Intro
Plasmodium falciparum is a protist parasite that causes malaria and has killed 881,000 people for the most recently tabulated year of 2006.1 This loss of life occurs despite recent reductions in malaria prevalence by intervention with insecticide treated bed nets and artemisinin-based combination drug therapy.2 Unfortunately, artemisinin drug resistance has already been observed and threatens to reverse recent reductions in malaria prevalence.3,4 Continued emphasis on malaria drug development and concomitant characterization of parasite biology are necessary to pursue the ultimate goal of malaria eradication.
Discovery that P. falciparum contains a plastid housing many proteins of prokaryotic origin led to immediate speculation that the metabolic pathways resident in this plastid, termed the apicoplast, could be targeted for drug development without side effects to the human host.5 Fatty acid biosynthesis is one such pathway that has received considerable attention since many antibiotics already exist to target this pathway.6 Two types of fatty acid synthase (FAS) pathway are found in nature. A type I FAS pathway is composed of large multifunctional proteins which catalyze all of the reactions required for de novo fatty acid synthesis, and is typically found in the cytosol of eukaryotic cells. A type II FAS pathway uses independent polypeptides to perform each catalytic step, and is typically bacterial. P. falciparum is unusual for a eukaryote in that it relies upon a complete type II pathway which is located in the apicoplast.
Central to the type II FAS pathway is a protein called acyl carrier protein (ACP). ACP acts as a hub for the type II FAS, shuttling the nascent fatty acid between enzymes of the pathway.7 All ACPs contain a flexible phosphopantetheine prosthetic group derived from coenzyme A, which is attached to a conserved serine residue. The phosphopantetheine group is attached post-translationally by holo-ACP synthase, converting apo-ACP into holo-ACP, here referred to as ACP for simplicity. Acyl groups are covalently bound to ACP by forming a thioester bond with the terminal sulfhydryl of this prosthetic group. Within bacteria such as E. coli (Ec), EcACP is thought to be monomeric, with the phosphopantetheine sulfhydryls either reduced or acylated.8 The phosphopantetheine sulfhydryl can also be readily oxidized, forming disulfide-linked ACP dimers. ACP dimer was observed in early work in E. coli9 and more recently in a selection of ACPs expressed heterologously in E. coli10, although the dimeric species is believed to form during gel electrophoresis and not in vivo.
In P. falciparum, the most dangerous form of malaria, PfACP has been localized exclusively to the apicoplast.11PfACP contains a phosphopantetheine prosthetic group12 and can serve as a cofactor for fatty acid biosynthesis enzymes when it is reduced in the monomeric state.13 While oxidized dimers of PfACP are observed in vitro,14 presence of the dimer has not been reported in vivo in any model system. In E. coli, EcACP is present at 0.1 mM,15 and glutathione is present at 0.1-10 mM.16 When EcACP is not acylated, glutathione prevents oxidized EcACP dimers by forming an EcACP-glutathione mixed disulfide, as observed in stationary phase E. coli.9 The apicoplast shares common ancestry with chloroplasts,17 which also contain ACP functioning in a type II FAS. In spinach chloroplasts, ACP-glutathione conjugates are not observed, and ACP with free sulfhydryl was the major species in seedlings and leaves under light conditions, and acyl-ACP dominates in leaves left in the dark.18
During the human blood stage of P. falciparum, it is not clear that oxidized PfACP dimer can be prevented by the same mechanisms used in either bacteria or chloroplasts. Inactivity of FAS during the blood stage19,20 means that acyl-PfACP species are probably not synthesized. Since glutathione biosynthesis occurs in the cytosol of P. falciparum,21PfACP-glutathione mixed disulfides are probably not formed in the apicoplast of the parasite. One candidate for maintaining a pool of reduced PfACP in the apicoplast is ferredoxin. Ferredoxin has been localized to the apicoplast and has been proposed as a source of reducing equivalents for the organelle.22 While metabolic models predict that many enzymes located in the apicoplast require a reducing environment to function, explicit experimental data on the redox potential within the apicoplast have not yet been described.
To address the biochemical and structural implications of an oxidized PfACP dimer, here we report the crystal structures of PfACP in both the reduced and oxidized disulfide-linked states. Surprisingly, we find constituents of the disulfide-linked dimer to be in close molecular contact, raising questions about accessibility of the disulfide bond to cellular reducing agents. We examine solvent accessibility of the disulfide bond in vitro as a function of ionic strength and reducing agent, finding that the dimer interface protects the disulfide bond when probed by BME, but less so for other common reducing agents. Finally, we directly probe the oxidative state of PfACP in blood stage parasites under conditions of either normal growth or oxidative stress. Using hydrogen peroxide or the disulfide inducing compound diamide, our studies show that PfACP is maintained in the reduced state in vivo, thus suggesting that the apicoplast is a highly buffered reducing compartment.