The Toxoplasma oxygen-sensing protein, TgPhyA, is required for resistance to interferon gamma-mediated nutritional immunity in mice

As Toxoplasma gondii disseminates through its host, the parasite must sense and adapt to its environment and scavenge nutrients. Oxygen (O2) is one such environmental factor and cytoplasmic prolyl 4-hydroxylases (PHDs) are evolutionarily conserved O2 cellular sensing proteins that regulate responses to changes in O2 availability. Toxoplasma expresses 2 PHDs. One of them, TgPHYa hydroxylates SKP1, a subunit of the SCF-E3 ubiquitin ligase complex. In vitro, TgPHYa is important for growth at low O2 levels. However, studies have yet to examine the role that TgPHYa or any other pathogen-encoded PHD plays in virulence and disease. Using a type II ME49 Toxoplasma TgPHYa knockout, we report that TgPHYa is important for Toxoplasma virulence and brain cyst formation in mice. We further find that while TgPHYa mutant parasites can establish an infection in the gut, they are unable to efficiently disseminate to peripheral tissues because the mutant parasites are unable to survive within recruited immune cells. Since this phenotype was abrogated in IFNγ knockout mice, we studied how TgPHYa mediates survival in IFNγ-treated cells. We find that TgPHYa is not required for release of parasite-encoded effectors into host cells that neutralize anti-parasitic processes induced by IFNγ. In contrast, we find that TgPHYa is required for the parasite to scavenge tryptophan, which is an amino acid whose levels are decreased after IFNγ up-regulates the tryptophan-catabolizing enzyme, indoleamine dioxygenase (IDO). We further find, relative to wild-type mice, that IDO knockout mice display increased morbidity when infected with TgPHYa knockout parasites. Together, these data identify the first parasite mechanism for evading IFNγ-induced nutritional immunity and highlight a novel role that oxygen-sensing proteins play in pathogen growth and virulence.


Introduction
Toxoplasma gondii is an intracellular apicomplexan parasite that chronically infects up to onethird of the world's population [1].This opportunistic pathogen causes toxoplasmosis, a disease that can be life-threatening in immunocompromised individuals and developing fetuses.Infection in humans follows consumption of meat harboring bradyzoite-containing tissue cysts or water and produce contaminated with sporozoite-containing oocysts [2].Gastric enzymes rupture cysts or oocysts and the released parasites go on to infect the small intestine.At this juncture, the parasites convert into replicative tachyzoites and trigger an immune response that results in recruitment of innate immune cells such as inflammatory monocytes, innate lymphoid cells, and neutrophils.These cells are in turn infected and used to disseminate to peripheral tissues where tachyzoites differentiate into bradyzoites and form tissue cysts that can persist for the host's lifetime [3].
As Toxoplasma traffics through a host, it encounters diverse environments that vary in essential nutrients and other compounds.O 2 is one such compound that all aerobic cells must sense and respond to changes in its levels to optimize growth and viability [4].Cytoplasmic prolyl 4-hydroxylases (PHDs) are dioxygenases that use α-ketoglutarate and O 2 as substrates and function as the key cellular O 2 sensors [5].PHDs hydroxylate proline residues in substrates and in metazoans the best-recognized PHD substrate is the α subunit of the hypoxiainducible transcription factor (HIFα).Prolyl-hydroxylated HIFα is ubiquitinated by the Von Hippel-Lindau ubiquitin ligase complex, which targets it for proteasomal degradation [6].Under hypoxic conditions, HIFα is stabilized and able to regulate the expression of genes important for growth and survival at low O 2 .While HIFα is not conserved in protist genomes, PHDs are, indicating that protists likely respond differently to changes in O 2 availability [4,7].The best characterized protozoan PHD is DdPhyA from Dictyostelium discoideum that modifies a proline in DdSkp1, which is an adaptor protein in the SCF (Skp1/Cullin/F-box protein) E3 polyubiquitin ligase complex [8].The resulting hydroxyproline is subsequently modified by a series of glycosyltransferases that alters the binding affinity of SKP1 towards different F-Box proteins and thus differential protein ubiquitination by the SCF complex [8][9][10].Two PHDs have been identified in Toxoplasma and one of these, TgPHYa also modifies SKP1 [11,12].While TgPHYa is not essential, TgPHYa knockouts in type 1 strain parasites display reduced in vitro growth at low O 2 [12].But TgPHYa's function in vivo remains to be determined.
Interferon gamma (IFNγ) is a cytokine critically required for host resistance to Toxoplasma [13,14].It acts by triggering the expression of IFNγ-responsive genes whose products are involved in oxygen radical generation and degradation of the parasitophorous vacuole membrane via the activity of immunity-related guanosine triphosphatases (IRGs) and guanylatebinding proteins (GBPs) [15].In addition, IFNγ induces the expression of indoleamine-pyrrole 2,3-dioxygenase (IDO) that catabolizes tryptophan, which Toxoplasma must scavenge from its host [16].The importance of IDO in nutritional immunity to Toxoplasma has remained enigmatic as it is important for resistance in some human cells but appears to be dispensable in mice [16][17][18].The most likely reason for this is that, unlike humans, mice express 2 IDO isoforms.Thus, it remains unknown whether tryptophan scavenging/utilization is a viable anti-Toxoplasma drug target [18].
Here, we report that TgPHYa deletion in a cystogenic strain leads to decreased virulence in both oral and intraperitoneal infection models.Furthermore, TgPHYa is required for resistance to IFNγ-mediated killing and does so by evading IFNγ-inhibition of tryptophan utilization.Significantly, TgPHYa knockout parasites display increased virulence in IDO1 knockout mice.Taken together, these data reveal that a parasite-encoded cytoplasmically localized prolyl hydroxylase mediates Toxoplasma evasion of a protective host nutritional immune response pathway.

Generation of TgPHYa knockout in the type II ME49 strain
To assess a role for TgPHYa during in vivo infection, CRISPR was used to generate a TgPHYa knockout mutant in a type II Toxoplasma strain that also expresses red fluorescent protein (RFP) (TgPHYaKO II ).PCR analysis indicated that TgPHYa was properly targeted by the CRISPR construct and that the TgPHYa gene was disrupted (Fig 1A and 1B).Because TgPHYa antisera is lacking, we assessed whether the gene disruption affected TgPHYa protein by analyzing TgSKP1 mobility by SDS-PAGE.We observed that the apparent molecular weight of TgSKP1 was reduced in TgPHYaKO II parasites indicating decreased TgSKP1 prolyl  hydroxylation/glycosylation (Fig 1C) [12].Similar to the type I strain TgPHYa knockout [12], in vitro TgPHYaKO II growth was slightly decreased at 21% O 2 but significantly at 0.5% O 2 (Fig 1D and 1E).Next, brain cyst burdens were enumerated in mice 30-, 60-, or 90-days post infection (dpi) with 50 cysts.We found that at 30 dpi, there was a statistically significant decrease in numbers of cysts in brains of TgPHYaKO II infected mice (Fig 2C).But, at later time points differences were not noted.Brain cyst sizes were also analyzed and at each time point the TgPHYaKO II cysts were significantly smaller than the parental strain cysts (Fig 2D).To test whether TgPHYaKO II brain cyst phenotypes were a result of the prolyl hydroxylase functioning in cyst formation, cyst formation was examined in parasites exposed to alkaline medium (pH 8.2) at ambient CO 2 [19].After 5 days of induction, cysts were examined using DBA-FITC and no differences in the ability for TgPHYaKO II to form in vitro cysts were noted (Fig 2E).Together, these data indicate that TgPHYa contributes to virulence and brain cyst development.

TgPHYa is important for parasite dissemination
One possible explanation for decreased numbers of TgPHYaKO II brain cysts at 30 dpi is delayed dissemination from the gut to the brain and other tissues.To examine this, we first assessed a role for TgPHYa in establishing an infection in the gut by gavage infecting mice with 50 WT or TgPHYaKO II cysts and 7 days later enumerating parasite burdens by quantitative real time PCR (qPCR) in different segments of the small intestine.No significant differences in parasite numbers were observed throughout the intestine (Fig 3A).We also assessed intestinal inflammation by histological analysis at 7 dpi (Fig 3B) and 12 dpi (Fig 3C ) and no significant differences were noted at either time point.
Next, parasite burdens in peripheral tissues at 7 and 12 dpi were determined by qPCR.Except for the spleen, there were no remarkable differences in parasite burdens in the organs on day 7 of infection (Fig 3D).In contrast, significantly fewer TgPHYaKO II parasites were observed in the spleens, hearts, and brains of the infected mice at 12 dpi.Parasite burdens in lungs were also lower although this difference was not statistically significant (P = 0.16).These data indicate that while TgPHYaKO II parasites can establish an infection in the intestine, they are unable to disseminate efficiently to peripheral tissues.

TgPHYaKO II burdens are reduced in inflammatory monocytes and neutrophils
Toxoplasma dissemination within and from the gut is dependent on parasite infection and survival within inflammatory monocytes, neutrophils, and CD103 + dendritic cells [3].To assess whether TgPHYa was required for Toxoplasma interactions with those cells, we first examined by flow cytometry the numbers of each cell type in intestines of mice 7 days after they were gavage-infected with 50 WT or TgPHYaKO II tissue cysts.Relative to uninfected mice, increased numbers of inflammatory monocytes and neutrophils, but not CD103 + dendritic cells, were noted in intestines of mice infected with either WT and TgPHYaKO II parasites but differences between the two were not significantly different (Fig 4A and 4B).
The infection status of these cells was assessed by flow cytometric detection of parasiteexpressed RFP.First, we compared total numbers of RFP + cells and found significantly decreased numbers and percentages of TgPHYaKO II -infected intestinal cells (Fig 4C).Using the same gating strategies as described above to identify cell populations, we noted a significant decrease in numbers and percentages of TgPHYaKO II -infected inflammatory monocytes and neutrophils (Fig 4D and 4E).Differences between the qPCR data in Fig 3 and the flow cytometry data here likely reflect that the latter is gated specifically on live infected cells.
To determine if these results were specific to the intestinal environment, we intraperitoneally infected mice with 10 4 WT or TgPHYaKO II tachyzoites.After 7 days, peritoneal exudates were collected and total numbers and infection status of recruited immune cells were (C) post infection.Asterisks highly inflammatory infiltrates in the villi and LP and arrowheads highlight destruction of villi.Shown are representative images and plots represent mean ± SEM of histological scores of 50 randomly selected section per mouse.(D) Parasite burdens were determined by qPCR of genomic DNA isolated from spleens, lungs, hearts, and brains of mice gavage infected with 50 of ME49 WT or TgPHYaKO II cysts.Data shown are mean ± SEM from a total of 7 to 9 mice from 3 independent experiments (ns = not significant, *P < 0.01, ***P < 0.001, using oneway ANOVA).The data underlying the graphs in this figure can be found in S2

TgPHYa is required for resistance to IFNγ
Because IFNγ is critically required for host resistance to Toxoplasma [14], survival of WT and IFNγ -/-mice gavage-infected with 50 ME49 WT or TgPHYaKO II cysts was monitored for 30 days.Infection with either parasite strain led to rapid death of IFNγ -/-mice (Fig 6A).Loss of IFNγ also led to comparable numbers of infected peritoneal exudate cells harvested from mice IP infected with either strain although we did observe fewer percentage of cells were infected in the TgPHYaKO II -infected IFNγ -/-mice (Fig 6B).We next compared the peritoneal immune cell infiltrate between WT and IFNγ -/-mice and found significantly higher numbers of inflammatory monocytes in IFNγ -/-mice infected with TgPHYaKO II .Moreover, loss of IFNγ led to increased recruitment of neutrophils in mice infected with either WT or TgPHYaKO II parasites (Fig 6C).Finally, we found that IFNγ restricted growth of TgPHYaKO II parasites in both inflammatory monocytes and neutrophils (Fig 6D and 6E) and while numbers of infected cells were similar for both inflammatory monocytes and neutrophils, the percentages of either were lower in TgPHYaKO II -infected IFNγ -/-mice.These data indicate that a parasite-encoded cytoplasmically localized prolyl hydroxylase, TgPHYa, is important for Toxoplasma to survive immune killing by IFNγ.Natural killer (NK) cells are critical for early resistance to Toxoplasma infection vial the release of IFNγ and other cytokines [22].We found that both WT and TgPHYaKO II strain parasites were able to stimulate IFNγ expression in NK cells to similar levels (S2A Fig) .We next asked whether NK cells as an innate source of IFNγ were important for controlling TgPHYa-KO II parasites.Thus, NK cell-depleted mice (or those treated with irrelevant IgG) were IP infected.The peritoneal exudate was collected 5 days later and we found that numbers of infected cells were increased in NK cell-depleted mice infected with both WT and TgPHYa-KO II strains (S2B and S2C Fig).

Toxoplasma growth and virulence is dependent on TgPHYa expression levels
Two TgPHYaKO II complementation constructs were generated in which the TgPHYa ORF with a 3X-hemagluttin (HA) C-terminal epitope tag was cloned downstream of either the strong TUB1 promoter (TgPHYaKO II :PHYa TUB ) or its endogenous promoter (1,000 bp upstream of its predicted transcription started site; TgPHYaKO II :PHYa PHYa ) (S3A and S3B Fig) .We confirmed in TgPHYaKO II :PHYa PHYa and TgPHYaKO II :PHYa TUB parasites by RT-qPCR expression of the transgene to be low and high, respectively (not shown).TgPHYa-KO II :PHYa TUB protein levels were readily detected by western blotting while TgPHYaKO II : PHYa PHYa was not detectable, even by immunoprecipitation (Fig 7A).To assess transgene function, SKP1 mobility was assessed by western blotting and in both strains SKP1 was found to migrate at a higher molecular weight than it did in TgPHYaKO II lysates indicating that SKP1 was modified as expected (Fig 7A).We also noted that SKP1 protein levels were reduced in the knockout strain suggesting that TgPHYa not only modifies SKP1 but also may affect its abundance.This decrease in SKP1 abundance was not observed in type I strains with TgPHYa knockout mutations [12,23] and at this point it is not possible to predict how altered SKP1 levels would impact SCF-E3 activity and parasite growth.
To assess complementation of TgPHYaKO II phenotypes, we first examined in vitro growth at 0.5% O 2 .We observed that the TgPHYaKO II growth defect was restored in TgPHYaKO II : PHYa PHYa but not TgPHYaKO II :PHYa TUB parasites (Fig 7B).We assessed the importance of TgPHYa expression levels on virulence by first IP infecting WT mice with 10 4 parasites of each strain and found that TgPHYaKO II :PHYa PHYa was as virulent as WT strain parasites and the TgPHYaKO II :PHYa TUB strain was less virulent (Fig 7C) even when mice were IP infected with 10 6 tachyzoites (S4A and S4B Fig) .Similarly, parasite growth within peritoneal cells was partially restored by TgPHYaKO II :PHYa PHYa but not TgPHYaKO II :PHYa TUB at either infectious dose (Figs 7D and S4B).However, both strains were similarly virulent in IFNγ -/-mice (Fig 7C).
Next, mice were intraperitoneally infected with 500 tachyzoites of each strain and 30 days later, brain cysts were enumerated.Our results showed that while cyst numbers were partially restored in mice infected with TgPHYaKO II :PHYa PHYa parasites, brain cysts were undetectable in mice infected with up to 10 3 TgPHYaKO II :PHYa TUB parasites (Fig 7E).

TgPHYa is important to evade IFNγ-dependent nutritional immunity
IFNγ kills Toxoplasma by up-regulating the expression of proteins that degrade the parasitophorous vacuole membrane, increase reactive oxygen species, and limit essential nutrient (e.g., tryptophan) availability [24].In response, Toxoplasma evades IFNγ killing mechanisms via the injection of parasite-encoded effector proteins that act to either prevent IFNγ from up-regulating IFNγ-stimulated genes (if Toxoplasma infects a cell before exposure to IFNγ) or inhibit the gene products' activities (if Toxoplasma infects a cell after its exposure to IFNγ) [25].To assess whether TgPHYa interferes with the ability for Toxoplasma to inhibit IFNγ-stimulated gene expression, HFFs were infected with WT, TgPHYaKO II , or the complemented strains and then cultures were treated ± IFNγ at 24 h post infection.The cultures were grown for 12 d, fixed, stained with crystal violet, and plaques numbers and sizes measured.We observed that TgPHYa was dispensable for Toxoplasma resistance to IFNγ-dependent vacuole elimination under these conditions although we did note that IFNγ had a slight but reproducible effect on the size TgPHYaKO II :PHYa PHYa plaques (Figs 8A and S5A).TgIST is a dense granule-localized, secreted parasite effector protein that inhibits expression of IFNγ-stimulated genes, such as IRF1, by preventing STAT1 from binding to and activating the promoters of these genes [26,27].HFFs were therefore infected with WT, TgPHYaKO II , TgPHYaKO II :PHYa PHYa , or TgPHYaKO II :PHYa TUB tachyzoites and 24 h later treated with IFNγ or PBS as a vehicle control.The cells were fixed 12 h later and stained to detect IRF1.We found that IRF1 could not be detected in parasite-infected cells but could be detected in neighboring, uninfected cells (Fig 8B).Given that secretion of dense granule-localized effectors such as TgIST utilize a common translocon [28], these data indicate that secretion of these types of effectors is apparently not dependent on TgPHYa and that TgPHYa has no obvious role in TgIST-dependent silencing of IFNγ-induced host cell gene expression.
IFNγ increase levels of oxygen radicals such as hydrogen peroxide and nitric oxide.We therefore tested whether TgPHYa was important for resistance by first assessing plaque formation when parasite were grown in the presence of 0.2 mM H 2 O 2 , which we and others found to be the approximate EC 50 for killing both intracellular and extracellular tachyzoites [11,29].We found that loss of TgPHYa did not impact parasite susceptibility to H 2 O 2 (S6A Fig) .In addition, we found that TgPHYa also was not apparently involved in parasite resistance to NO as parasite growth was not impacted by the addition of the NO donor, sodium nitroprusside (SNP) (S6B Fig) .Together, these data suggest that TgPHYa is dispensable for effector protein secretion or resistance to reactive O 2 species.
Tryptophan is an essential amino acid for Toxoplasma and IFNγ limits available pools of tryptophan for parasites to scavenge as a result of up-regulating the expression of the tryptophan catabolizing enzyme, indoleamine 2,3 dioxygenase (IDO) [16].To test whether TgPHYa is important for tryptophan scavenging in IFNγ-treated cells, HFFs were stimulated with IFNγ for 24 h (or PBS as a vehicle control) and then infected with WT, TgPHYa-deficient, or TgPHYaKO II complemented tachyzoites.The cells were then grown in complete medium or medium supplemented with increasing levels of tryptophan.As expected, growth of all strains was prevented by IFNγ pretreatment ("+IFNγ" samples in S5B and S5C Fig) .When the medium was supplemented with 16 or 32 μg/ml of tryptophan, the numbers of plaques formed were restored for all strains (S5B Fig) as were the sizes of the plaques formed by WT and TgPHYaKO II :PHYa PHYa parasites (S5C Fig) .In contrast, significantly smaller plaques were formed by knockout and TgPHYaKO II :PHYa TUB parasites in tryptophan supplemented medium (S5C Fig) .Importantly, we noted in these plaquing assays that even though the size of TgPHYaKO II plaques were smaller, there was no difference in the numbers of plaques indicating that TgPHYa is not involved in avoiding IFNγ-triggered xenophagic elimination of parasitophorous vacuoles [30,31]

(S5B Fig).
To assess parasite replication more directly, HFFs plated on coverslips were incubated overnight with IFNγ and then infected for 24 or 48 h ± 32 μg/ml of tryptophan.The cells were then fixed, stained to detect the surface antigen SAG1, and numbers of parasites per vacuole counted.The data revealed that growth of WT and TgPHYaKO II :PHYa PHYa parasites were restored by the addition of exogenous tryptophan but growth of either TgPHYaKO II or TgPHYaKO II :PHYa TUB was not (Figs 8C and S5E).To test if TgPHYa is also important for parasite growth under tryptophan restricted conditions in other cell types, we repeated the replication assays using murine bone marrow-derived macrophages (BMDMs) and similarly observed that only WT and TgPHYaKO II :PHYa PHYa parasite growth was restored when exogenous tryptophan was added to IFNγ-treated BMDM (Figs 8D and S5F).The data that loss of TgPHYa, an enzyme localized to the parasite's cytoplasm, does not impact effector secretion suggest that its role in survival under tryptophan depleted conditions is most likely in efficient uptake and/or utilization of the amino acid (not shown).We were precluded from directly assessing amino acid uptake as we could not detect radiolabeled tryptophan uptake by the parasites and to our knowledge this has not been previously reported.Alternatively, we hypothesized that growth of TgPHYaKO II parasites would be increased by elevating the pools of tryptophan within the host cell.We therefore pretreated HFFs with IFNγ, infected them in the absence or presence of the IDO inhibitors, 1-MT or epacadostat [32,33], and counted parasites per vacuole 48 h later.We found that growth of WT, TgPHYa-KO II , and the complemented strains was restored by the inhibitors (Fig 9A).
Next, we tested whether virulence of TgPHYaKO II parasites would be increased in IDO1 knockout mice since loss of IDO1 would result in increased levels of tryptophan available for parasites to scavenge.Thus, we assessed whether TgPHYa had an in vivo role in tryptophan utilization by comparing survival, weight loss, and numbers of brain cysts of WT and IDO1 -/- mice infected IP with parental or TgPHYaKO II parasites.While differences in survival were not observed between the infected mice (Fig 9B), IDO −/− mice infected with the TgPHYaKO II had significantly increased morbidity as revealed by greater weight loss and increased brain cyst burdens in the IDO -/-mice (Fig 9C and 9D).

Discussion
Hypoxic responses by pathogens and their hosts are important determinants for infection outcomes.Host immune responses are highly dependent on HIF-1-dependent transcription that is triggered by the hypoxic environment of infection foci [34][35][36].Conversely, bacterial and fungal pathogens utilize O 2 as a signaling molecule to regulate biofilm formation, virulence factor expression, and host cell interactions [37][38][39][40][41].In contrast to bacterial pathogens, less is known about the role of O 2 sensing in protozoan virulence [4].Toxoplasma expresses 2 oxygen sensing proteins-TgPHYb is important for growth at high O 2 [11] and TgPHYa is important for growth at low O 2 [12].Yet, the role for either PHD in virulence was unknown.In this study, we generated a TgPHYa knockout mutant in the cystogenic type II ME49 strain background and demonstrated that it is important for virulence and brain cyst formation.While TgPHYaKO II mutant parasites can establish an infection in the gut, they are unable to efficiently disseminate to peripheral tissues because TgPHYa is required to resist IFNγ-triggered killing and does so by overcoming IDO scavenging of tryptophan.
These data represent, to our knowledge, the first examples of an enzyme localized within the parasite cytoplasm that mediates resistance to IFNγ and of a PHD functioning as a virulence factor.One possible mechanism for how TgPHYa functions is that a proline in a tryptophan transporter is a TgPHYa substrate.But, we believe that is unlikely since growth defects are similar in vitro between parasites with knockout mutations in TgPHYa and TgGNT1, which is the first of a series of glycosyltransferases that modify prolyl hydroxylated TgSKP1 to generate a novel pentasaccharide knockout parasites [23].Addition of this pentasaccharide is Graph represents the averages ± SD of 3 independent experiments.(C) HFFs were pretreated with IFNγ (1 U/ml) for 24 and then infected with tachyzoites of the indicated strain.The cells were grown for 24 and 48 h in the presence of tryptophan (32 μg/ml) and then the monolayers were fixed and numbers of parasites per vacuole determined by IFA staining.(D) BMDMs were treated with or without IFNγ (1 U/ml) for 24 h and then infected with tachyzoites of the indicated strain for 24 h and 48 h in the absence or presence of 32 μg/ ml tryptophan.The cells were fixed and numbers of parasite/vacuole determined by IFA staining.Shown are the means ± SEM, pooled from 3 independent experiments; ns = not significant, **P < 0.01, using a multiple comparison two-way ANOVA test.The data underlying the graphs in this figure can be found in S2 thought to alter the F-box protein repertoire that associates with the SCF-E3, which would impact SCF-E3-dependent protein ubiquitination [9].Thus, we favor a model where TgPHYa alters SCF-E3-dependent ubiquitination of a tryptophan transporter or a transporter-associated protein that impacts the transporter's stability, localization, and/or activity.
We found that TgPHYa overexpression has a dominant negative effect on IFNγ-controlled parasite virulence during the acute phase of the infection.This phenotype is reminiscent of the dominant negative effect that DdPHYa has on O 2 -dependent culmination during Dictyostelium development [42].Since SKP1 was fully hydroxylated in both TgPHYaKO II :PHYa PHYa and TgPHYaKO II :PHYa TUB parasites, we do not believe that overexpression impacts TgPHYa activity.Rather, TgPHYa overexpression likely leads to aberrant complex formation that acts to sequester TgPHYa substrates and/or interacting proteins such as SKP1.Regardless of mechanism, the finding that TgPHYaKO II :PHYa TUB parasites were virulent in IFNγ -/-mice was surprising given its highly attenuated growth in tissue culture and avirulence in wild-type mice.This likely is due to the diverse repertoire of Toxoplasma virulence factors that target distinct host resistance pathways [25].
We also observed that regardless of IFNγ expression, growth of wild-type parasites was greater than that of TgPHYa knockout parasite growth.This could be due to TgPHYa also contributing to evasion of IFNγ-independent host resistance mechanisms such as TNFα [43] and IL-6 [44,45].Alternatively, TgPHYa may be important for the efficient utilization of other essential nutrients such as arginine.These hypotheses are not necessarily mutually exclusive since inducible nitric oxide synthase (iNOS), whose expression is regulated by cytokines such as IFNγ, TNFα, and IL-6 [44,46], consumes arginine as a substrate.Earlier work by Denkers and colleagues demonstrated that the Toxoplasma virulence factor ROP16 impacts parasite growth by regulating iNOS expression and subsequent arginine consumption [47].
The role of IDO in host resistance to Toxoplasma has been enigmatic.Early studies revealed that IFNγ-dependent induction of IDO1 expression inhibited parasite growth in human fibroblasts [16,48,49].In addition, pharmacological inhibition of IDO enzymatic activity leads to increased mortality in Toxoplasma-infected mice [18].But, IDO1 knockout mice did not display increased susceptibility to Toxoplasma infection and this is likely due to functional redundancy with a second IDO isoform, IDO2 [18,50] or intrinsic differences in IDO enzymatic activity between mice and humans [51].Our data that TgPHYa knockout parasites are more virulent in IDO1-deficient mice suggests that IDO1 is activated in parasite-infected mice but that it is unable to degrade tryptophan to levels sufficient to inhibit parasite growth.Moreover, we demonstrated that increasing cellular tryptophan pools within the host cell via IDO inhibition led restored growth of TgPHYaKO II parasites.Since TgPHYa is an intracellular enzyme and TgPHYaKO II parasites have no apparent defects in secretion of parasite effectors into the host cell, these data together lead us to hypothesize that loss of TgPHYa leads a defect in tryptophan uptake that is exacerbated under conditions of limited tryptophan availability.This is reminiscent of increased virulence of IDO −/− mice infected with Francisella tularensis mutants with defects in tryptophan biosynthesis [52].And while our data most likely indicate a role in amino acid uptake, we cannot rule out that TgPHYa functions in other pathways such as tryptophan catabolism or inefficient tRNA charging.Our future work will address these questions.
In summary, our data reveal a novel role for a pathogen-encoded cytosolic enzyme mediating resistance to IFNγ-tryptophan utilization efficiency.They also suggest that dual inhibition of host IDO and parasite tryptophan utilization represents a novel drug target against Toxoplasma and other tryptophan auxotrophic intracellular pathogens such as Chlamydia spp and Trypanosoma cruzi [53,54].

Ethics statement
Mice were treated in compliance with the guidelines set by the Institutional Animal Care and Use Committees of the University at Buffalo (protocol #PROTO202200022:) and the University of Wyoming (protocol # 20190201JG00344-03).All experiments were carried out in accordance with Public Health Service policy on the humane care and use of laboratory animals and AAALAC accreditation guidelines.

Cells and parasites
The Toxoplasma gondii type II strain ME49 expressing red fluorescent protein and other strains generated here were maintained by passage in human foreskin fibroblasts (HFFs) (from American Type Tissue Culture, Reston, Virginia, United States of America) in complete medium (Dulbecco's modified Eagle medium supplemented with 10% heat-inactivated fetal calf serum, 1% L-glutamine, and 1% penicillin-streptomycin) as previously described [23].Cultures are routinely tested for Mycoplasma contamination (Lonza; Basel, Switzerland) and found to be negative.
To complement TgPHYaKO II parasites, a PHYA-3×HA complementation plasmid with TgPHYa expressed under the Toxoplasma tubulin (TgTUB) promoter was generated by modifying the UPRT Vha1 cDNA shuttle vector containing TgVhaI cDNA [57] using the NEB HiFi Builder method.The vector backbone, containing 973 nt of 5 0 -and 993 nt of 3 0 -UPRT flanking sequences, TgTUB promoter, and a 3×HA (C-terminal) sequence, was PCR amplified from the shuttle vector using primers G and H.The coding sequence of TgPHYa was PCR amplified from pET15b-TgphyA [12] using primers I and J. PCR fragments were gel-purified and incubated with HiFi DNA assembly enzyme mix (NEB) and transfected into E. coli Top10 cells (Thermo Fisher; Waltham, Massachusetts, USA) to yield pUPRTPHYaKO II :TgPHYa TUB and sequence verified using primers I and J. TgPHYaKO II parasites were transfected by co-electroporation with a PCR amplicon (generated using primers U and V) from pUPRTPHYaKO II : TgPHYa TUB (1 μg) and pSAG1::CAS9-U6::sgUPRT (10 μg), whose guide RNA targets the UPRT locus (S1B Fig) .Transfectants were selected in the presence of 10 μm fluorodeoxyuridine (FUdR, Sigma), and drug-resistant clones were screened by PCR with primers listed in S1 Table .A second complementation construct was generated in which TgPHYa is expressed under the control of 1 kb of sequence upstream of the TgPHYa start codon (pUPRTPHYaKO II : TgPHYa PHYa ).The upstream TgPHYa DNA fragment was generated by PCR of genomic DNA from ME49-RFP parasites using primers K and L, and the vector DNA fragment was generated from pUPRTPHYaKO II :TgPHYa TUB using primers M and N, which excluded the tubulin promoter sequence.Transfection, selection, and validation of were performed as described above.

Western blotting
Tachyzoites were harvested from HFFs, counted with a hemacytometer, pelleted by centrifugation at 2,000 × g for 8 min at 4˚C, and lysed in boiling SDS-PAGE sample buffer containing 5% 2-mercaptoethanol.Lysates from equivalent cell numbers were separated on a 4% to 12% gradient SDS-PAGE gel (Invitrogen; Carlsbad, California, USA), transferred to nitrocellulose membranes that were blocked with 5% milk in TBS, and incubated with primary and secondary antibodies (S1 TableAU : Table1iscitedinthesentencesLysatesfromequivalentcellnumberswere:::andC ).Blots were visualized and analyzed using a Li-Cor Odyssey scanner and software (Li-Cor; Lincoln, Nebraska, USA).
For SKP1 western blots, extracts were prepared by vortexing parasites in 1% SDS, 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, boiling for 5 0 , and centrifugation for 15 0 at 21,000 × g.For reduction and alkylation [58], the supernatant was diluted with 4× sample buffer (0.9 M Tris (pH 8.45), 24% glycerol, 8% SDS, 0.01% phenol red, 0.01% Coomassie G) to which DTT was added (10 mM final concentration) and incubated at 65˚C for 30 0 .Iodoacetamide was then added to a final concentration of 25 mM and incubated at room temperature in the dark for 1 h.Samples were electrophoretically separated on an Invitrogen NuPAGE 4% to 12% Bis-Tris gel (Thermo Fisher) in MES-SDS buffer (pH 7.3) (Thermo Fisher), transferred to nitrocellulose membrane, and western blotted as above.

Parasite growth assays
Plaquing assays were performed essentially as described [23].Briefly, tachyzoites were added to confluent HFF monolayers and grown undisturbed for 12 days.Cells were fixed with icecold methanol, stained with crystal violet, and plaques counted.Plaque areas were measured using ImageJ software (https://imagej.nih.gov/ij/).When assessing IFNγ/tryptophan supplementation, HFFs were pretreated for 24 h with 1 U/ml IFNγ (Sigma; St. Louis, Missouri, USA) in media prepared using amino acid free DMEM (USBiological; Salem, Massachusetts, USA) supplemented with 3% dialyzed heat-inactivated fetal bovine serum (FBS), individual essential amino acids, and indicated concentrations of tryptophan.In vitro bradyzoite development was assessed as described [59].Briefly, parasites were added to confluent HFFs seeded on glass coverslips in pH 8.1 medium and grown for 72 h at 37˚C at ambient CO 2 .Coverslips were fixed and cysts detected using Dolichos Biflorus Agglutinin lectin (Vector Lab, Burlingame, California, USA) and tachyzoites with anti-SAG1 antisera; 1-MT and SNP were purchased from SIGMA and epacadostat from Selleckchem (Houston, Texas, USA).
To assess growth in murine BMDMs, bone marrow cells isolated from the tibia and femur of C57BL/6 J mice were cultured in DMEM, 10% FBS, 1% L-glutamine, 1% penicillin/streptomycin, and 20% L929 conditioned medium for 5 to 7 days.BMDMs were then grown on coverslips in 24-well plates and stimulated with 1 U/ml murine IFNγ (PeproTech, Cat#315-05) for 24 h pre infection with or without tryptophan as indicated.Both naïve BMDMs and IFNγactivated BMDMs were infected at an MOI of 1.After 24 or 48 h, coverslips were fixed with 4% paraformaldehyde in PBS for 20 min at room temperature, blocked in 5% bovine serum albumin in PBS for 60 min, parasites detected by immunofluorescence staining with anti-SAG1 antibodies and Alexa Fluor 594-conjugated secondary antibodies.Coverslips were mounted with DAPI-containing VECTASHIELD mounting medium (Vector Labs; Burlingame, California, USA).At least 50 randomly selected vacuoles from 3 independent experiments were counted.

Mouse infections
Animal protocols and procedures were approved by the IACUCs at either the University at Buffalo and University of Wyoming.All mice were on a C57BL/6 background.C57BL/6 wildtype, IFN-γ −/− , CCR2 −/− and IDO -/-mice were purchased from The Jackson Laboratory (Jackson Laboratory, Bar Harbor, Maine, USA).NK cells were depleted using anti-NK1.1 (or mouse IgG2A as a control) as previously described [60].Morbidity was assessed by daily weighing and mortality by death or moribund state.Tissue cysts were enumerated 30 days post infection by counting numbers of RFP + cysts within three 25 μl aliquots of brain homogenates.

Quantitative PCR
Mice were killedAU : PleasenotethatasperPLOSstyle; donotusesacrificeinreferencetokillingofanimalsdu , tissues harvested and weighed, and genomic DNA isolated using the DNA Tissue kit (Omega, Norcross, Georgia, USA).gDNA.DNA (100 ng) was analyzed by real-time qPCR as described previously [11] using Toxoplasma B1 primers.Numbers of parasites were calculated from a standard curve generated in parallel using purified genomic DNA from known numbers of parasites.mRNA was analyzed by qRT-PCR as previously described [61] using the Toxoplasma gene-specific primers against TgPHYa and ß-actin (S1 Table ).

Histology
Mice were euthanized and small intestines were removed, and sections of the duodenum, jejunum, and ileum immediately fixed in a 10% formalin solution.Paraffin-embedded sections were cut longitudinally at 0.5 μm and hematoxylin and eosin stained.Inflammation was scored in a blinded manner using a 0 to 4 scoring scale: 0, no inflammation; 1, slight infiltrating cells in LP with focal acute infiltration; 2, mild numbers of infiltrating cells in the LP with increased blood flow and edema; 3, diffuse and massive infiltrating cells leading to disturbed mucosal architecture; and 4, crypt abscess and necrosis of the intestinal villi.For each animal, at least 2 sections (at least 100 μm from each other) were evaluated.

Enzyme-linked immunosorbent assay (ELISA)
IFNγ ELISA assays were performed in plasma according to manufacturer's instructions (from RnD Biosystems, Minneapolis, MN).

Immunoprecipitation
Tachyzoites grown in HFF were harvested by syringe lysis and washed in ice-cold PBS.Total protein (1 mg) was resuspended in 50 mM Tris-HCl (pH 7.4) with 1% Triton X-100, 100 mM NaCl, 1 mM NaF, 0.5 mM EDTA, 0.2 mM Na3VO4, 1× protease inhibitor cocktail (Thermo Fisher), incubated on ice for 30 min, and then subjected to sonication.Lysates were clarified by centrifugation at 16,000 × g, incubated with rabbit α-anti-HA clone C29F4 (Cell Signaling Technology) conjugated with protein G beads (Sigma-Aldrich) for 16 h at 4˚C.Immune complexes were separated with SDS-PAGE and then western blotted with antibodies against mouse anti-HA clone 6E2 (Cell Signaling Technology) or rabbit anti-TgSKP1 UOK75 [19].
were fixed and numbers and sizes of plaques determined (mean ± SEM, pooled from 3 independent experiments; ns = not significant, **P < 0.01 using a multiple comparison two-way ANOVA test).(B) HFFs were pretreated with IFNγ (1 U/ml) for 24 h and then infected with tachyzoites of the indicated strain.The cells were grown for 12 days in the presence of increasing tryptophan concentrations and then the monolayers were fixed and numbers of plaques enumerated.(C) HFFs were pretreated with IFNγ (1 U/ml) for 24 h and then infected with tachyzoites of the indicated strain.The cells were grown for 12 days in the presence of increasing tryptophan concentrations and then the monolayers were fixed and plaque sizes enumerated.Shown are the means ± SEM, pooled from 3 independent experiments; ns = not significant, *P < 0.05, **P < 0.01, using a multiple comparison two-way ANOVA test.

Fig 1 .
Fig 1. TgPHYa knockout in type II strain parasites.AU : AbbreviationlistshavebeencompiledforthoseusedinFigs1to9:Pleaseverifythatallentriesarecorrect: (A) Scheme for disrupting the TgPHYa locus by CRISPR.(B).Genomic DNA from ME49 WT and TgPHYaKO II parasites were analyzed by PCR using the primers depicted in (A).(C) Western blot analysis to detect SKP1 in lysates from the indicated strains.Note the increased mobility of the SKP1 in the TgPHYaKO II lysate indicating is loss of prolyl hydroxylation/glycosylation.(D) HFFs infected with WT or TgPHYaKO II parasites were grown for 12 days at 21% or 0.5% O 2 .The cells were then fixed, stained with crystal violet, and plaque sizes measured.Shown are averages ± SD from 3 independent experiments.(E) Representative images of plaques from (D).The data underlying the graph in this figure can be found in S2Table and blot in S1 Raw Images.HFF, human foreskin fibroblast; WT, wild-type.
C57BL/6J mice were infected by oral gavage with 50 or 100 wild-type (WT) or TgPHYaKO II tissue cysts and survival and body weight were monitored for 30 days.All mice infected with 50 TgPHYaKO II cysts survived up to 30 days post infection, while 40% of mice infected with the parental WT strain died (p = 0.06 Mantel-Cox test) (Fig 2A).Moreover, all mice infected with 100 WT cysts succumbed during the acute stage of the infection (within 8 days post infection), while <40% of the mice succumbed when infected with 100 TgPHYaKO II cysts (p = 0.02 Mantel-Cox test) (Fig 2A).Similar to decreased mortality, mice infected with the TgPHYa knockout lost less weight than those infected with WT parasites and their weight loss rebounded, whereas weight loss was sustained in the surviving WT infected mice (Fig 2B).

Fig 2 .
Fig 2. C57BL/6 mice are less susceptible to TgPHYaKO II strain infection correlating with fewer number of brain cysts.(A) Survival of C57BL/6J mice infected orally with 50 or 100 cysts of the indicated strains.Cumulative data from 3 independent trials and 9 mice for each strain and dose.Curves were compared by log-rank survival analysis of Kaplan-Meier curves, 50 cysts p = 0.0649, 100 cysts *P = 0.02.(B) Weight loss was monitored in mice infected with 50 cysts.Shown are cumulative data from 3 independent trials (mean ± SD *P < 0.05, multiple Student's t test with Holm-Sidak correction).Brain cyst burden (C) and diameter (D) were calculated from mice 30, 60, or 90 days after infection.Shown are cumulative data from 3 independent trials (mean ± SD *P < 0.05, multiple Student's t test with Holm-Sidak correction).(E) In vitro cyst development was assessed by Dolichos biflorus agglutinin lectin staining of parasites grown in HFFs for 5 days in in pH 8.2 medium.Shown is percentage of lectin + vacuoles from 3 independent assays (50 randomly selected vacuoles were counted for each sample).The data underlying the graphs in this figure can be found in S2Table.HFF, human foreskin fibroblast; WT, wild-type.

Fig 3 .
Fig 3. TgPHYaKO II is not required for colonization of intestinal epithelium but is important for parasite dissemination to peripheral tissues.(A) ME49 WT or TgPHYaKO II burdens were analyzed by quantitative RT-PCR in the indicated intestinal sections 7 days following gavage infection with 50 cysts.Shown are means ± SEM of 3 independent experiments with 3 mice per experiment (ns = not significant, using one-way ANOVA).(B, C) HEstained small intestinal sections from mice gavage infected with 50 WT or TgPHYaKO II cysts on day 7 (B) or day 12

Fig 4 .
Fig 4. TgPHYa is important for survival in inflammatory cells within the small intestine.Small intestine LP from uninfected mice and mice gavage-infected with 50 WT or TgPHYaKO II tissue cysts were harvested 7 dpi and analyzed by flow cytometry.Cells were gated on live/CD45 + /CD11b + cells to identify (A) inflammatory monocytes (IM; Ly6C hi Ly6G -) and neutrophils (Ly6G hi Ly6C int ), and (B) CD11b + CD103 + dendritic cells.(C) Parasite burdens were determined by assessing RFP expression of live gated cells.(D, E) Numbers and percentage of parasite-infected inflammatory monocytes and neutrophils, respectively, were determined by flow cytometry.Shown are representative flow cytometry plots.Graphs represent pooled analyses (mean ± SEM, pooled from 5 independent experiments); ns = not significant, *P < 0.05, **P < 0.001, ***P < 0.0001, using unpaired Student t test.The data underlying the graphs in this figure can be found in S2 Table.dpi, days post infection; LP, lamina propria; RFP, red fluorescent protein; WT, wild-type.https://doi.org/10.1371/journal.pbio.3002690.g004

Fig 5 .
Fig 5. TgPHYaKO II is important for survival within the intraperitoneal cavity.Mice were intraperitoneally infected with 10 4 WT or TgPHYaKO II tachyzoites and 7 days later the peritoneal exudate was harvested and processed for flow cytometry.(A) Live/CD45 + /CD11b + gated cells were stained to identify inflammatory monocytes (IM; Ly6C hi Ly6G low ) and neutrophils (Ly6G hi Ly6C int ).(B) Number and percentage of parasite-infected cells.Parasites burdens were identified by assessing RFP expression of live gated cells.(C, D) Numbers and percentages of parasite-infected inflammatory monocytes and neutrophils, respectively, were determined by flow cytometry.(E) Kaplan-Meier curve showing survival of mice IP infected with 10 6 WT or TgPHYaKO II tachyzoites.Cumulative data from 3 independent experiments (n = 9 total for each strain).The data underlying the graphs in this figure can be found in S2 Table.RFP, red fluorescent protein; WT, wild-type.https://doi.org/10.1371/journal.pbio.3002690.g005

Fig 8 .
Fig 8. TgPHYa is important for tryptophan utilization in IFNγ-treated cells.(A) Toxoplasma-infected HFFs were mock or IFNγ treated 24 hpi and 12 days later the monolayers were fixed and numbers and sizes of plaques determined (mean ± SEM, pooled from 3 independent experiments; ns = not significant, **P < 0.01 using a multiple comparison two-way ANOVA test).(B) HFFs were infected with the indicated strains for 24 h and then treated with IFNγ for 24.The cells were fixed and then stained to detect IRF1.Arrowhead highlights a parasite within a host cell that lacks IRF1 staining unlike neighboring uninfected cells.Note that the parasites are RFP + .

Fig 9 .
Fig 9. TgPHYaKO II is hypervirulent in IDO1-deficient mice.(A) HFFs were pretreated with IFNγ (1 U/ml) for 24 h in the absence or presence of 1-MT or epacadostat and then infected with tachyzoites of the indicated strain in the absence or presence of 1-MT or epacadostat.The cells were grown for 48 h in the presence of tryptophan (32 μg/ml) and then the monolayers were fixed and numbers of parasites per vacuole determined by IFA staining.(B, C) Mortality (B) and weight loss (C) were monitored in WT and IDO -/-mice infected with IP with 1,000 WT or TgPHYaKO II tachyzoites.Shown are cumulative data from 3 independent experiments.* P < 0.05 comparing TgPHYaKO II -C57/ BL67 to TgPHYaKO II IDO -/-using a multiple Student's t test.(D) Brain cyst burdens in mice 30 days post infection.Shown are means ± SD (ns = not significant, **P < 0.001, using a multiple Student's t test with Holm-Sidak correction).The data underlying the graphs in this figure can be found in S2 Table.HFF, human foreskin fibroblast; IFNγ, interferon gamma; WT, wild-type.https://doi.org/10.1371/journal.pbio.3002690.g009 (D) Representative images of plaques from S5C Fig. (E and F) Representative images of fields used for quantification in Fig 8C and 8D, respectively.The data underlying the graphs in this figure can be found in S2 Table.
(TIF) S6 Fig.TgPHYais dispensable for parasite resistance to reactive oxygen species.(A) Equal number of parasites were added to HFFs and grown in the presence of 0.2 mM H 2 O 2 .After 12 days, monolayers were fixed, numbers of plaques were counted, and results are expressed as the percentage of plaques formed relative to each strain grown without H 2 O 2 .(B) HFFs infected with tachyzoites of the indicated strain and then grown for 48 h in the presence of the indicated concentration of SNP.The monolayers were then fixed and numbers of parasites per vacuole determined by IFA staining.The data underlying the graphs in this figure can be found in S2 Table.LP, lamina propria; WT, wildtype. https://doi.org/10.1371/journal.pbio.3002690.g003 Table.BMDM, bone marrow-derived macrophage; HFF, human foreskin fibroblast; IFNγ, interferon gamma; RFP, red fluorescent protein.

S1 Table. List of primers and antibodies used for this study. (XLSX) S2 Table. Numerical values used for graphs.
(XLSX)