Parasite-induced IFN-γ regulates host defense via CD115 and mTOR-dependent mechanism of tissue-resident macrophage death

Host resistance to a common protozoan parasite Toxoplasma gondii relies on a coordinated immune response involving multiple cell types, including macrophages. Embryonically seeded tissue-resident macrophages (TRMs) play a critical role in maintaining tissue homeostasis, but their role in parasite clearance is poorly understood. In this study, we uncovered a crucial aspect of host defense against T. gondii mediated by TRMs. Through the use of neutralizing antibodies and conditional IFN-γ receptor-deficient mice, we demonstrated that IFN-γ directly mediated the elimination of TRMs. Mechanistically, IFN-γ stimulation in vivo rendered macrophages unresponsive to macrophage colony-stimulating factor (M-CSF) and inactivated mTOR signaling by causing the shedding of CD115 (CSFR1), the receptor for M-CSF. Further experiments revealed the essential role of macrophage IFN-γ responsiveness in host resistance to T. gondii. The elimination of peritoneal TRMs emerged as an additional host defense mechanism aimed at limiting the parasite’s reservoir. The identified mechanism, involving IFN-γ-induced suppression of CD115-dependent mTOR signaling in macrophages, provides insights into the adaptation of macrophage subsets during infection and highlights a crucial aspect of host defense against intracellular pathogens.

Monocyte-derived macrophages play a central role in pathogen elimination via multiple mechanisms including production of proinflammatory cytokines, nitric oxide, and reactive oxygen species [3,4,7].Monocytes and monocyte-derived macrophages closely cooperate with dendritic cells in the regulation of T cell responses [23][24][25].In contrast to monocyte-derived macrophages, our knowledge of TRM functions in the context of microbial infection is relatively limited.The unique response of TRMs to microbial infection is called the macrophagedisappearance reaction (MDR) which is characterized by contraction of the TRM population due to poorly understood mechanisms including tissue adherence, migration, and cell death during the first phase of the inflammatory response [26,27].Following their diminution during microbial or viral insult, the remaining TRMs undergo M-CSF-dependent proliferation, which contributes to their repopulation.This repopulation is bolstered by recruited monocytes that acquire a TRM-like phenotype during the resolution phase of the inflammatory response.Being located in a sterile environment, peritoneal macrophages serve as an excellent experimental model to study the first response of TRMs to microbial infections.Peritoneal macrophages are predominantly comprised of TRMs that can be identified as CD11b+F4/80+CD102 + cells expressing low levels of MHCII in naïve mice [28].Monocyte-derived macrophages represent a minor subset and express low levels of F4/80 and CD102, and high levels of MHCII.Here we investigate the dynamic response of the peritoneal TRMs to T. gondii, a common intracellular pathogen that triggers a highly polarized Th1 response [29,30].We observed that T. gondii infection and the associated immune response results in elimination of TRMs in peritoneal cavity, liver, and intestine during the acute response to the parasitic infection.We established that the loss of TRMs is mediated by IFN-γ and resulted in the shedding of CD115 from the macrophage cell surface.Our additional experiments revealed that the loss of CD115 triggered the cell death of TRMs via an mTOR-dependent mechanism that we recently identified in the context of the epithelial cell responses to IFN-γ [31].Additional experiments uncovered that this TRM response to IFN-γ played a role in host resistance to the parasite, and the preservation of TRMs achieved by selectively targeting the IFN-γ receptor in TRMs caused an increased parasite burden responsible for the acute susceptibility of infected mice to T. gondii infection.When combined, our results established that IFN-γ was required and sufficient for triggering the CD115 and mTOR-dependent mechanism of LPM death in vivo, revealing a mechanism of host resistance to T. gondii mediated by limiting a reservoir for the common intracellular parasite.

Loss of tissue-resident macrophages during T. gondii infection in vivo
Macrophages are sentinels of the immune system and are present in all lymphoid organs and tissues.These cells are involved in early interactions with pathogens, including T. gondii [25,[32][33][34].Two distinct macrophage subsets coexist in the peritoneal cavity [28].The large peritoneal macrophages (LPMs) are seeded during embryonic development and represent the dominant population in naïve mice (Figs S1 and 1).LPMs are characterized by high levels of the canonical surface markers CD11b and F4/80 (Figs 1A and S1).The LPMs can also be identified as Lin-CD11b+CD102+ cells that express low levels of MHCII (Figs 1C and S1).In contrast, monocyte-derived small peritoneal macrophages (SPMs) express low levels of F4/80 and CD102, and high levels of MHCII and represent a minor population in naïve mice (Figs 1A and 1B and S1).It has been established that monocytes and monocyte-derived macrophages play a crucial role in host defense against the intracellular parasite T. gondii [30,32,35]; but the roles of resident macrophages in controlling the immune response during infection with T. gondii is largely unknown.
We noted major changes in the composition of macrophage subsets during the acute response to T. gondii.We observed a nearly complete loss of the LPMs by day 5 post-infection (Fig 1).This observation was established independently of the gating strategies used to define the LPMs as either CD11b+F4/80+ (Fig 1A and 1B) or CD11b+CD102+MHCII-cells (Fig 1C *1F), strongly suggesting that the acute response to T. gondii led to LPM loss (Fig 1).In striking contrast, SPMs became the dominant macrophage subset during the acute response to the parasite.This was evident from the analysis of the absolute and relative frequencies of SPMs on day 5 post infection (Fig 1B,1D and 1F).While massive infiltration of monocytederived macrophages to the site of T. gondii infection has been previously described [34,[36][37][38], the loss of the LPMs was unexpected and prompted us to further investigate the response of this macrophage subset to the parasite.We observed that even during the chronic stage of infection, LPMs were practically undetectable in the peritoneal cavity of the infected mice, indicating the parasitic infection caused a long-term change in the macrophage subsets (Fig 1).Not only was there a dramatic reduction in the relative frequencies of LPMs, which could be explained by the migration of inflammatory monocytes, but an absolute quantification of LPMs revealed a complete or nearly complete loss of this macrophage subset that persisted past the acute stage of infection (Fig 1B , 1D and 1F).
Furthermore, experimental peritoneal infection resulted in a near complete elimination of hepatic resident macrophages, known as Kupffer cells (KCs), that were identified via VSIG4 (also known as CRIg) that is selectively expressed by KCs (S2A Fig) .The disappearance of KCs coincided with the loss of LPMs (S2B Fig).
We also discovered that, similar to the experimental intraperitoneal T. gondii infection, the natural oral route of infection led to the loss of the intestinal macrophages (Fig 2A -2C).Interestingly, oral infection with T. gondii also resulted in the loss of LPMs (Fig 2D and 2E), but it did not lead to significant recruitment of SPMs to the peritoneal cavity (Fig 2E).Therefore, the

PLOS PATHOGENS
IFN-γ induced tissue-resident macrophage death in host resistance to Toxoplasma death of LPMs and the recruitment of SPMs represent distinct events during immune responses to T. gondii.Although these events occur simultaneously during intraperitoneal infection with the parasites, they likely happen independently, as evidenced by the natural oral route of infection (Fig 2D and 2E).This is evident from the dramatic reduction in the total number of LPMs in orally infected mice compared to naïve controls ( These findings indicate that the recruitment of SPMs and the loss of LPMs can occur independently during immune responses to T. gondii.While the mechanism of SPM recruitment to the site of infection has been previously established as a CCL2 dependent process [39][40][41][42], the loss of LPMs regardless of the infection route, prompted us to investigate the underlying mechanism behind TRM depletion during acute responses to the parasite.

IFN-γ is required and sufficient for the loss of large peritoneal macrophages
IFN-γ is a crucial cytokine that regulates both host resistance and immunopathological responses to T. gondii infection [43][44][45][46].IFN-γ mediated activation of macrophages, known as priming, is required for the elimination of intracellular pathogens including T. gondii  https://doi.org/10.1371/journal.ppat.1011502.g002[7,30,44,47,48].At the same time, the results from our and other laboratories revealed that IFN-γ can trigger major changes in tissue cellular composition via triggering cell death during an inflammatory response [31,[49][50][51][52][53].To determine if IFN-γ played a role in the observed loss of LPMs seen in T. gondiii infected mice, we first analyzed macrophages in parasite-infected mice in the presence of IFN-γ blocking antibodies.We observed that neutralization of IFN-γ prevented the loss of LPMs during the acute response to the parasite (Fig 3A and 3B).These results suggested that IFN-γ was required for the disapperance of LPMs seen in T. gondiiinfected mice (Fig 3A and 3B).We next examined whether IFN-γ was sufficient for the elimination of LPMs in vivo in the absence of parasitic infection.This was crucial as T. gondii frequently infects macrophages and can modulate their functions via multiple virulence factors injected into the infected cells [54][55][56][57][58]. To test this possibility, naïve mice were injected with recombinant IFN-γ (rIFN-γ) and the macrophage subsets in the peritoneal cavity were analyzed 3 days later.We observed that a single injection of rIFN-γ, resulting in a comparable level of the circulating cytokine seen in T. gondii-infected mice (S3 Fig), was sufficient to drive the loss of LPMs (Fig 3C and 3D).These results strongly indicated that even in the absence of T. gondii infection, IFN-γ can trigger the dissaperance of LPMs.We also noted that the injection of rIFN-γ alone was sufficient for recruiting SPMs to the peritoneum (Fig 3D), an observation that requires additional further studies.When combined, our results formally established that IFN-γ was both required and sufficient for the changes in peritoneal macrophage composition characterized by the loss of LPMs (Fig 3A -3D).

IFN-γ signaling in macrophages is required for the loss of large peritoneal macrophages
IFN-γ has pleiotropic effects on multiple cell types [59].Experiments with systemic blockage of or supplementation with IFN-γ do not discriminate between the direct effects of this cytokine on peritoneal macrophages or indirect effects caused by IFN-γ activating other cell types (Fig 3A -3D).This knowledge is crucial for deciphering the molecular mechanisms required for the loss of LPMs during T. gondii infection.
To examine whether the direct effects of IFN-γ on macrophages are responsible for LPM loss during the acute response to T. gondii, mice with an IFN-γ receptor deficiency specific to myeloid cells were generated by crossing Mlys-cre mice with Infgr2 flox/flox mice.Myeloid-cellspecific IFN-γRII knockout mice (M-IFNgR KO) and their cre-negative littermate controls were next infected with the parasite and changes in macrophage populations were analyzed on day 5 post-infection (Fig 3E and 3F).As anticipated, cre-negative mice were indistinguishable from the wild-type (WT) controls and lost their LPMs during the acute response to T. gondii (Fig 3E and 3F).In contrast, we observed that both the frequencies and absolute numbers of LPM and SPM populations were practically unchanged when comparing day 5 infected and naïve control M-IFNgR KO mice (Fig 3E and 3F).These results indicate that IFN-γ acting directly on peritoneal macrophages is responsible for the loss of LPMs (Fig 3E and 3F).

IFN-γ treatment drives the death of large peritoneal macrophages
Several non-mutually exclusive mechanisms can explain the rapid loss of LPMs during the acute response to T. gondii.Among the major possibilities are migration of LPMs caused by T. gondii infection, or IFN-γ mediated death of LPMs, a mechanism recently reported for intestinal epithelial cells [31,49].
We first examined whether IFN-γ can trigger LPM migration to the omentum.The omentum is located in the peritoneal cavity and is composed of two mesothelial sheets enclosing adipose tissue.Recenly, it was observed that T. gondii infected macrophages and dendritic cells can rapidly relocate from the peritoneum to the omentum for induction of T cell responses to the parasite [60].To minimize artifacts associated with cell isolation or incomplete analysis of the omentum, the entire omenta from naïve and T. gondii infected mice were next analyzed by whole-mount imaging (S4 Fig) .While an increase in CD11b+ cells was observed in the omentum analyzed from T. gondii infected mice, only an incremental amplification in peritoneal LPM-specific CD102+ cells [61,62] was seen (S4 Fig) .These results indicate that the loss of

PLOS PATHOGENS
IFN-γ induced tissue-resident macrophage death in host resistance to Toxoplasma LPMs during parasitic infection was not the result of detectable cell relocation from the peritoneal cavity to the omentum.
To test a hypothesis that T. gondii-induced IFN-γ lead to the death of LPMs via a mechanism recently reported by us for Paneth cells [31], we first applied an in vitro system with enriched peritoneal LPMs stimulated with rIFN-γ (Fig 4A and 4B).Timelapse imaging analysis of macrophages treated with rIFN-γ was performed for 18 hours and dying cells were detected with propidium iodide (PI) staining (Fig 4A ).We observed that IFN-γ treatment resulted in the rapid appearance of PI+ cells (red) during the course of the experiment (Fig 4A and 4B).In contrast, only minimal cell death was seen in the untreated controls.These results revealed that IFN-γ can trigger the death of LPMs, but does not formally demonstrate the direct causality of this cytokine in the induction of macrophage death.To examine the direct effect of IFNγ signaling on macrophages, we isolated macrophages from the peritoneal cavity of M-IFNgR KO mice and treated them with rIFN-γ prior to utilizing imaging flow cytometry to assess cellular death.We observed that a lack of responsiveness to IFN-γ precluded the death of LPMs (Fig 4C and 4D).At the same time and analogous to the WT macrophages, the cre-negative Ifngr flox/flox macrophages were PI+ within 18 hours of treatment with rIFN-γ (Fig 4C and 4D).When combined, these results revealed that IFN-γ signaling in macrophages directly leads to the loss of LPMs due to their death.We also observed that LPMs exhibited substantially higher sensitivity to cell death triggered by IFN-γ compared to SPMs when both were exposed to the same cytokine in vitro (S5 Fig) .This heightened sensitivity suggests a potential explanation for the selective loss of LPMs observed during potent type I immune responses elicited during T. gondii infection (Fig 1).

IFN-γ triggers a CD115-and mTOR-dependent death of large peritoneal macrophages
IFN-γ triggers profound metabolic changes in multiple cell types, including macrophages [63,64].One of the key molecules that senses cell metabolism is the mTORC1 kinase complex, which plays a critical role in integrating growth receptor signaling with cellular functions, including cell survival [65].In addition, we have recently observed that in the context of T. gondii infection, IFN-γ leads to the death of intestinal Paneth cells via an mTOR-dependent mechanism distinct from apoptosis, necroptosis, or pyroptosis [31].Furthermore, it has been demonstrated that IFN-γ can suppress the activity of mTORC1 in other cell types including macrophages [66].Considering that, we next examined if mTOR supression was also responsible for the death of LPMs.
To test this hypotheiss, we first examined the requirements for mTORC1 activity in LPMs and SPMs.We observed that treatment of mice with the mTORC1 inhibitor rapamycin resulted in the rapid and selective loss of LPMs (Fig 5A and 5B).While this treatment blocked mTOR activities in both LPMs and SPMs, only LPMs were susceptible to this treatment in vivo as evidenced by the analysis of dying cells in response to rapamycin treatment (Fig 5C and 5D).We observed that only a small fraction of SPMs were stained with the dead cell Among the various factors that are responsible for sustained mTORC1 activity in macrophages, the M-CSF-dependent activation of its receptor, CD115, is of particular interest.M-CSF signaling plays an indispensable role in the development of macrophages and is essential for the survival and maintenance of LPMs in the peritoneal cavity [9,20].For the above reasons, we examined the expression of CD115 in naïve or T. gondii infected mice, along with the animals stimulated with rIFN-γ.We observed that both parasitic infection and cytokine treatment in vivo resulted in the rapid loss of CD115 cell surface expression (Fig 6A).Furthermore, when examining the peritoneal exudate fluid collected from the same mice, we observed the release of the extracellular portion of CD115 (Fig 6B ), suggesting that IFN-γ triggers shedding of CD115 from peritoneal macrophages.
To determine if direct IFN-γ stimulation of macrophages leads to the loss of CD115 from the cell surface during T. gondii infection, CD115 expression was next analyzed in M-IFNgR KO mice infected with the parasite.We observed that the lack of IFNgR prevented the loss of CD115 cell surface expression in myeloid cells (Fig 6C).In addition, the lack of IFNgR prevented the release of soluble CD115 into the peritoneal cavity of the infected mice (Fig 6D).As anticipated, the appropriate cre-negative IfngrII flox/flox mice behaved indistinguishable from the WT controls, and lost CD115 from the cell surface of the peritoneal macrophages (Fig 6).When combined, these data revealed that IFN-γ leads to the shedding of CD115 in both LPMs and SPMs, but considering that M-CSF is essential for the survival and maintenance of TRMs but not SPMs, the loss of CD115 resulted in the selective death of LPMs in vivo (Fig 6).Of note, blocking CD115 in vitro failed to induce rapid death of peritoneal macrophages (S7 Fig) , suggesting distinct requirements for M-CSF signaling in the survival of peritoneal macrophages in vivo and in vitro.

Macrophage IFN-γ signaling is required for host survival during acute toxoplasmosis
The IFN-γ-mediated selective loss of LPMs in response to T. gondii infection prompted us to investigate whether this phenomenon is a host defense mechanism or an immunopathological response induced by IFN-γ.To distingushing among those possibilities, we first examined the survival of M-IFNgR KO mice infected with T. gondii.We observed acute susceptibility in these mice, displaying kinetics similar to those of complete IFN-γ-deficient mice (Fig 7).These findings established that the macrophage response to IFN-γ is crucial for resistance against parasitic infection, even when other cell types can respond to IFN-γ (Fig 7A).The acute susceptibility observed in the mice was directly associated with an uncontrolled parasite burden (Fig 7B and 7C).
We next investigated whether the IFN-γ-dependent death of f LPMs limits the replication niche for the parasite.In vitro experiments revealed that T. gondii parasites frequently target LPMs (Fig 7D and 7E), supportring a model that LPMs are an important reservour for the parasite.To further investigate this in an in vivo setting, we employed a strategy to deplete LPMs by blocking the CD115 receptor (Fig 7F).This approach allowed us to achieve depletion of LPMs in an IFN-γ-independent manner.We observed that the depletion of LPMs through a CD115 blockade had no impact on the parasite burden in wild-type (WT) mice, as they were already experiencing LPM loss due to IFN-γ, irrespective of the additional treatment (Fig 7F ,  7G and 7H).However, when the response to IFN-γ was abolished due to complete inactivation of the receptor for this cytokine, the depletion of LPMs resulted in a significant reduction in successful parasite replication (Fig 7H).Although CD115-mediated depletion did not decrease the parasite burden in IFN-γ receptor-deficient mice to the level observed in WT controls (Fig 7H ), our experimental findings strongly support the hypothesis that the IFN-γ-mediated elimination of LPMs plays a role in host defense by controlling the elimination of a reservoir for the intracellular parasite T. gondii (Fig 7H).

PLOS PATHOGENS
IFN-γ induced tissue-resident macrophage death in host resistance to Toxoplasma

Discussion
Our study unveils a novel host defense mechanism against T. gondii infection, which involves the IFN-γ-mediated elimination of LPMs.We have demonstrated that IFN-γ directly impacts LPMs, leading to their death through CD115 and mTOR-dependent pathways in vivo.This host defense process plays a role in restricting the available niche for the intracellular parasite.Our findings expand upon the existing understanding of IFN-γ-mediated host resistance to intracellular pathogens, characterized by the activation of macrophages prior or during infection with T. gondii.
TRMs are recognized as key regulators in tissue homeostasis, organ development, and the resolution of infections.Although extensive research has been conducted on the development and lineage of TRMs using lineage tracing and RNAseq techniques [61], their specific role in host defense remains incompletely understood.The loss of TRMs during microbial infection is a distinctive feature known as the MDR [27].This response has been observed in relation to diverse microbial components, such as zymosan, LPS, and live bacterial pathogens, and the current study extends this concept to IFN-γ.Notably, the MDR or MDR-like loss of TRMs is not confined to the peritoneal cavity but has also been observed in other tissues.For instance, in the intestine, microbiota drives replacement of TRMs by monocyte-derived macrophages, suggesting that even low concentrations of microbial ligands can trigger the MDR [27].Similarly, inflammatory responses can lead to the replacement of liver TRMs with monocytederived cells [67], indicating that the loss of TRMs is a common physiological or pathophysiological response to microbial ligands and inflammatory cytokines.
In our investigation, we discovered that IFN-γ stimulation in vivo renders macrophages unresponsive to M-CSF by inducing the shedding of its receptor, CD115.Given the critical role of M-CSF in TRM development and survival, the loss of CD115 is a crucial step contributing to the elimination of TRMs.M-CSF is a tighly controlled growth factor that contributes to monocyte to macrophage differentiation.Both macrophages and fibroblasts can prodcue M-CSF, and its levels regulate macrophage presence through autocrine and paracrine mechanisms [9,68].Thus, the regulation of M-CSF responsiveness is an effective strategy to control macrophage populations.Functionally, M-CSF can enhance phagocytic activity [69], antigen presentation [70], and cytokine production [71][72][73], thereby augmenting the ability of macrophages to eliminate pathogens and modulate immune responses.The distinct requirement for mTOR signaling may suggest that monocyte-derived macrophages are better equipped to combat intracellular pathogens in the highly inflammatory environment caused by large amounts of IFN-γ seen during T. gondii infection.This may explain why TRMs cannot be reprogrammed by IFN-γ to acquire the functions of proinflammatory SPMs, although further investigation is needed to fully explore this hypothesis.Our study suggests that the IFN-γinduced CD115 shedding is a rapid and efficient strategy to eliminate the preexisting TRMs.Intriguingly, IFN-γ-induced CD115 shedding was not limited to LPMs, but the distinct dependence on M-CSF determines the selective loss of TRMs during the acute response to T. gondii infection.This mechanim of host defense cooperates with TLR-dependent and independent production of CCL2 that leads to the recruitment of monocytes to the site of infection that are indispensible for host resistance to T. gondii [39][40][41][42].Our results also revealed that the loss of LPMs and the recruitment of monocytes and monocyte-derived macrophages are distinct events that can be separated, as seen in orally infected mice.These findings indicate that the death of LPMs alone is insufficient for the recruitment of SPMs.Instead, while not formally proven, our data suggest that the presence of the parasite or parasite-derived molecules is required for the recruitment of monocytes during T. gondii infection, providing a locationspecific immune response to the pathogen.While future biochemical studies are needed to define IFN-γ inducible enzymes responsible for CD115 shedding, it has been reported that metalloproteases, particularly ADAM17 (a disintegrin and metalloprotease 17) and ADAM10 (a disintegrin and metalloprotease 10) cleave the extracellular domain of CD115, releasing it as a soluble form into the extracellular environment [74][75][76].These enzymes can be activated in response to various stimuli, including cytokines and microbial components [77], suggesting a general strategy to regulate macrophage responses in different physiological and pathological conditions via regulation of responsivness to M-CSF.
In summary, our study reveals an IFN-γ-induced mechanism of host resistance to T. gondii, characterized by suppression of CD115-dependent mTOR signaling in macrophages at the site of infection and potential the sites of parasite dissemination.Although additional studies are necessary to fully understand the necessity of TRM elimination, we speculate that it is a common host defense mechanism against intracellular infections.We provide experimental evidence supporting the concept that the elimination of TRMs and their subsequent replacement by monocyte-derived cells represents an IFN-γ-mediated host defense strategy against a common protozoan parasite capable of infecting macrophages.

Ethics statement
All animal experimentation in this study was reviewed and approved by the University of Rochester's University Committee on Animal Resources (UCAR), the Institutional Animal Care and Use Committee (IACUC).

Mice
C57BL/6 mice were originally purchased from the Jackson Laboratories and maintained in a pathogen-free animal facility at the University of Rochester of Medicine and Dentistry.Ifngr2flox/flox mice were generated using targeted embryonic stem cells obtained from the Knockout Mouse Project repository and injected into C57Bl/6 albino blastocysts by the Fox Chase Cancer Center Transgenic Mouse Facility as previously described [31,78].Mlys-Cre (LysM-Cre, Strain #004781) mice were purchased from the Jackson laboratory.Mice for all experiments were age-and sex-matched.

Visualization of macrophage death in vitro
For time-lapse imaging, 2x10 5 peritoneal macrophages enriched for LPMs through cell adhesion exhibited a purity of at least 95% were imaged using a TE2000-U microscope (Nikon) coupled to a CoolSNAP HQ CCD camera with a 20X objective.Images were taken at the following intervals: brightfield images every 30 seconds and DAPI/Cy5 channels every 1 minute.Imaging took place over 18 hours in a humidified, temperature-controlled chamber maintained at 37˚C for the duration of imaging.To assemble the time-lapse move, images were compressed at normal quality and set at 100 ms speed.
For quantitative analysis of cell death, 2x10 4 freshly isolated PECs were plated in each well of a black-walled, clear bottom 96 well plates, and macrophages were enriched for a final density of 1x10 4 macrophages per well.Prior to imaging, an additional 100 μl complete media was added to the wells containing either rIFN-γ or PBS (untreated) and a mixture of Hoescht 3364 and PI.Images were then captured using the Total (Brightfield) + Live (Blue-Hoescht) + Dead (Red-PI) setting on a Celigo Imaging Cytometer (Nexcelom).In some experiments, cell-sorted purified LPMs and SPMs were incubated with the indicated amounts of rIFN-γ or rapamycin for 18 hours (as well as for 12 and 24 hours) prior to the analysis of cell death by Zombie Yellow staining.

Determination of pathogen loads during infection
To determine T. gondii pathogen loads, total genomic DNA from animal tissue was isolated by using the DNeasy Blood and Tissue Kit (Qiagen) according to manufacturer's instructions.PCR were performed by using SSOFast Eva Green Supermix (BioRad).Samples were measured by qPCR using a MyiQ Real-Time PCR Detection System (BioRad), and data from genomic DNA was compared with a defined copy number standard of the T. gondii gene B1.

Immunofluorescence staining of the liver
For Immunofluorescence images, livers were fixed in 4% PFA in PBS at 4˚C for 1 hour.Following fixation, samples were incubated in a 30% sucrose solution in PBS overnight at 4˚C.Samples were then frozen in OCT (Tissue Tek) compound.10 μm liver sections were washed three times with 50mM NH 4 Cl in PBS for 3 minutes prior to permeabilization with 0.25% Triton X-10 solution in PBS for 10 minutes at room temperature.Primary antibody staining for VSIG4 was performed in 2% serum in PBS overnight at 4˚C.Slides were mounted in ProLong Gold (Molecular Probes).Specimens were imaged with a Leica SPE system (Leica DMi8) fitted with a Leica 63X objective NA 1.4.

Identification of TRMs within the small intestine
For flow cytometric analysis of intestinal TRMs in the lamina propria, small intestine segments were washed, flattened, cut into 1.5cm pieces and shaken in 2 changes of HBSS with 5% FBS and 2 mM EDTA for 20 min at 37C.The suspension was passed through a mesh strainer to

PLOS PATHOGENS
IFN-γ induced tissue-resident macrophage death in host resistance to Toxoplasma remove epithelial cells.The intestinal pieces were chopped using scissors and shaken in digestion media containing HBSS (Gibco) with 5% FBS, 1mg/ml Collagenase D, 2U/ml DNase I, 0.1U/ml Dispase for 30 mins at 37C.The digested samples were then vortexed briefly and passed through a 100-micron filter followed by staining for flow cytometry.

Whole mount immunofluorescence of the omentum
Mice were sacrificed and whole omenta were excised.Samples were placed in 6ml polypropylene tubes and blocked using FC Block (PharMingen, San Diego, CA) at 10μg/ml in 200 μl of FACs buffer and placed on a shaker for 15 minutes at 4˚C.Following incubation, samples were stained with the following antibodies for 2 hours at 4˚C: CD11b-PE (eBioscience; Clone M1/70), CD102-Alexa Fluor 647 (BioLegend; Clone 3C4), and CD115-BV421 (BioLegend; Clone AFS98) or left unstained.Samples were washed twice by the addition of 4ml FACs buffer and rotated at 4˚C for 30 minutes.After the final wash, samples were collected from the tubes using a snipped off P1000 tip and placed onto glass slides with FACs buffer.A cover slip was placed on top of the omentum and pressed down.Samples were viewed via fluorescence microscopy and digital images were acquired.
Fig 2E).LPMs still constituted a relatively large cellular population (Fig 2D) due to the low abundance of SPMs in the peritoneal cavity of orally infected mice (Fig 2E).

Fig 4 .
Fig 4. IFN-γ mediates the death of LPMs.(A) LPMs were cultivated in a temperature, CO 2 , and humidity-controlled chamber in the presence or absence of rIFN-γ.Time lapse imaging was performed for 18 h in the presence of Hoescht 3364 and Propidium Iodide (PI), and the representative snapshots at 3, 6, 12, and 18h are shown.(B) Quantification of live LPMs detected as PI-negative cells at 3, 6, 12, and 18 h post treatment with rIFN-γ during the time lapse imaging.(C) LPMs prepared from M-IFNgR KO mice or the appropriate WT controls (Cre-negative Ifngr2 flox/flox littermates) were plated onto black walled, flat-bottom, 96 well plates and the cells were either left untreated or treated with rIFN-γ for 18 hours.Images were captured using the Celigo imaging cytometry and (D) dead cells were quantified as PI+ cells after 18 h.Scale bar = represents 200 μm.Error bars = mean ± SEM. ns P > 0.05, ****P < 0.00001.https://doi.org/10.1371/journal.ppat.1011502.g004