Metabolic interactions between Toxoplasma gondii and its host

Toxoplasma gondii is an obligate intracellular parasite belonging to the phylum Apicomplexa that infects all warm-blooded animals, including humans. T. gondii can replicate in every nucleated host cell by orchestrating metabolic interactions to derive crucial nutrients. In this review, we summarize the current status of known metabolic interactions of T. gondii with its host cell and discuss open questions and promising experimental approaches that will allow further dissection of the host–parasite interface and discovery of ways to efficiently target both tachyzoite and bradyzoite forms of T. gondii, which are associated with acute and chronic infection, respectively.


Introduction
Toxoplasma gondii is an obligate intracellular parasite that infects all warm-blooded animals, including birds and humans, where it can replicate in all nucleated cells within a nonfusogenic compartment termed the parasitophorous vacuole (PV). The fast-replicating tachyzoites cause the acute phase of the infection. However, in response to immune pressure, differentiation into the slow-growing or quiescent bradyzoites, which is accompanied by the transformation of the PV membrane (PVM) into a cyst wall, marks the onset of the chronic phase. In general, T. gondii transmission to humans occurs via ingestion of tissue cysts contained in contaminated undercooked meat products. In addition, humans can be infected following exposure to oocysts shed by the definitive feline host via contaminated food or water. T. gondii infections are largely asymptomatic during both the acute and chronic phases, and the chronic stage persists for the life of the host. However, upon severe immunosuppression, rapid replication of T. gondii tachyzoites, derived from the reactivation of encysted bradyzoites, can cause severe symptoms such as encephalitis, which is potentially fatal. Additionally, primary infection during pregnancy is associated with serious consequences for newborns, ranging from blindness and deafness to mental retardation and stillbirth. However, as is the case with many other intracellular pathogens, the intruder is not unnoticed, as evidenced by transcriptomic and metabolomic disturbances within the host. The latter uses these cues to counteract the infection on biochemical as well as immunological levels (for review, see 1-3).
Both tachyzoites in PVs and bradyzoites in cysts integrate tightly with their host cells to ensure nutrient supply for optimal proliferation and persistence, which define the characteristics of the specific life cycle stage 4-6 . In this review, we will briefly outline recent findings of how T. gondii tachyzoites and bradyzoites interact with their host cells for the acquisition of essential nutrients. Recent progress in molecular genetics as well as biochemical and metabolomic methods has resulted in a better understanding of the competition for the shared pool of nutrients. In many respects, T. gondii exhibits metabolic traits similar to those of Plasmodium species, the causative agents of malaria, particularly with regard to the understudied hepatic stage. Therefore, insights from this pathogen's metabolism can also instruct efforts to exploit metabolic dependencies as drug targets in more than one apicomplexan parasite.

Nutrient acquisition from the host cell: not only a transporter issue
T. gondii tachyzoites and bradyzoites replicate intracellularly and therefore need to acquire nutrients from their host cells. The parasite establishes a vacuole that is initially composed of host lipids but during its active invasion process excludes most host proteins 7 . The PVM is then heavily modified by parasite proteins that mediate protein export and the import of lipidic and polar metabolites. It has long been known that the PVM is freely permeable for molecules as large as 1,300 Da 8 , but only recently were two parasite proteins-GRA17 and GRA24defined as the molecular constituents of this pore 9 . It is thought to be permissive for non-directional passive transport of small nutrients, such as vitamins, sugars, amino acids, nucleobases, nucleosides, and nucleotides. It provides a putative mechanism for the export of catabolites, such as lactic acid and, to a lesser extent, alanine and bicarbonate 10 . Besides this pore, an intravacuolar membranous tubulo-vesicular network (membranous tubules and vesicles that are bridging the PVM with the parasite 11 ) has been shown to be involved in the uptake of host proteins and lipids [12][13][14] . Whether other host metabolites can enter the PV via this route and likewise whether it is used on the other hand by the parasite as a "waste pipe" are unknown.
Most polar metabolites are imported through a range of transporters in both the PVM and the parasite's plasma membrane. To date, a small subset has been functionally characterized; hence, there likely exists a larger interaction surface and undiscovered redundancy. Strikingly, however, there are lower numbers of computationally annotated transporter families in the genomes of parasitic protozoa 15 when comparing both intracellular and extracellular living parasites with unicellular free living organisms ( Figure 1). This is despite the need to scavenge as many nutrients as possible from the environment. Explanations for these lower numbers could include a broader substrate specificity of individual transporters combined with a minimal need for diversification due to host niches with complex but predictable compositions. Several recent studies have started to shed light on the importance of transporter families and channels for parasite survival. Consequently, these molecular entities constitute potential drug targets, some of which are currently being exploited 16 .

Ingredients to make a tachyzoite
Global genome-based models of T. gondii metabolism can provide an estimated minimal set of required nutrients. These flux balance models (FBMs) use gene annotations to predict the presence of metabolic pathways and use estimated parasite biomass composition and ATP consumption rates to model corresponding fluxes. To date, two FBMs of T. gondii metabolism have been generated and published: iCS382 17 and ToxoNet1 18 . The recent, more comprehensive study by Tymoshenko et al. resulted in a proposed minimal set of required nutrients for tachyzoite growth that are illustrated in Figure 2 18 . It should be emphasized that validation of nutrient dependencies is no easy feat for intracellular parasites, as they are in constant competition with the host and no apicomplexan parasite has yet been reported to replicate indefinitely in axenic media. In addition nutrients are conditionally dependent on the presence of other nutrients and environmental conditions.
In accordance with ToxoNet1 predictions, tachyzoite growth depends on successful competition for a number of co-factors and vitamins with the host cell. These small molecules fulfill vital metabolic functions in the parasite (Figure 2). We linked these nutrient-dependent pathways with the set of essential metabolic pathways as identified through FBM modelling 18 and assigned the respective "phenotype scores" of the genome-wide CRISPR (clustered regularly interspaced short palindromic repeats) screen of Sidik et al. 19 to the respective enzymes (Table S1). As expected, these pathways contain mostly essential genes with exceptions such as putative phosphatidylinositol synthetase, NAD + synthetase, and a sphingolipid desaturase, a loss of which do not cause signifcant fitness penalties 19 . This may indicate that T. gondii is able to scavenge products of these enzymes in some form; it might also reflect incorrect gene annotations or insufficient time elapsed between gene knockout and expression of a fitness defect in those experiments 19 .
The major carbon sources of T. gondii are glucose and glutamine 21 . In addition, the parasite can use exogenous acetate to elongate fatty acids if available 10,22 .
Amino acid auxotrophies of T. gondii tachyzoites have been a long recurring subject of research but, to our knowledge, have not been systematically and comprehensively tested. As is true for many unicellular pathogens, T. gondii requires exogenous tryptophan 23,24 . It is also auxotroph for arginine 25 and tyrosine 26 . Besides glutamine, a range of amino acids are imported, as shown recently by gas chromatography-coupled mass spectrometry (GC/MS). These include alanine, valine, leucine, isoleucine, proline, glycine, serine, threonine, methionine, and tyrosine 24,27 . Phenylalanine import 28 was observed by using Raman spectroscopy. Complementarily, T. gondii has been shown to be able to synthesize alanine, glutamine, and aspartate by stable isotope-resolved GC/MS 10 . Recently, several amino acid transporters have been identified as part of a transporter family that appears to be specific to apicomplexans 24 . Arginine and tyrosine transport activity has been assigned to T. gondii apicomplexan amino acid transporter 1 and 5-3 (TgApiAT1 and TgApiAT5-3), respectively 24,27,29 . Other amino acids such as histidine, lysine, branched-chain amino acids, and cysteine or methionine have been considered to be essential 18 but this remains to be experimentally verified.
In addition to proteinogenic amino acids, T. gondii tachyzoites are thought to rely on uptake of ornithine and spermine to reverse-synthesize polyamines 30 . The metabolic roles that are fulfilled by these compounds in T. gondii are still ill defined 30,31 . Red arrowheads indicate uptake of compounds that can be synthesized by mammalian cells, whereas brown arrowheads mark uptake of those substances for which both organisms are auxotrophic and thus they compete. Metabolites in black are scavenged when available but are not strictly essential whereas the pathways they are involved in are. Blue arrowheads denote the end products for which the imported metabolites are precursors or essential co-factors (or both) for their synthesis. Lipoic acid (pink) is special since it is synthesized in one compartment of the parasite (apicoplast) but apparently unable to reach another organelle (mitochondrion) where it is also required; thus, it has to be scavenged from the host 32 . 1 L-Arg; L-Trp; L-His; L-Lys (or L-2-aminoadipate 6-semialdehyde); L-Ile (or (S)-3-methyl-2-oxopentanoate); L-Val (or 3-methyl-2-oxobutanoate); L-Leu (or 4-methyl-2-oxo-pentanoate); L-Phe (or phenylpyruvate). 2 Nicotinate or nicotinate-D-ribonucleoside or nicotinamide. 3 Putrescine, spermine, ornithine. 4 D-fructose or D-glucosamine or D-glucose or D-mannose or D-ribose or D-sorbitol or 2-Deoxy-D-ribose. 5 Adenine or adenosine or guanine or guanosine or inosine or hypoxanthine or xanthine. 6 Phosphatidylserine, phosphatidylcholine, phosphatidylethanolamine, phosphatidic acid. 7 Uracil, uridine, cytidine, deoxyuridine, deoxycytidine.
Besides amino acids, purines are a major nutrient for T. gondii and required for nucleic acid synthesis (reviewed in 33). As a purine auxotroph, T. gondii imports purines in various forms through three transporters. TgAT1 is a low-affinity nucleoside transporter for adenosine and inosine 34 . Furthermore, the existence of two uncloned transporters has been postulated by detailed radiolabeling experiments. TgNBT1 is a high-affinity purine base transporter that accepts hypoxanthine, xanthine, and guanine as substrates. TgAT2 is a broad-spectrum nucleoside transporter 35 .
T. gondii imports a number of vitamins that are needed as co-factors for essential enzymes.
Biotin is a co-factor for several carboxylase reactions, including the acetyl-CoA carboxylase (ACCase), which is essential for de novo fatty acid synthesis in the apicoplast (FAS2), a relict plastid 19,36,37 (Table S1). Biotin was not mentioned as an essential co-factor in previous lists, but T. gondii appears to lack the canonical biotin synthesis pathway 38 . It is readily taken up through the host cell as indicated by its use for tagging proteins via a promiscuous ubiquitin ligase BirA 39 . The qualification of biotin as an essential vitamin for T. gondii is supported by recent data showing that a biotin ligase of Plasmodium falciparum, responsible for post-translational modification of its ACCase, is required for the development of liver stages 40 . Accordingly, we added the respective homolog of T. gondii to the list of essential genes in Table S1.
Folates take part in the synthesis of pyrimidines and amino acids and presumably are taken up through the putative TgBT1 transporter 41 and are also synthesized (reviewed in 33).
Thiamine is a co-factor for transketolases and dehydrogenases that act on pyruvate, alpha-ketoglutarate, and branched-chain keto amino acids 42 . To date, there is no biochemical evidence of thiamine uptake. Interestingly, T. gondii (M. Blume, unpublished data) and the related apicomplexan P. falciparum 43 are sensitive to the analog oxythiamine, indicating that this co-factor can be salvaged and is incorporated into enzymes.
Similarly, lipoic acid is an essential co-factor for a number of dehydrogenases and is being scavenged from the host cell in spite of being synthesized by the parasite in the apicoplast 32 ( Figure 2).
In contrast to these co-factors, tachyzoites appear unable to scavenge pantothenate, a precursor for coenzyme A, in significant amounts. Instead, they possess the entire synthesis machinery for pantothenate 44 and parasite growth is sensitive to its pharmacological inhibition. The uptake of nicotinate or nicotinamide and riboflavin has been proposed to be required for growth of T. gondii 18 ; however, to date, no biochemical evidence has been reported. Not surprisingly, FBMs also predict that T. gondii requires inorganic ions such as iron 45,46 and phosphates. Also, their uptake mechanisms remain unknown.
In addition to importing small molecules that act as building blocks, T. gondii imports proteins from the cytosol of its host cell and digests them in its endolysosomal system via cathepsin L and other proteases 12 . The underlying uptake mechanism is unclear but may involve the intravacuolar network of tubular membranes as a potential delivery route. Interestingly, a transient endocytic structure in the parasite plasma membrane that can be triggered by excess supply of oleic acid 13 has been implicated in lipid droplet uptake from the host and may also be responsible for import of host proteins.

Lipid uptake: the dose is the poison
To sustain its rapid replication, T. gondii tachyzoites need to synthesize large amounts of lipids and import precursors.
Exciting new research has established that host lipid droplets present a mobilizable source of neutral lipids and fatty acids for tachyzoites. Infected cells exhibit an increased abundance of lipid droplets 47,48 which can be further increased by exogenous oleic acid 13 . Under these conditions, T. gondii scavenges oleic acid or neutral lipids (or both) from Rab7positive host lipid droplets and stores excess lipids in deposits as triacylglycerides in its cytosol. The disruption of this storage pathway has been shown to be detrimental to tachyzoites and bradyzoites in vitro 13,49 . Under normal culture conditions, fatty acid import results in an even cellular distribution of the acquired fatty acids and confers enhanced growth 50 . Interestingly, the fusion of host mitochondria around the PV has been shown to limit access of T. gondii to host fatty acids and decrease its replication rate 50 . These studies report different intracellular distributions of internalized fatty acids that indicate distinct fates of host lipid droplet content. However, it is important to note that the employed fluorophore-linked lipids such as C4-BODIPY-C9 or BODIPY-FL-C12, though conceptually attractive, differ substantially in their biophysical properties from the lipids of interest. These include altered flip-flop rate and transfer between organelles compared with their parental lipids 51 . Regardless, these studies suggest that the parasite undertakes largely unregulated import of lipids and depends on control mechanisms from its host cell instead. There appears to be a need to fine-tune the fatty acid supply of T. gondii since fatty acid scavenging is supportive of growth only in low doses 50 and becomes harmful at higher concentrations 13 . Shielding parasites from fatty acid uptake by enwrapping their PV in host mitochondria appears to be a defense mechanism 50 . It will be important to see how general this proposed mechanism is operating since, in the low virulence type II strains, the host mitochondria do not associate with the PVM 52 .
The uptake of fatty acids is complemented by T. gondii's comprehensive capabilities to produce fatty acids on its own. GC/MS-based methods revealed that the parasite can synthesize myristic and palmitic acid de novo via the FAS2 pathway in the apicoplast and also possesses a set of elongases and the FAS1 pathway 53 . In addition, GC/MS-resolved stable isotope labelling patterns revealed salvage of unsaturated long chain and very long chain fatty acids from its host 54 .
Other precursors for major lipids include choline, which is needed for phosphatidylcholine synthesis 55 , and this pathway can be poisoned with the choline analog di-methylethanolamine 56 . Phosphatidylinositol is part of the lipidome of T. gondii but its synthase appears non-essential (Table S1). Also, its precursor myo-inositol is not synthesized from glucose 10 , indicating that T. gondii can access exogenous inositol and may also import this phospholipid. In addition, T. gondii acquires several other lipids directly from its host. These include low-density lipoprotein-derived host cholesterol for which the parasite is auxotroph (reviewed in 57) and sphingolipids from Golgiderived vesicles 55,58 .
Another important class of lipidic molecules are isoprenoids. These are implicated in a number of essential processes that include signaling, trafficking, energy metabolism, and protein translation (reviewed in 59). T. gondii synthesizes the precursors isopentenyl pyrophosphate and dimethylallyl pyrophosphate in its apicoplast via the non-mevalonate pathway 60 . These molecules are the building blocks for terpenoids via a multifunctional farnesyl diphosphate synthase (TgFPPS) 61 . Interestingly, T. gondii also scavenges products of this enzyme, including farnesyl-and geranylgeranyl-pyrophosphates from its host, making TgFPPS knockout mutants dependent on corresponding host isoprenoid synthesis 62 .

The state of bradyzoite-host interactions: initial insights into a black box
Owing to their ease in cultivation, tachyzoites are clearly the most investigated stage of T. gondii. However, the vast majority of T. gondii's lifetime in intermediate hosts is spent in the bradyzoite stage. Much less is known about how these stages interact with their host cell and which metabolic factors permit persistence. T. gondii tissue cysts are frequently found in brain, muscle, eye, and cardiac tissue 63 . It is unclear whether the parasite displays tropism toward particular cell types or whether some are more permissive for long-term residence. Interestingly, in vitro parasites display elevated spontaneous stage conversions in cell cycle-arrested fibroblasts 64 and terminally differentiated skeletal muscle cells 65 . This suggests that non-proliferating cells express an as-yet-unknown effector or provide a generally stressful environment. In this respect, it is noteworthy that proliferating cells maintain a glycolytic metabolic profile, with large pools of glycolytic intermediates and high rates of lactic acid production, to support their biosynthetic activities 66 . This environment may equally favor the biosynthetic activity of growing T. gondii tachyzoites. Consistently, host cells that produce high levels of lactic acid promote tachyzoite growth over bradyzoite formation 67 . The induction of bradyzoites in vitro is considered to work by slowing down tachyzoite growth via a range of stressors, such as limitation of arginine and bicarbonate as well as basic and acidic pH stress 63 . The resulting encysted bradyzoites adopt a spectrum of cell division rates 68 that may introduce a degree of flexibility in their need for pyrimidines for nucleic acid synthesis and consequently contribute to their resistance to antifolates such as pyrimethamine. However, bradyzoites appear sensitive to inhibition of nutrient storage and turnover pathways including autophagy 69 , lipid storage 70 , and carbohydrate storage 71 . These findings may reflect the fact that T. gondii deals with oversupply of nutrients by storage rather than by regulating respective import pathways. In particular, slowly dividing bradyzoites appear to depend on the integrity of the regulation of these storage pathways to avoid detrimental accumulation of material.

Future methods to study host-parasite interactions
As outlined above, comprehensive metabolomic approaches were instrumental in defining metabolic interactions 72 as well as synthesis and import capabilities of fatty acids and amino acids in tachyzoites 10,24 . A similar approach with bradyzoites requires the development of efficient methods for the in vitro culture and purification of tissue cysts in their natural metabolic state. Strategic introduction of stable isotopes into parasite polymers such as DNA and proteins might be helpful to measure the activity of metabolic pathways and decrease the number of cysts needed 73 .
Comprehensive liquid chromatography-coupled mass spectrometry (LC/MS)-based metabolomic approaches will also help to define which amino acid synthesis pathways are active in T. gondii. By measuring the levels and turnover of intermediates directly, metabolomics can bridge the gap that exists between annotated gene sets and a functional pathway. Such associations are very indirect since function will depend on regulatory mechanisms at the transcriptomic, proteomic, and metabolomic levels 74 . In addition, the accuracy of functional predictions based on homology are limited. For instance, a putative lysine decarboxylase implicated in lysine degradation has been shown to act as a glutamate decarboxylase within the GABA shunt instead 10 .
Similar limitations apply to conclusions based on transcriptomic and proteomic data. While these approaches are powerful to functionally associate genes and resolve differences of physiological states of the parasite under different conditions, such as host environments 64 and stage conversion states 75,76 , functional implications on the phenotype level and mechanistic insights are constrained by functional gene annotations. Combined multi-omics approaches might provide insights beyond the metabolic dimension of host-parasite interactions in the future 77 .
In addition to the use of metabolomic techniques to study parasite-host interactions, the use of more sophisticated cellular systems such as neurons and myotubes, organoids 77 , or mice 78 holds great potential. This is illustrated by the fact that T. gondii and coccidian-specific genes are the least fitnessconferring ones for the survival in fibroblasts 19 .
In conclusion, there remain many open questions about how parasite and host metabolism interact, in particular with respect to dormant bradyzoites that are key for transmission and global prevalence of T. gondii. Future metabolomic approaches will be instrumental in further dissecting the host-parasite interface and discovering ways to efficiently target both tachyzoite and bradyzoite forms of T. gondii.

Abbreviations
ACCase, acetyl-CoA carboxylase; FAS2, fatty acid synthesis in the apicoplast; FBM, flux balance model; GC/MS, gas chromatography-coupled mass spectrometry; PV, parasitophorous vacuole; PVM, parasitophorous vacuole membrane; TgFPPS, farnesyl diphosphate synthase from Toxoplasma gondii The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Open Peer Review Current Referee Status:
Editorial Note on the Review Process are commissioned from members of the prestigious and are edited as a F1000 Faculty Reviews F1000 Faculty service to readers. In order to make these reviews as comprehensive and accessible as possible, the referees provide input before publication and only the final, revised version is published. The referees who approved the final version are listed with their names and affiliations but without their reports on earlier versions (any comments will already have been addressed in the published version).
The referees who approved this article are:

Version 1
The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias You can publish traditional articles, null/negative results, case reports, data notes and more The peer review process is transparent and collaborative