New insights into apicoplast metabolism in blood-stage malaria parasites

Abstract

The apicoplast of Plasmodium falciparum is the only source of essential isoprenoid precursors and coenzyme A in the parasite. Isoprenoid precursor synthesis relies on the iron-sulfur cluster (FeS) cofactors produced within the apicoplast, rendering FeS synthesis an essential function of this organelle. Recent reports provide important insights into the roles of FeS cofactors and the use of isoprenoid precursors and Coenzyme A both inside and outside the apicoplast. Here, we review the recent insights into the roles of these metabolites in blood-stage malaria parasites and discuss new questions that have been raised in light of these discoveries.

Introduction

Most of the apicomplexans, including the malaria parasite Plasmodium falciparum, contain a four-membrane relict plastid organelle called the apicoplast (Apicomplexan plastid), with Cryptosporidium spp. and some gregarines being notable exceptions [1]. Following initial discord on the evolutionary origin of this organelle [2–4], it is now widely accepted that the apicoplast is a product of secondary endosymbiosis of a red alga [5–7]. Ultimately, the two-membrane plastid of the alga was retained and acquired two additional membranes: a third membrane corresponding to the algal plasma membrane and a fourth outer membrane thought to be the product of phagocytic engulfment of the algal cell. The organelle contains a highly reduced circular genome (28.7–39.5 kilobases [kb]) which primarily contains genes related to transcription and translation [8–10]. Over the course of evolution, many of the genes originally found in the organelle were lost, while others were transferred to the nuclear genome through the process of endosymbiotic gene transfer [11,12]. These genes share lineage with the original cyanobacterial endosymbiont of chloroplasts and are responsible for a number of distinctly prokaryotic metabolic pathways found in apicomplexans. Owing to its prokaryotic origin and unique features, the apicoplast organelle has intrigued scientists since its discovery.

In P. falciparum, the 35 kb apicoplast genome contains a complete set of transfer RNA (tRNA) genes and ribosomal RNA (rRNA) genes, but only nine genes that could encode proteins [13]. Most of the proteins that function in the apicoplast (several hundred) are nucleus-encoded proteins that are trafficked to the apicoplast via the secretory pathway with the help of topogenic N-terminal bi-partite targeting sequences [11]. The apicoplast contains several metabolic pathways that are primarily composed of these trafficked proteins, including pathways for the synthesis of isoprenoid precursors, iron-sulfur cluster (FeS) cofactors, and fatty acids. The apicoplast also houses several pathways that are shared between different compartments of the parasite, for example, heme biosynthesis (apicoplast, mitochondrion, and cytoplasm) [14] and Coenzyme A (CoA) biosynthesis (apicoplast and cytoplasm) [15]. The fatty acid synthesis (FASII) and heme biosynthesis pathways are essential for the liver and mosquito stages of the parasite life cycle but are dispensable in blood-stage parasites [14,16–19]. A growing body of evidence shows that three apicoplast metabolic pathways (isoprenoid precursor synthesis, FeS synthesis, and CoA synthesis) are essential for blood-stage parasites [15,20–25]. In this review, we will focus on these metabolic pathways and the uses of their products in the apicoplast organelle and in the rest of the parasite cell.

Isoprenoid precursor synthesis and inhibition

Isoprenoids are diverse metabolites that play roles in key cellular processes including post-translational modification of proteins (prenylation), dolichol-mediated glycosylation and GPI synthesis, tRNA modifications, and cofactor synthesis (for example, Heme A and ubiquinone). In P. falciparum, the apicoplast contains the methylerythritol phosphate (MEP) pathway for isoprenoid precursor synthesis [11,21,26]. This pathway consists of seven enzymes all of which are encoded in the nuclear genome and trafficked to the apicoplast (Figure 1). The identification of DOXP synthase (DXPS) and DOXP reductoisomerase (DXPR) transcripts was the first sign of an active MEP pathway in P. falciparum [21]. Subsequently, all seven genes were found in the parasite genome [11] and several pathway metabolites were identified in blood-stage parasites [27]. The MEP pathway uses pyruvate and glyceraldehyde-3-phosphate as initial substrates and ultimately generates the isoprenoid precursors IPP (isopentenyl pyrophosphate) and its regioisomer DMAPP (dimethylallyl pyrophosphate), which are the building blocks for all downstream isoprenoid products (Figure 1).

Figure 1. Isoprenoid precursor biosynthesis and Coenzyme A (CoA) biosynthesis pathways of Plasmodium falciparum and the utilization of their end products.

Isoprenoid precursors isopentenyl pyrophosphate (IPP) and its regioisomer dimethylallyl pyrophosphate (DMAPP) generated by apicoplast methylerythritol phosphate (MEP) pathway are utilized both within and outside the apicoplast. Within the organelle, isoprenoid precursors are used for tRNA modification and the synthesis of long-chain polyprenyls, whereas outside the organelle they are used for protein prenylation in cytosol (black outlined box), ubiquinone synthesis in the mitochondrion (orange outlined box), and dolichol synthesis in the endoplasmic reticulum (blue outlined box). Iron-sulfur clusters (FeS) generated by the FeS synthesis pathway are essential for the activity of two terminal enzymes of MEP pathway. DXPR is the target of fosmidomycin (Fos). The CoA synthesis pathway is shared between the cytosol and apicoplast, where the final product CoA is produced in the apicoplast. Within the organelle, CoA is used for fatty acid biosynthesis (FASII) while outside the organelle it is used in the TCA cycle and acetyl CoA synthesis in the mitochondrion (orange outlined box). Boxes filled in grey are non-essential pathways/processes in the blood-stage parasites. GA3P, glyceraldehyde-3 phosphate; DOXP, 1-deoxy-D-xylulose 5-phosphate; DXPS, DOXP synthase; DXPR, DOXP-reductoisomerase; IspD, 2C-methyl-D-erythritol 4-phosphate cytidyltransferase; IspE, 4-diphosphocytidyl-2C-methyl-D-erythritol (CDP-ME) kinase; CDP-MEP, CDP-ME phosphate; IspF, 2C-methyl-D-erythritol-2,4-cyclodiphosphate (MEcPP) synthase; IspG, hydroxylmethylbutenyl diphosphate (HMB-PP) synthase; IspH, HMB-PP reductase; PPS, polyprenyl synthase; MiaA, tRNA isopentenyltransferase; MiaB, tRNA-i⁶A37 methylthiotransferase; FASII, type II fatty acid synthesis pathway; CoA, Coenzyme A, ACP, acyl carrier protein; ACPS, ACP synthase; DPCK, dephospho-CoA kinase; Pan, pantothenate; GPP, geranyl pyrophosphate; FPP, farnesyl pyrophosphate; GGPP, geranylgeranyl pyrophosphate; FPP/GGPPS, farnesyl/geranylgeranyl pyrophosphate synthase; PPRD, polyprenol reductase.

Fosmidomycin (Fos) blocks DXPR activity (and also inhibits IspD [28]) and has proven to be a valuable reagent in the study of the MEP pathway. Fos has in vitro EC₅₀ of 0.3 to 1.2 μM in parasite cultures [21,27] and has been evaluated as a partner drug with clindamycin [29,30] and piperaquine [31] in phase II human trials. Drug resistance studies link Fos resistance to copy number variations in the gene encoding DXPR [32]. Interestingly, Fos resistance was also associated with mutations in cytosolic enzymes thought to alter the abundance of glycolytic intermediates, including two substrates (pyruvate and glyceraldehyde-3-phosphate) used by the MEP pathway [33–35]. Overall, the MEP pathway is a validated drug target for malaria parasites and significant effort has been put into inhibitor discovery in recent years [36–45].

Apicoplast disruption and metabolic bypass

A variety of antibacterial drugs (such as azithromycin, clindamycin, and doxycycline) kill blood-stage malaria parasites by inhibiting the prokaryotic ribosomes found in the apicoplast [46,47]. A key insight into apicoplast metabolism came from the observation that the isoprenoid precursor IPP can bypass the toxic effects of these antibacterial drugs [23]. Blood-stage parasites treated with lethal doses of translation inhibitors lose the four-membrane apicoplast organelle and its organellar genome, but can be maintained indefinitely in culture as long as at least 200 μM IPP is provided in the culture medium [23]. This phenomenon demonstrated that the apicoplast is essential to produce isoprenoid precursors needed for the synthesis of isoprenoid products elsewhere in the cell. Building on this result, Wiley et al. showed that isoprenoid precursor synthesis is also an essential function of the apicoplast in the sexual stages of parasite development [48].

It is surprising that IPP alone can serve as an apicoplast bypass since the synthesis of complex isoprenoids require both IPP and DMAPP. A computational study [49] identified a candidate IPP isomerase and this enzyme (or one like it) may be responsible for balancing IPP/DMAPP levels. Since IPP is prohibitively expensive, a parasite line called PfMev was engineered to synthesize IPP and DMAPP in the cytosol of P. falciparum parasites [50]. The PfMev line expresses the last three enzymes of the mevalonate pathway which convert mevalonate (Mev; added to the growth medium) to IPP; a fourth enzyme isomerizes IPP to provide DMAPP. Chemical bypass (either IPP or Mev with the PfMev line) made it possible to genetically delete several apicoplast-specific genes that would otherwise result in parasite death [22,50–56], as well as to screen for compounds with apicoplast-specific targets [47,57].

Delayed death and the link to isoprenoid synthesis

Inhibitors that target apicoplast functions fall into two categories based on how quickly treatment leads to growth arrest (reviewed in [58]). Treatment of parasites with some inhibitors, like Fos, kills the parasites in the first parasite life cycle [21,27], which contrasts with delayed death in the second growth cycle following treatment with inhibitors that target different apicoplast ‘housekeeping’ processes such as translation, transcription, and genome maintenance (Figure 2) [58–61]. In a recent study, Kennedy et al. [62] investigated the delayed death phenotype by treating parasites with translation-blocking inhibitors and tracking cellular and metabolic phenotypes. They found that drug-treated parasites still produce intermediates of the MEP pathway and synthesize isoprenoid products such as geranylgeranyl pyrophosphate (GGPP) in the first cell cycle. Presumably, MEP pathway enzymes continue to function despite the lack of newly-synthesized proteins during this period. The inheritance of dysfunctional apicoplasts in the second cell cycle leads to a reduction of isoprenoid products and decreased protein prenylation, leading to vesicular trafficking defects, aberrant uptake of red blood cell cytoplasm, and reduced hemoglobin degradation [62]. Protein prenylation and the vesicular trafficking defects can be restored by providing geranylgeraniol (GGOH), a prenyl precursor that has previously been reported to significantly shift the EC₅₀ of Fos [28]. These results establish protein prenylation as the proximal cause of the delayed death effect.

Figure 2. Inhibition of apicoplast housekeeping functions causes delayed death.

Inside the apicoplast (green outlined box), isoprenoid precursors IPP and DMAPP are important for apicoplast maintenance and have a possible role in membrane fluidity. Outside the organelle, they are required for protein prenylation in the cytosol (necessary for vesicular trafficking), ubiquinone synthesis in the mitochondrion (necessary for the electron transport chain), and dolichol synthesis in the endoplasmic reticulum (necessary for GPI anchor synthesis and N-linked glycosylation). Several classes of antibiotics inhibit housekeeping functions (replication, transcription, or translation of the apicoplast genome), causing apicoplast disruption in the second growth cycle. Loss of the apicoplast genome prevents FeS synthesis and the activity of FeS-dependent MEP pathway enzymes. Parasites fail to grow in the second cycle (delayed death) due to lack of protein prenylation (Left panels). Second cycle defects caused by a reduction of isoprenoid products and decreased protein prenylation can be rescued with geranylgeraniol (GGOH) supplementation (green arrow). However, the delayed death phenotype cannot be rescued with GGOH beyond second cycle, presumably due to depletion of ubiquinone and/or dolichols (Right panels).

The critical role of protein prenylation highlights the significance of the enzyme farnesyl/geranylgeranyl pyrophosphate synthase (FPPS/GGPPS). This polyprenyl synthase is located in the cytosol [63] and is required for the synthesis of FPP and GGPP from the isoprenoid precursors IPP and DMAPP [64,65]. Bypassing this process with GGOH supplementation allows drug-treated parasites to survive a second growth cycle, but they fail to complete a third growth cycle, presumably due to deficiencies in other essential isoprenoid products (such as dolichols or ubiquinone) [62]. Based on experiments that should bypass the need for ubiquinone, Kennedy et al. [62] propose that the third cycle death is likely due to perturbation of dolichol-mediated GPI synthesis (Figure 2). Although little is known about dolichol production in malaria parasites [66,67], a recent study links the activity of two enzymes to dolichol synthesis and localizes one of these enzymes to the ER [68]. Additional work will be needed to firmly establish that these enzymes and/or dolichol synthesis are essential.

The phenomenon of delayed death is linked to disruption of the apicoplast organelle. Inhibition of essential housekeeping functions leads to loss of the four-membrane organelle, loss of its organellar genome, and subsequent loss of essential isoprenoid products. Apicoplast disruption, however, can also accompany inhibitors such as the metalloprotease inhibitor actinonin, which kill parasites in the first growth cycle [69,70]. Other first-cycle inhibitors, such as the MEP pathway inhibitors Fos and FR900098, do not lead to loss of the apicoplast or its genome as long as parasites are supplemented with IPP [47]. Genetic experiments in the PfMev bypass line (with mevalonate supplementation) reinforce the conclusion that loss of essential apicoplast protein function can result in disrupted or intact organelles. The apicoplast remains intact in four MEP pathway enzyme deletion lines [22,50,52], consistent with the results for known MEP pathway inhibitors. Although specific inhibitors of pyruvate kinase II and ferredoxin reductase have not been developed, their phenotypes can be predicted from knockout studies. With Mev supplementation, deletion of ferredoxin or ferredoxin reductase in the PfMev line does not lead to apicoplast disruption [22], whereas deletion of pyruvate kinase II and the two transporters that are responsible for importing its phosphoenolpyruvate substrate (iTPT and oTPT [71]) results in apicoplast disruption [52].

Genetic experiments provide clear answers to the roles of potential drug targets and the phenotypes associated with their loss of function. Inhibitor experiments can be complicated by the effects that they have on multiple apicoplast processes. For example, the tetracycline analog doxycycline acts as a delayed death inhibitor when used at 1 μM or 3 μM, but alters apicoplast morphology and inhibits parasite growth in the first cycle when used at a concentration of 10 μM [60]. First-cycle inhibition can be rescued by exogenous iron, indicating that first-cycle activity involves a dominant metal-dependent mechanism beyond the delayed-death mechanism [72].

Isoprenoids are required for maintenance of the apicoplast organelle

Several studies noted that Fos treatment blocks apicoplast branching and elongation [70,73,74]. Intrigued by this phenomenon Okada et al. [75] investigated this phenotype in detail. Treatment with 10 μM Fos resulted in punctate apicoplasts that do not divide during schizogony and are not passed on to the next generation of parasites. Addition of IPP (or mevalonate in PfMev parasites) rescued apicoplast defects and parasite growth. A similar phenomenon was found for PfMev deletion parasites which are dependent on mevalonate due to loss of the MEP pathway enzyme DXPS [52]. Withdrawal of mevalonate resulted in punctate apicoplasts in the first growth cycle and apicoplast disruption in subsequent cycles. These studies suggested that isoprenoid precursors synthesized de novo within the organelle are required for its own maintenance. The only predicted use of IPP/DMAPP in the organelle is tRNA modification mediated by MiaA and MiaB, both of which have been shown to be dispensable in blood-stage parasites (Figure 1) [22,75]. In search of other isoprenoid metabolism within the organelle, the authors identified an apicoplast-localized polyprenyl synthase (PPS), inducible knockdown of which resulted in similar apicoplast defects as those observed in the Fos inhibition and DXPS knockout studies. A key experiment showed that PPS knockdown was rescued with the long-chain polyprenol decaprenol (C₅₀-OH), but not with GG-OH (C₂₀-OH) or shorter polyprenols. The authors proposed that polyprenols might be critical for apicoplast membrane fluidity, since other possible functions of long-chain polyprenols are not predicted to be relevant for apicoplast biology (Figure 1) [11,76,77].

Transport of isoprenoid precursors in and out of the apicoplast

Studies by Okada et al. [75] and Kennedy et al. [62,67] provide better understanding of the roles of isoprenoid precursors both within and outside the organelle. These studies raise the question of how IPP and DMAPP are exported from the apicoplast for use in the cytosol, a process that should require one or more transporters to cross the four membranes of the organelle. The requirement for isoprenoid metabolism in the apicoplast also raises the question of how apicoplast-intact parasites can be maintained after inhibition [47] or deletion [22,50,52] of MEP enzymes. One explanation is that cytosolic IPP and DMAPP provided by either IPP supplementation or the mevalonate pathway in PfMev parasites is able to enter the apicoplast, possibly by using the same transport mechanism that normally functions to export these metabolites.

Coenzyme A is an essential product of the apicoplast

Coenzyme A (CoA) is a universal cofactor that supports diverse biosynthetic pathways including fatty acid biosynthesis and modification, acetyl-CoA generation, and cellular oxidation and metabolism [24,78]. Pantothenate (also known as vitamin B5) is an essential nutrient for P. falciparum and is exclusively used for the de novo synthesis of CoA [79–81]. Imported pantothenate is converted into dephospho-CoA in four enzymatic steps that are believed to be in the cytoplasm [24,82,83]. The final enzyme, dephospho-CoA kinase (DPCK), generates CoA and was recently localized to the apicoplast (Figure 1) [15]. This arrangement is somewhat surprising and suggests that a transport mechanism exists to import the large, charged substrate dephospho-CoA into the apicoplast and export product CoA back out to the cytosol.

DPCK appears to be essential for blood-stage malaria parasites since efforts failed to delete this enzyme in rodent parasites [25]. Deletion of DPCK was also attempted without success using the PfMev apicoplast bypass line under mevalonate supplementation conditions [15]. To investigate the role of DPCK, a PfMev line was generated to express an apicoplast-targeted construct of E. coli DPCK. The parasite DPCK could now be deleted, indicating that the well-characterized E. coli DPCK could functionally complement loss of the parasite ortholog [15]. Finally, the E. coli DPCK was depleted using a combination of a conditional localization system to trigger secretion of the protein [84,85] and a conditional knockdown system to reduce protein levels [86,87]. Conditional depletion of E. coli DPCK led to parasite death, demonstrating that DPCK activity in the apicoplast is essential for parasite survival.

CoA may be required both within the apicoplast as well as outside the organelle (Figure 1). Within the apicoplast, acyl-carrier protein synthase (ACPS) uses CoA to modify apo acyl carrier protein (apo-ACP) with a 4′-phosphopantetheine group to generate holo-ACP. Holo-ACP plays an integral role in fatty acid synthesis and elongation by shuttling nascent fatty acids between enzymes of FASII pathway [11,17,24,88,89]. CoA is also used for generation of acetyl-CoA and malonyl-CoA [90–92] which feed the FASII pathway. Although FASII pathway enzymes are dispensable in blood-stage parasites [17–19], forward genetic datasets indicate that both ACP and ACPS are essential [93,94] and may be required for a function other than fatty acid biosynthesis. The mitochondrion is another compartment that should require CoA. Although the TCA cycle is not essential in blood-stage parasites [95,96], a recent investigation shows that TCA cycle enzymes produce acetyl-CoA that is used for acetylation reactions throughout the cell [97].

Apicoplast disruption and the role of FeS clusters

Parasites with a disrupted apicoplast accumulate vesicles throughout the cell (Figure 3) [20,23]. These vesicles contain nucleus‐encoded apicoplast‐specific proteins that presumably would have been trafficked to the organelle under typical conditions [47,51,61,73]. PfMev parasites were used to collect transcriptomic and metabolomic data sets comparing apicoplast-intact and apicoplast-disrupted parasites over the course of the parasite developmental cycle [50]. Expression of genes from the apicoplast genome were not detectable in apicoplast-disrupted parasites, as expected due to loss of the organellar genome, but there was no significant change in the expression of genes from the nuclear genome, including the cohort of genes known to encode apicoplast-specific proteins [50]. Consistent with these findings, metabolite levels were largely unchanged after disruption of the apicoplast [50].

Figure 3. Active pathways in apicoplast-disrupted parasites with isopentenyl pyrophosphate (IPP) supplementation.

In apicoplast-disrupted parasites, protein prenylation in the cytosol (black outlined box), ubiquinone synthesis in the mitochondrion (orange outlined box), and dolichol synthesis in the endoplasmic reticulum (blue outlined box) can remain functional as long as the parasites are exogenously supplemented with IPP. The dephospho-CoA kinase (DPCK) enzyme of the Coenzyme A (CoA) synthesis pathway remains active in the vesicles (green box) seen in apicoplast-disrupted parasites which provide CoA for mitochondrial processes (orange outlined boxes). Boxes shaded grey are non-essential pathways/processes in blood-stage parasites. Pan, pantothenate; DMAPP, dimethylallyl pyrophosphate; GPP, geranyl pyrophosphate; FPP, farnesyl pyrophosphate; GGPP, geranylgeranyl pyrophosphate; FPP/GGPPS, farnesyl/geranylgeranyl pyrophosphate synthase; PPRD, polyprenol reductase.

It is possible that vesicle‐localized proteins remain biochemically active in apicoplast-disrupted parasites (Figure 3). This appears to be the case for DPCK and explains why this enzyme could not be deleted, even in the presence of the mevalonate bypass. To test this idea, azithromycin was used to trigger apicoplast disruption in the PfMev line expressing E. coli DPCK (with the endogenous DPCK deleted) [15]. Parasites could be maintained with disrupted apicoplast vesicles containing E. coli DPCK, but conditional knockdown and mislocalization of DPCK led to parasite death, even with mevalonate supplementation [15]. These experiments demonstrated that DPCK is still active and essential after apicoplast disruption. DPCK activity in disrupted apicoplast vesicles would require the import or synthesis of substrates dephospho-CoA and ATP and export of product CoA. Identification of apicoplast membrane proteins in similar vesicles of Toxoplasma gondii, a related apicomplexan [98], suggests apicoplast transporters could remain active.

CoA production in disrupted apicoplast vesicles provides an explanation for why CoA supplementation is not required in apicoplast bypass experiments. This raises the question of why IPP is required for apicoplast bypass and why isoprenoid precursors are not synthesized in the vesicles. One explanation to this question has to do with loss of the apicoplast genome which encodes the FeS synthesis protein SufB [20,22]. Loss of SufB should lead to inactivation of all FeS cluster enzymes in the apicoplast. FeS cluster proteins were systematically deleted in a recent study, revealing that only proteins necessary for the MEP isoprenoid precursor pathway are essential in blood-stage parasites [22,99]. Thus, apicoplast disruption should lead to loss of MEP pathway activity but may not lead to the loss of other apicoplast metabolic pathways.

New paradigms expand our understanding of apicoplast biology

The apicoplast of blood-stage malaria parasites has been described as producing only a single product (isoprenoid precursors) for use in other compartments of the cell. Okada et al. [75] now show that isoprenoid products are essential for the maintenance of the apicoplast organelle itself, while Kennedy et al. [62] provide new insight into the utilization of isoprenoid precursors outside the organelle and how these requirements are responsible for the delayed death phenomenon. The discovery that CoA is the second essential product of the apicoplast raises questions about how this metabolite is still produced in apicoplast-disrupted parasites and whether other metabolic pathways remain active after apicoplast disruption [15]. Loss of the apicoplast genome and the ability to synthesize FeS cofactors could be a key factor in determining which metabolic pathways will function and which will not after apicoplast disruption. Future experiments will build on recent insights into apicoplast biology to address these gaps in knowledge.

Highlights.

• Isoprenoid precursors and Coenzyme A are essential metabolites generated by apicoplast localized enzymes

• Isoprenoids play an essential role in apicoplast biogenesis

• Depletion of isoprenoids outside the apicoplast results in the delayed-death phenomenon

• Some apicoplast metabolic pathways remain active in vesicles following organelle disruption, while other pathways do not.