Malaria is a parasitic disease that might lead to acute lung and kidney injuries and, if not treated in time, possibly death. Most malarial incidents are due to the Plasmodium falciparum (P. falciparum) parasite. Malaria can be treated by for instance the drug Malarone® whose effective molecule is the drug molecule atovaquone. Unfortunately, an emergence of P. falciparum resistance to atovaquone and other antimalarial medicaments is seen globally. Therefore, to combat the increasing drug resistance in malarial parasites and to aid in future antimalarial drug development, it is of great importance to investigate the precise inhibition mechanism of antimalarial drugs, such as atovaquone, and the cause of resistance in known resistant P. falciparum mutants. Here we investigate the effect of two bc1 complex mutations, Y279S and L282V (see Figure 1), that are known to cause atovaquone resistance in malarial parasites.
The bc1 complex and the Q-cycle
The effectiveness of atovaquone is due to its interference with the adenosine triphosphate (ATP) synthesis in the mitochondria of the parasite cells. ATP synthesis relies on the electron transport chain (ETC) proteins which are four protein complexes that through a series of electron transfers pump protons across the inner mitochondrial membrane. Because the inner membrane itself is impermeable to protons an electrochemical potential is created across the membrane. The created electrochemical potential is vital for ATP synthesis and thereby the parasitic cell. Atovaquone targets the cytochrome bc1 complex of the ETC proteins in P. falciparum. The bc1 complex is composed of two identical monomers, each consisting of three protein subunits: cytochrome b (Cyt b), the Rieske iron-sulfur protein (ISP), and cytochrome c1 (Cyt c1) see Fig. 2. Furthermore, each monomer contains an iron-sulfur cluster (Fe2S2) in the ISP and three heme groups.
In the bc1 complex, the proton pumping is driven by the Q-cycle. During a Q-cycle two ubiquinol molecules (QH2) are oxidized at the Qo-site where they each release two protons to the intermembrane space and two electrons into the bc1 complex. Two of the released electrons reduce a ubiquinone (Q) molecule at the Qi-site, where simultaneously an uptake of two protons from the mitochondrial matrix completes the reduction. Atovaquone inhibits the Qo-site and thereby blocks the binding site for the QH2 molecules and stops the Q-cycle, which in turn disturbs the ATP synthesis and through that seriously harms or kills the cell and ultimately the parasite.
Binding modes and interaction energies
Using molecular dynamics simulations, the binding modes of two known bc1 complex inhibitors, atovaquone, and stigmatellin, were investigated in the wild type (WT) bc1 complex and the Y279S and L282V mutants. By studying the percentage of simulation time (occupancy) that hydrogen bonds are present between the inhibitors and the three main residues involved in inhibitor binding, see Fig. 3, it was found that in the WT bc1 complex both inhibitors mainly bind to the H181 residue of the ISP. In the Y279S mutant, the number of hydrogen bonds to the H181 residues drastically decreases both in the case of atovaquone and stigmatellin.
The total interaction energy between the inhibitors and the bc1 complex was sampled for each simulation trajectory. Especially, the interaction strength between stigmatellin and the H181 residue was significantly weakened in the Y279S mutant as compared to the WT bc1 complex as can be seen by the peak shifts of the probability distribution functions, P(E), of the total interaction energy, E, in Fig. 4. The shift in interaction energy and hydrogen bond occupancy was in the case of stigmatellin related to a clear displacement of stigmatellin from the H181 residues towards the mutated residue Y279.
Free energy perturbation
Free energy perturbation simulations were performed to determine the binding free energy of atovaquone in the bc1 complex. It is no possible to set up one simulation that directly computes the exact binding free energy, ∆G0. Therefore, one must perform a series of sub-simulations that in steps transform an unbound system into a bound system as illustrated in the thermodynamic cycle in Fig. 5. In the end, the free energy changes evaluated from the sub-simulations, ∆Gi, are added together to obtain the binding free energy. In the case of atovaquone, we identified a ~1.2 kcal/mol weaker binding free energy in the mutated bc1 complexes. However, because the binding environment is relatively elaborate in the bc1 complex, large errors were associated with the results and it was not possible to make a reliable conclusion based on the free energy perturbation investigation.