Apicoplast metabolism

What functions are potentially localized to the apicoplast? As noted above, the apicoplast genome does not encode any proteins that provide clues to its metabolic role. Most of the genes encode proteins required for transcription or translation. A potential ortholog of ClpC (a chaperone component of the Tic complex) is encoded by the 35-kb element.

One gene (originally named ORF470, then ycf24, and now sufB) was first thought to encode an ABC transporter, but has more recently been suggested to be involved in iron metabolism (Rangachari et al., 2002; Wilson et al., 2003). Hence, most of the clues to apicoplast function come from the predicted NEAT proteins. Using PlasmoAP, Ralph et al. (2004) identified more than 500 P. falciparum proteins as potentially targeted to the apicoplast. The identification of the constellation of potential apicoplast proteins provides clues as to why the organelle is essential. Nonetheless, about 70 percent of the predicted NEAT proteins are of unknown function, being annotated as hypothetical proteins (Ralph et al., 2004). Further analysis will be required to determine which of these predicted proteins are expressed and are targeted to the apicoplast.

As yet, the P. falciparum data have not been systematically compared with the T. gondii predicted proteins. Nonetheless, like the genetic systems in the apicoplast, most NEAT proteins are likely to be cyanobacterial in origin, with an evolutionary history that does not overlap with that of the human host. Therefore, many of these pathways would, if essential, provide excellent potential drug targets. Some of the functions mapped to the apicoplast have already been experimentally demonstrated to be essential through the use of inhibitors, or have a high probability of being essential given their biological importance. A summary of metabolic pathways predicted or demonstrated to reside in the apicoplast is given in Figure 9.10.

Acyl carrier protein, part of the type II fatty acid biosynthesis pathway, was one of the first proteins recognized to be targeted to the apicoplast (Waller et al., 1998). The type II pathway, typical of plants and algae, is mediated by a series of individual enzymes, all of which have been identified in the P. falciparum and T. gondii genomes. In contrast, the type I pathway used by mammals is mediated by a single, large multidomain molecule. Scans of the T. gondii genome database indicate that the parasites also contain a gene encoding a protein related to this multidomain molecule (cited as Crawford and Roos, unpublished, in Wiesner and Seeber, 2005). While it is clear that T. gondii can take up fatty acids from the host cell and perhaps

FIGURE 9.10 Metabolism of the apicoplast and mitochondrion. Several metabolic pathways have been associated with the apicoplast (depicted with four membranes) and the mitochondrion (depicted with two membranes), based on predicted targeting sequences. For some, localization of the relevant enzymes to the organelle has been experimentally confirmed. Two distinct pathways for synthesis of Fe-S cluster biosynthesis appear to be present. In contrast, heme biosynthesis requires the collaboration of enzymes in the two organelles as well as the cytosol.

FIGURE 9.10 Metabolism of the apicoplast and mitochondrion. Several metabolic pathways have been associated with the apicoplast (depicted with four membranes) and the mitochondrion (depicted with two membranes), based on predicted targeting sequences. For some, localization of the relevant enzymes to the organelle has been experimentally confirmed. Two distinct pathways for synthesis of Fe-S cluster biosynthesis appear to be present. In contrast, heme biosynthesis requires the collaboration of enzymes in the two organelles as well as the cytosol.

utilize the type I pathway to synthesize additional fatty acids, new data indicate that the type II pathway is essential (Bisanz et al., 2006). Thiolactomycin, which inhibits the type II pathway for fatty acid synthesis in bacteria, can inhibit growth of both T. gondii and P. falciparum (Waller et al., 1998; unpublished data cited in McFadden and Roos, 1999). Similarly, triclosan, which inhibits enoyl-ACP-CoA reductase (which mediates the last step in the fatty acid synthesis cycle) kills both T. gondii and P. falciparum (McLeod et al., 2001; Perozzo et al., 2002). Interestingly, P. falciparum lacks the type I pathway. Conversely, Cryptosporidium parvum has only the type I pathway (Zhu et al., 2004). This observation correlates with the apparent lack of an apicoplast in C. parvum (Riordan et al., 1999; Zhu et al., 2000a; Keithly et al., 2005), and the absence of apicoplast-encoded and NEAT genes in C. parvum (Abrahamsen et al., 2004) and Cryptosporidium hominis (Xu et al., 2004), as well as the lack of Cryptosporidium genes corresponding to known apicoplast pathways.

Acetyl CoA is required for fatty acid synthesis, and is generated from pyruvate by the action of the multienzyme complex pyruvate dehydrogenase. Recent studies show that T. gondii and P falciparum possess genes encoding the subunits for an apicoplast pyruvate dehydrogenase, even though they lack a mitochondrial form (Foth et al., 2005; McMillan et al., 2005). The first step in production of fatty acids from acetyl CoA is catalyzed by acetyl CoA carboxylase (ACC). Plastid ACCs are inhibited by aryloxyphenoxypropionate herbicides, and T. gondii growth is inhibited by these compounds (Zuther et al., 1999). There are two T. gondii ACCs; one is cytosolic and the other has been identified and localized to the apicoplast (Jelenska et al., 2001). Recombinant ACC corresponding to the apicoplast-localized protein has been confirmed as a target of aryloxyphenoxypropionates (Jelenska et al., 2002).

Isoprenoid synthesis is another pathway localized to the apicoplast. The enzymes mediating this pathway are distinct from those of the mevalonate pathway for isoprenoid synthesis found in the human host. The parasite pathway, called the DOXP pathway for its early intermediate 1-deoxy-D-xylulose 5' phosphate, has been more thoroughly studied in P. falciparum. Fosmidomycin, an antibiotic which inhibits DOXP reductase, is toxic to the malaria parasite (Jomaa et al., 1999). However, it shows little activity against T. gondii despite presence of a T. gondii DOXP reductase gene. Resistance could result from specific changes in the target or lack of bioavailability.

How do the substrates for these pathways enter the apicoplast? By analogy with the chloroplast, it is likely that specific carriers are required. The malaria parasite possesses two sugar phosphate transporters, one of which carries an apicoplast targeting sequence, which could mediate the uptake of substrates (such as glyceraldehyde-3-phosphate or phosphoenolpyruvate) into the plastid (Mullin et al., 2006). Our analysis of the T. gondii genome at 10x coverage suggests a single sugar phosphate transporter is present (Karnataki et al., unpublished results). It is challenging to predict the specificity of this transporter based on sequence alone, and in some cases plastid sugar phosphate transloca-tors transport multiple substrates, albeit with differing affinities (Knappe et al., 2003).

Accessory systems required for the functions of the above biosynthetic processes also must reside in the apicoplast. A redox system is present, comprised of ferredoxin and its partner ferredoxin-NADP+ reductase (Vollmer et al., 2001). The ferredoxin reductase is most similar to those of non-photosynthetic tissues of plants, which provide reducing equivalents for biosynthetic processes. In the apicoplast, NADPH is a required co-factor for DOXP reductase in isoprenoid synthesis and for the KAR (FabG) enzyme involved in apicoplast fatty acid biosynthesis. Lipoylation is needed for activity of the pyruvate dehydrogenase complex, and T. gondii does possess an apicoplast-targeted lipoic acid synthase (Thomsen-Zieger et al., 2003). Fatty acid biosynthesis requires the action of an acetyl CoA carboxylase, which contains a cova-lently attached biotin prosthetic group. The biotin group can be visualized with FITC-labeled strepta-vidin and co-localizes with apicoplast markers, confirming the predicted localization (Jelenska et al., 2001). An enzyme required for biotin ligation is also expected to be apicoplast-localized, and indeed P. falciparum possesses a gene encoding such an enzyme with a predicted apicoplast targeting sequence (Gardner et al., 2002).

Iron-sulfur clusters fulfill a variety of functions in proteins, most notably facilitating electron transfer. Several of the proteins that reside in the apicoplast contain Fe-S clusters, including ferredoxin and the enzymes GcpE (IspG) and LytB (IspH) involved in the last steps of isoprenoid synthesis. Lipoic acid synthase is also an Fe-S protein. Iron-sulfur clusters do not form spontaneously - they require the action of multiple proteins for their efficient formation. Seeber (2002) has recently summarized the data concerning the generation of Fe-S clusters in apicomplexans and contrasted it with those in other species. The series of reactions are highly conserved. The generation of elemental sulfur from cysteine is mediated by the action of cysteine desulfurase, and its reduction is coupled to the oxidation of iron (brought in by a siderophore) on a specific protein scaffold. Additional proteins are required for the release of Fe-S clusters from the scaffold, and their coordination with other proteins. Ferredoxin is needed for reductive steps, and several additional proteins are also important. The compartmentation of these reactions is likely beneficial due to the generation of reactive intermediates. In this way, damage to other cellular components is avoided. In most eukaryotes, Fe-S cluster synthesis occurs in the mitochondrion. Recent work indicates that the apicoplast is likely to be a site for the synthesis of Fe-S clusters. The apicoplast encoded ORF470 is homologous, if not orthologous, to sufB (Wilson et al., 2003), which lies in the same bacterial operon as other genes involved Fe-S biosynthesis in bacteria (Nachin et al., 2001, 2003). Genes encoding other proteins likely to be involved in Fe-S cluster synthesis have been identified in P. falciparum, and several have been used to predict apicoplast targeting sequences (Ellis et al., 2001).

Heme biosynthesis is another organellar process in eukaryotes, occurring in the mitochondrion of the human host. Once again, compartmentation may serve to protect the cell from potentially damaging intermediates or byproducts. In apicom-plexaus, heme biosynthesis is accomplished through a metabolic cooperation between the apicoplast and mitochondrion. All eight enzymes required for the de novo synthesis of heme are present in the genome of P. falciparum (Gardner et al., 2002). The enzyme mediating the first step in the pathway is S-aminolevulinic acid synthase. It contains a predicted mitochondrial targeting sequence, and, in malaria parasites, GFP fusions with this sequence are targeted to the mitochondrion (Sato and Wilson, 2002; Sato et al., 2004). The next two enzymes in the pathway, porpho-bilinogen synthase and hydroxymethylbilane synthase, have bipartite leader sequences, and GFP fusions are targeted to the plastid (Sato et al.,

2004). Porphobilinogen synthase belongs to the cyanobacterial/plastid clade (Obornik and Green,

2005), as expected for a plastid-targeted protein. However, the same authors note that two enzymes functioning later in the pathway, porphobilinogen deaminase and uroporphyrinogen decarboxylase, do not cluster within this clade (Obornik and Green, 2005). Nonetheless, the latter has a predicted apicoplast targeting sequence (Sato et al., 2004). Phylogenetic inference indicates that apicomplexan ferrochelatase, the last enzyme in the pathway, evolutionarily clusters with red algal and proteobacterial sequences. Its location has not been experimentally determined, but is proposed to be mitochondrial (Sato et al., 2004). The variety in evolutionary origins of these enzymes illustrates the mosaic nature of the apicomplexan genome.

0 0

Post a comment