Toxoplasma gondii requires its plant-like heme biosynthesis pathway for infection

Loading metrics

Open Access
Peer-reviewed

Research Article

Amy Bergmann, 

Katherine Floyd, 

Melanie Key, 

Carly Dameron, 

Kerrick C. Rees, 

L. Brock Thornton, 

Daniel C. Whitehead, 

Iqbal Hamza, 

Zhicheng Dou

x

Published: May 14, 2020

https://doi.org/10.1371/journal.ppat.1008499

?This is an uncorrected proof.

AbstractHeme, an iron-containing organic ring, is essential for virtually all living organisms by serving as a prosthetic group in proteins that function in diverse cellular activities ranging from diatomic gas transport and sensing, to mitochondrial respiration, to detoxification. Cellular heme levels in microbial pathogens can be a composite of endogenous de novo synthesis or exogenous uptake of heme or heme synthesis intermediates. Intracellular pathogenic microbes switch routes for heme supply when heme availability fluctuates in their replicative environment throughout infection. Here, we show that Toxoplasma gondii, an obligate intracellular human pathogen, encodes a functional heme biosynthesis pathway. A chloroplast-derived organelle, termed apicoplast, is involved in heme production. Genetic and chemical manipulation revealed that de novo heme production is essential for T. gondii intracellular growth and pathogenesis. Surprisingly, the herbicide oxadiazon significantly impaired Toxoplasma growth, consistent with phylogenetic analyses that show T. gondii protoporphyrinogen oxidase is more closely related to plants than mammals. This inhibition can be enhanced by 15- to 25-fold with two oxadiazon derivatives, lending therapeutic proof that Toxoplasma heme biosynthesis is a druggable target. As T. gondii has been used to model other apicomplexan parasites, our study underscores the utility of targeting heme biosynthesis in other pathogenic apicomplexans, such as Plasmodium spp., Cystoisospora, Eimeria, Neospora, and Sarcocystis.
Author summary
Toxoplasma gondii infects essentially all warm-blooded animals due to its broad species and tissue tropism. Almost one-third of the human population carry Toxoplasma infection, which can cause severe morbidity and mortality in immunocompromised individuals. The current antibiotics against Toxoplasma trigger strong side effects in some groups of patients and have limited efficacy on congenital toxoplasmosis. Thus, an urgent need for novel therapeutics exist. Here, we show that Toxoplasma gondii actively produces heme, a key nutrient for many subcellular activities, via its plant-like heme biosynthesis pathway for intracellular growth and acute virulence. We found that several herbicidal heme biosynthesis inhibitors and their derivatives show inhibitory effects against intracellular Toxoplasma growth. Our findings provide evidence that disrupting heme production in Toxoplasma could be an effective therapeutic strategy to control infection.

Citation: Bergmann A, Floyd K, Key M, Dameron C, Rees KC, Thornton LB, et al. (2020) Toxoplasma gondii requires its plant-like heme biosynthesis pathway for infection. PLoS Pathog 16(5):
e1008499.

https://doi.org/10.1371/journal.ppat.1008499Editor: Laura J. Knoll, University of Wisconsin Medical School, UNITED STATESReceived: November 8, 2019; Accepted: March 25, 2020; Published: May 14, 2020Copyright: © 2020 Bergmann et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Data Availability: All relevant data are within the manuscript and its Supporting Information files.Funding: This work was supported by the Clemson Startup fund (to Z.D.), Knights Templar Eye Foundation Pediatric Ophthalmology Career-Starter Research Grant (to Z.D.), NIH R01AI143707 (to Z.D.), and a pilot grant of an NIH COBRE grant P20GM109094 (to Z.D.), NIH R01AI067979 (to I.H.). Additionally, the Clemson University Open Access Publishing Fund provided partial financial support to defray the publication fee. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.Competing interests: The authors declare no competing interests.

IntroductionHuman protozoan pathogens share some common nutrient metabolism pathways with their counterparts in the host but show distinct features. For example, two well-known representative species within the Apicomplexa Phylum, Toxoplasma gondii and Plasmodium spp., encode a complete heme biosynthesis pathway in their genomes [1] (Fig 1A and S1 Table). The eight heme biosynthetic enzymes residing within this pathway are delivered to three subcellular locations [1–4], including the mitochondrion, cytoplasm, and apicoplast. The apicoplast is a remnant chloroplast and specifically exists in apicomplexan parasites. Heme typically serves as a prosthetic group in proteins, such as cytochromes, which play an essential role in mitochondrial respiration through the electron transport chain in parasites [5,6]. In addition, Toxoplasma encodes orthologs of two other hemoproteins, cytochrome P450 and catalase, in its genome, suggesting that heme is likely involved in detoxication in Toxoplasma. In contrast, the malaria parasites lack both genes [1].
Fig 1. Toxoplasma gondii encodes its de novo heme biosynthetic pathway within 3 subcellular locations.a, The working model of the de novo heme biosynthesis in Toxoplasma parasites. The enzymes catalyzing the de novo heme biosynthesis are distributed within three subcellular locations in the parasites, whereas they are only localized in the mitochondria and cytoplasm in mammals. b, Determination of the expression of the heme biosynthetic genes in Toxoplasma during its acute infection and their subcellular locations by endogenous gene tagging with 3xHA or 3xmyc epitopes. A subunit of Toxoplasma mitochondrial ATPase (TgF1β) and an apicoplast-associated thioredoxin family protein (TgATrx1) were used as the mitochondrial and apicoplast markers, respectively. TgActin was used as a cytoplasm marker. Bar=2 μm. ALA, 5-aminolevulinic acid; ALAS, 5-aminolevulinic acid synthase; Api, apicoplast; Copro III, coproporphyrinogen III; CPOX, Coproporphyrinogen III oxidase; FECH, Ferrochelatase; Gly, glycine; HMB, hydroxymethylbilan; IVN, intravacuolar network; Mito, mitochondria; PBG, porphobilinogen; PBGD, Porphobilinogen deaminase; PBGS, Porphobilinogen synthase; PPIX, protoporphyrin IX; PPO, Protoporphyrinogen oxidase; Protogen, protoporphyrinogen IX; PV, parasitophorous vacuole; PVM, parasitophorous vacuole membrane; SucCoA, Succinyl-CoA; Toxo, Toxoplasma gondii; UROD, Uroporphyrinogen III decarboxylase; Urogen III, uroporphyrinogen III; UROS, Uroporphyrinogen III Synthase.

https://doi.org/10.1371/journal.ppat.1008499.g001The previous studies have successfully expressed active recombinant Toxoplasma porphobilinogen synthase (TgPBGS), the second enzyme residing within the parasite’s heme biosynthesis pathway, in E. coli, and solved its three-dimensional structure [7,8]. Moreover, succinylacetone (SA), an inhibitor targeting TgPBGS activity, was shown to suppress intracellular Toxoplasma growth at a half maximal inhibitory concentration (IC50) of ~ 2 mM, which sheds light on the therapeutic potential of targeting the heme biosynthetic pathway against toxoplasmosis [2]. A recent genome-wide CRISPR screen in Toxoplasma calculated the fitness scores of all eight heme biosynthetic genes to be below -2.7, indicating that the heme biosynthesis pathway is critical in parasite growth [9]. However, it still remains unknown whether the entire pathway is active for heme production during Toxoplasma infections and the extent to which Toxoplasma relies on this pathway for its infection.
As an intracellular parasite, Toxoplasma utilizes the host plasma membrane to create its own membrane-bound compartment for intracellular replication, termed the parasitophorous vacuole (PV) [10]. The PV membrane (PVM) is permeable to small solutes. Studies have demonstrated that putative nutrient pores exist on the PVM, which allow small substances with molecular weights less than ~1,300 Da to diffuse into the PV [11]. While heme molecules are significantly smaller than 1,300 Da, free heme, however, is toxic and therefore unlikely to be found in a free form. Instead, they are loosely associated with proteins or small ligands that collectively constitute the cellular heme labile pool within mammalian cells [12–14]. Therefore, it is unlikely that Toxoplasma is able to acquire free heme from the host cells via putative nutrient pores. Since Toxoplasma can ingest and digest host proteins to support its growth [15], Toxoplasma may acquire heme through liberating it from the digestion of the host’s hemoproteins. Collectively, both the parasite’s de novo heme production and/or heme acquisition from the host may contribute to parasite replication and infection.
Interestingly, the de novo heme biosynthesis pathway exhibits diverse patterns within the Apicomplexa Phylum. By ortholog search, several other human and animal pathogens besides Toxoplasma and Plasmodium, including Cyclospora cayetanensis, Cystoisospora suis, Eimeria tenella, Hammondia hammondi, Neospora caninum, and Sarcocystis neurona, encode a complete or a partial heme biosynthetic pathway (S1 Table). Orthologs missing for some heme biosynthetic genes could be due to incomplete coverage of genome sequencing or diverse primary sequences of these orthologs in apicomplexan parasites. For example, a UROS ortholog is not identified in the Plasmodium genome by primary sequence alignment with annotated UROS proteins, but a recent bioinformatic study reported a putative Plasmodium UROS ortholog with low similarity to annotated UROS [16]. In addition, the enzymatic activity of UROS has been observed in recombinant Plasmodium PBGD [17], suggesting its unique dual roles in the heme biosynthesis within Plasmodium. In contrast, Babesia, Theileria, and Cryptosporidium spp., completely lack the intact pathway [3], suggesting that these apicomplexan parasites have to scavenge heme from the host. Moreover, malaria parasites switch their requirement for de novo heme production at different infection stages. The de novo heme biosynthesis is dispensable in the blood-stage infection of Plasmodium falciparum [4,18], but is required for its liver-stage infection [19,20], suggesting that intracellular pathogens could switch their heme requirements based on heme availability.

Results and discussionTo test whether all 8 genes residing within Toxoplasma’s de novo heme biosynthesis pathway are expressed in the acute infection stage, we endogenously inserted epitope tags at their C-termini (S1A Fig). Immunoblotting revealed their active expression during acute toxoplasmosis (S1C Fig). In addition, fluorescence localization experiments confirmed that Toxoplasma distributes its heme biosynthesis throughout the mitochondrion, cytoplasm, and apicoplast (Fig 1B, S1B Fig, and S1 Text). Overall, our findings revealed that Toxoplasma maintains de novo heme biosynthetic components during acute infection.
Given the possibility that Toxoplasma could rely on its de novo heme production or scavenge heme or its intermediates from the host to support infection, we deleted 5-aminolevulinic acid (ALA) synthase (TgALAS, TGGT1_258690), the first enzyme in the pathway, in NanoLuc luciferase-expressing wildtype (WT::NLuc) Toxoplasma. Disruption of TgALAS will specifically block the parasite’s de novo heme biosynthesis but maintain the downstream pathway intact for possible utilization of host-derived heme intermediates (Fig 1A). Generation of Δalas in standard growth medium was unsuccessful until the medium was supplemented with 300 μM ALA, the product of TgALAS, suggesting that the de novo heme production is essential for parasite infection. To test this, we evaluated the growth of the resulting Δalas::NLuc mutant after it was starved in ALA-free medium for 144 h. The pre-starved Δalas::NLuc exhibited severe growth defects compared to WT::NLuc and ΔalasALAS::NLuc (a TgALAS complementation strain), and also grew more slowly relative to the non-starved Δalas::NLuc parasites in the medium lacking ALA (Fig 2A), suggesting that the stored heme reserve within parasites enhances their intracellular growth when the parasites encounter insufficient heme production. In contrast, the pre-starved Δalas::NLuc mutant showed comparable growth compared to the non-starved Δalas::NLuc when they were grown in the ALA-containing medium (Fig 2A), indicating that the parasites can quickly respond to extracellular ALA for heme production. The pre-starved Δalas::NLuc mutant displayed an extremely slow growth rate when it was grown in the ALA-free medium (Fig 2A), suggesting that the parasites may incorporate a residual level of heme and/or heme synthesis intermediates from the host. Similarly, we observed that the Δalas::NLuc parasites showed severe defects in intracellular replication (S3A Fig) and plaque development (S3B Fig). The addition of 300 μM ALA in the medium enhanced the replication of TgALAS-deficient parasites regardless of pre-starvation status (S3A Fig). We also observed that the plaque formation of Δalas::NLuc was partially restored in the medium supplemented with 300 μM ALA (S3B Fig and S1 Text). The addition of extracellular ALA could result in enhanced production and accumulation of toxic porphyrin intermediates in host cells and infected parasites [6], which may impair intracellular parasite growth. To test whether ALA can boost parasite growth to a greater extent at ALA concentrations below 300 μM, we compared the growth rates of Δalas::NLuc in media containing 300, 100, 33.3, 11.1, and 3.7 μM ALA. Our data showed that the growth of Δalas::NLuc was enhanced in the media supplemented with 100 μM and 300 μM ALA, and exhibited a higher increase under 300 μM ALA than 100 μM ALA (S3C Fig). This observation suggests that Toxoplasma parasites cannot easily access extracellular ALA, probably because ALA needs to cross multiple membranes to reach the parasite’s mitochondrion. To test the role of TgALAS in parasite acute virulence, we subcutaneously injected the Δalas::NLuc mutant along with WT::NLuc and ΔalasALAS::NLuc strains into CD-1 mice and did not observe mortality in the mice infected with the Δalas::NLuc strain, even when its inocula were 103- and 104-fold higher than that required for WT parasites to establish a lethal infection (Fig 2B). As expected, the infections derived from WT::NLuc and ΔalasALAS::NLuc strains were lethal at 10–12 days post-infection (Fig 2B). The parasite infection in the surviving mice was confirmed by seroconversion and their resistance to subsequent challenge with WT parasites.
Fig 2. Toxoplasma parasites principally rely on their de novo heme biosynthesis for intracellular growth and pathogenesis.a, Growth comparison of the ALA-starved and non-starved Δalas::NLuc parasites in media containing or lacking ALA. The parasites were grown in confluent HFFs and their luciferase activities were measured every 24 h for up to 96 h. Data represent mean ± SEM of n=3 biological replicates. b, Acute virulence determination of TgALAS-deficient parasites in a murine model. Ten mice of equal numbers of males and females were used for each strain. c, Evaluation of repression efficiency of TgFECH by ATc treatment via immunoblotting. TgFECH was endogenously tagged with a 3xmyc tag at its C-terminus for recognition by immunoblotting. The lysates were also probed against TgGRA7 as a loading control. d, Replication assessment of the TgFECH knockdown parasites. T7S4-TgFECH and its parental strains were pre-treated with ATc for the period described in the scheme before replication assay. Data represent mean ± SD of n=3 biological replicates. e-f, Replication assay of Δcpox and Δppo parasites. Data represent mean ± SD of n=3–4 biological replicates. g-h, Acute virulence measurement of Δcpox and Δppo parasites in a murine model. 5 male and 5 female mice were used for each strain. The statistical significance for each animal study in b, g, and h was calculated using the Log-rank (Mantel-Cox) test. Statistical significance in the rest of the studies was calculated by two-tailed unpaired Student’s t-test. *, p
Read More