Abstract
Malaria continues to be the most lethal protozoan disease of humans and the pathogenesis is fundamentally associated with the infection and hemolysis of red blood cells. Due to the emergence of resistance to most current drugs, there is an urgent need to develop a new generation of anti-parasitic agents. Drug development programs are expensive, long-term endeavors with numerous bottlenecks that exhibit a high rate of attrition. A major concern following the scientific and financial investment in drug discovery is the emergence of drug resistance. This is a well documented problem in malaria, and may be exceedingly rapid, classically demonstrated by pyrimethamine-resistant Plasmodium falciparum malaria. Strategies therefore that identify the most suitable drug target sites to minimize resistance are of major interest. In this study, a novel approach to select such sites based on the evolutionary rate of change is described, using the P. falciparum glycerol kinase (PfGK) as an example. The ratio of non-synonymous (dN) to synonymous (dS) nucleotide substitutions is defined as omega and was used to identify the patterns of evolutionary change at individual codons in the parasite and orthologous human (HsGK) coding sequences. The omega value of a particular codon reflects the evolutionary forces acting on the corresponding amino acid in the protein sequence. Natural selection will retain mutations that are beneficial to the organism and eliminate those that are detrimental. Omega values typically fall into three categories:
positive selection (omega>1.0), neutral (omega=1.0), or purifying selection (omega<1.0).
In this study, we quantified the relative intensity of selection and introduced the category of extreme purifying selection (omega≤0.1) to identify sites under the most severe evolutionary constraints. We have termed this novel approach to drug target selection “evolutionary patterning” (EP). EP describes the pattern of evolutionary change across a coding sequence, thereby identifying residues that make the most (omega<0.1) and least (omega>1.0) suitable drug target sites based on their potential to produce viable mutations. The EP approach was validated using the P. falciparum dihydrofolate reductase gene. Pyrimethamine targets the dihydrofolate reductase enzyme and five mutations conferring drug resistance have been identified. We hypothesized that none of these mutations would be under extreme purifying selection and our EP investigation confirmed this. EP analysis was thus applied to PfGK, which could be a potential novel drug target. PfGK is annotated as a putative glycerol kinase in the PlasmoDB database and to confirm this predicted function, the full length gene of 1506bp was cloned into a pGEX-4T2 expression vector, the recombinant GST-fusion protein was expressed in E coli and an in vitro assay showed that the enzyme was active and could phosphorylate glycerol. Glycerol-3-phosphate is a multifunctional metabolite that is essential for glycerolipid synthesis and also feeds into glycolysis, highlighting its essential role in parasite metabolism. EP analysis of the PfGK and HsGK genes was conducted separately as part of protozoan and metazoan clades, respectively, and key differences in the evolutionary patterns of the two molecules were identified. These differences were exploited to target the parasite selectively and six potential drug target sites were chosen, which contained residues under extreme purifying selection. To assess the functional and structural significance of these regions, as well as their accessibility to potential therapeutic molecules, they were mapped onto a 3D model of PfGK. This analysis ruled out three of the potential sites, since they were either not essential for enzyme activity or were embedded in the hydrophobic core of the enzyme. In collaboration with medicinal chemists the remaining three potential drug target sites will be used for in silico drug design and docking studies. The strategy of EP and refinement with structural modeling is generic in nature and will limit the development of drug resistance. This represents a significant advance for drug discovery programs in malaria and other infectious diseases.
Large scale enterohemorrhagic E coli population genomic analysis using whole genome typing reveals recombination clusters and potential drug target
