Structure-based drug discovery against a novel antimalarial drug target, S-adenosylmethionine decarboxylase/ornithine decarboxylase

Abstract

Malaria is one of the most life-threatening diseases affecting mankind, with over 3 billion people being at risk of infection, with most of these people living in Africa, South America and Asia. As the malaria parasite is rapidly becoming resistant to many of the possible treatments on the market, it is of upmost importance to identify new possible drug targets and describe drugs against these that are inexpensive, easy to manufacture and have a long shelf-life in order to combat malaria. One such target is the polyamine pathway. The polyamines putrescine, spermidine, and spermine are crucial for cell differentiation and proliferation. Interference with polyamine biosynthesis by inhibition of the rate-limiting enzymes ornithine decarboxylase (ODC) and S-adenosylmethionine decarboxylase (AdoMetDC) has been discussed as a potential chemotherapy of cancer and parasitic infections. Usually, both enzymes are individually transcribed and highly regulated as monofunctional proteins. However, ODC and AdoMetDC from P. falciparum (PfODC and PfAdoMetDC, respectively) are found as a unique bifunctional protein (PfAdoMetDC/ODC) in the malaria parasite, making it an enticing target for new, selective antimalarial chemotherapies. In order to apply structure-based drug discovery strategies to design inhibitors for PfAdoMetDC/ODC, the atomic resolution structures of these proteins are needed. Each individual domain has had its structure proposed through homology modelling; however atomic resolution structures of these domains are not yet available. The homology model of PfAdoMetDC/ODC has not yet been elucidated due to the interactions between the domains of the bifunctional protein not being fully understood. High levels of recombinant expression of the bifunctional protein have been either unsuccessful or resulted in the formation of insoluble proteins being produced. The purpose of this project is to optimise the recombinant expression of PfAdoMetDC/ODC, and the PfODC domain, to produce high yields of pure, soluble protein for subsequent atomic resolution structure determination. Ultimately, this will enable the utilisation of PfAdoMetDC/ODC in structure-based drug discovery strategies. Overexpression of P. falciparum proteins in E. coli is notoriously difficult, mainly due to the codon bias between the two species. Comparative studies were performed on four constructs of the PfAdoMetDC/ODC gene, containing either the wild-type, fully codon harmonised, or partially codon harmonised gene sequences to analyse the effect codon harmonisation had on protein expression and activity of both domains of PfAdoMetDC/ODC as well as on the monofunctional PfODC domain. Codon harmonisation did not improve the expression levels or the purity of recombinantly expressed PfAdoMetDC/ODC or the monofunctional PfODC domain. Truncated versions of both proteins, and contamination by the E. coli chaperone proteins DnaK and GroEL, were present in the protein samples even after purification by affinity chromatography. However, codon harmonisation improved the activity levels of the PfAdoMetDC domain, while decreasing the activity of the PfODC domain of PfAdoMetDC/ODC. Harmonisation of the monofunctional PfODC domain resulted in a decrease in the activity of the protein. In order to identify possible inhibitors of the PfODC domain of the bifunctional protein, a structure-based drug discovery study was initiated based on a homology model for PfODC. Four hundred compounds with known antimalarial activity were virtually screened against the PfODC homology model and the top two scoring compounds were selected for enzyme inhibition assays based on their predictive binding affinity against the enzyme, and two medium scoring compounds were selected as controls. Enzyme inhibition studies were performed on the bifunctional PfAdoMetDC/ODC to determine the effect the compounds had on both domains of the protein. Of the compounds assayed one of the compounds significantly reduced the activity levels of both domains of PfAdoMetDC/ODC. Additionally, one compound significantly reduced the activity level of the PfAdoMetDC domain of PfAdoMetDC/ODC. This work therefore contributes towards characterisation of the unique PfAdoMetDC/ODC in malaria parasites as a novel drug target.