached.
The development of integrative bioinformatics has made it possible to predict
the effect of enzyme inhibitors on the flux through metabolic routes
and on the concentrations of intracellular metabolites. Previously we constructed
a detailed and realistic model of glycolysis in the tropical parasite Trypanosoma
brucei. The principles of Metabolic Control Analysis were applied to the model
and it was calculated which enzymes exert most control on the flux of ATP
synthesis. This led us to the conclusion that glucose transport was the most
effective target for drugs directed against trypanosome glycolysis, followed
by glyceraldehyde 3-phosphate dehydrogenase, phosphoglycerate kinase, aldolase
and glycerol-3-phosphate dehydrogenase.
Our present research is focused on the selectivity of antiparasitic drugs.
In this context, selectivity is defined as the percentage inhibition of the
metabolic flux of interest in the parasite, divided by the percentage inhibition
of the corresponding flux in the host cells. The effect of an inhibitor on
a metabolic flux is determined simultaneously by 1. the direct effect of the
inhibitor on the target enzyme, as expressed quantitatively by an elasticity
coefficient, and 2. the control exerted by the target enzyme on the flux through
the pathway, as expressed by its flux control coefficient. The elasticity
coefficients can be optimized mainly, but not solely, by structure-based drug
design. The control coefficients are properties of the metabolic network as
a whole and therefore they can only be optimized by network-based drug design.
Optimization of overall drug selectivity requires the optimization of structure-based
as well as of network-based selectivity. We will demonstrate the different
effects that contribute to drug selectivity in a simple core model. Our aim
is to extend the analysis to realistic models of parasite and host metabolism.
|
|
