To dissect the cell death pathway responsible for Paneth cell loss during T. gondii infection, we performed experiments in which the role of apoptosis, which is known to be involved in intestinal cell death, was investigated via a terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay, and the expression of cleaved Caspase 3, an essential mediator of apoptosis was also evaluated. We observed that Paneth cells did not acquire TUNEL- or cleaved Caspase 3-positivity on any of the examined days post infection, ruling out apoptosis as the major mechanism of Paneth cell death (Figure 2), even though it was previously reported that in some rare cases IFN-γ is capable to activate caspase-3/7-dependent pathway of cell death in Paneth cells (Eriguchi et al., 2018).

In addition, inhibition of apoptosis by caspase inhibitor zVAD failed to prevent Paneth cell death (Figure 2—figure supplement 1). Importantly, zVAD treatment of the intestinal organoids prevented the loss of the stem cell-specific transcript, signifying a role for apoptosis in stem cell death caused by IFN-γ, and suggesting that in contrast to the intestinal stem cells (Kretzschmar and Clevers, 2019; Takashima et al., 2019), Paneth cells undergo a different mechanism of death in response to this cytokine (Figure 2—figure supplement 1). Two additional cell death pathways that are associated with inflammatory responses were next investigated. During pyroptosis, intracellular activation of inflammasomes and Caspase 1 or 11 results in rapid formation of plasma membrane pores and cell lysis (Broz et al., 2020; Place and Kanneganti, 2018). In agreement with the zVAD treated mice, the absence of Caspases 1 and 11 failed to prevent the loss of Paneth cells, further confirming no role for these caspases in Paneth cell death. The absence of the additional key proteins involved in pyroptosis, including NLRP3 and ASC (encoded by Pycard) (Orning et al., 2019), failed to prevent the death of Paneth cells, eliminating pyroptosis as the mediator of Paneth cell death during T. gondii infection (Figure 2—figure supplement 2A and B).

Similarly, we ruled out a role for necroptosis, a cell death mechanism that depends on RIPK1 and RIPK3, in Paneth cell death downstream of IFN-γ via analysis of mice deficient in RIPK3 (Figure 2—figure supplement 2C and D) and animals treated with necrostatin (Figure 2—figure supplement 2E), a potent and well-characterized inhibitor of RIPK1 (Choi et al., 2019). We observed that inactivation of RIPK1 or RIPK3 had no detectable effect on Paneth cell loss during T. gondii infection (Figure 2—figure supplement 2C–2E). Taken together, these data revealed that IFN-γ-dependent death in Paneth cells was independent of apoptosis, pyroptosis, and necroptosis (Figure 2, Figure 2—figure supplement 1, Figure 2—figure supplement 2).

Infections result in cytokine-mediated metabolic reprogramming, which has been primarily studied in immune cell populations (O’Neill et al., 2016) . Among several soluble mediators, IFN-γ is capable of triggering profound metabolic changes in multiple cell types, including epithelial cells (Nava et al., 2010). One of the key sensors in cell metabolism is the mTORC1 kinase complex, which plays a critical role in integrating metabolism with signal transduction (Condon and Sabatini, 2019). In addition, mTORC1 has been strongly implicated in controlling proper Paneth cell functions (Yilmaz et al., 2012; Barron et al., 2017; Sampson et al., 2016). Tracking the activity of mTOR measured by ribosomal S6 protein phosphorylation (pS6^{Ser 240/244}) in the intestinal crypts revealed high level of mTOR activity in Paneth cells when compared to stem cells in the PC-Cre x TdTomato x Lgr5-GFP double reporter mice that allowed simultaneous detection of these cell types (Figure 3—figure supplement 1). These observations prompted us to examine if IFN-γ leads to Paneth cell death via mTORC1 inactivation. We first analyzed the phosphorylation of the ribosomal S6 protein, a well-characterized marker of mTORC1 activity in mice infected with T. gondii and treated with anti-IFN-γ therapy throughout infection. We observed a rapid loss of ribosomal S6 protein phosphorylation in Paneth cells as early as day 3 post infection (Figure 3A), and this loss preceded the loss of Paneth cells detected by immunofluorescence staining (Figure 3A) or quantitative analysis of Paneth cell-specific transcripts (Figure 3B, Figure 3—source data 1 and data not shown). Successful blockade of IFN-γ in vivo with a neutralizing antibody (Figure 3A and C) prevented the loss of ribosomal S6 protein phosphorylation in Paneth cells, which were instead retained throughout the course of the experiments (Figure 3A and B). These results reveal that IFN-γ plays a role in modulating mTORC1 activity in Paneth cells in vivo. Additionally, Paneth cells lacking IFN-γRII retained ribosomal protein S6 phosphorylation during T. gondii infection (Figure 3D and E, Figure 3—source data 1), further suggesting that direct effects of IFN-γ on Paneth cells result in the loss of mTORC1 activity in this cell type. In addition, we observed a moderate increase in mTORC signaling in anti-IFN-γ treated animals or in mice lacking IFN-γRII in Paneth cells (Figure 3A). These observations suggest that basal IFN-γ, most likely driven by the intestinal microbiota (Benson et al., 2009), has a negative impact on mTOR activity.

To formally examine whether IFN-γ alone is capable of impacting mTORC1 activity in Paneth cells, naïve mice were injected with recombinant IFN-γ. This treatment resulted in levels of circulating IFN-γ comparable to those seen in T. gondii-infected mice (Figure 4A, Figure 4—source data 1) and triggered the induction of the representative IFN-stimulated genes, including Cxcl10, Ido1, and Irgm3, in the small intestine (Figure 4C). We observed that IFN-γ treatment alone triggered Paneth cell disappearance (Figure 4B, Figure 4—source data 1) and resulted in a rapid loss of S6 ribosomal protein phosphorylation that preceded Paneth cell loss (Figure 4D).

Finally, phosphorylation of the mTOR kinase itself was severely compromised in Paneth cells during IFN-γ treatment (Figure 4E,F, Figure 4—source data 1). Taken together, these experimental approaches reveal that IFN-γ alone is sufficient to elicit profound inhibition of mTORC1 activity in Paneth cells that precedes their loss in vivo.

We next formally examined the requirements for an intact Paneth cell-intrinsic mTORC1 signaling pathway in the survival of these cells. To achieve this, mice with targeted deletion of the mTORC1 activator Raptor (PC-Raptor KO mice) or the mTOR kinase (PC-Mtor KO mice) in Paneth cells were generated. We observed that a Paneth cell-restricted deficiency in Raptor or mTOR resulted in nearly complete disappearance of these cells in naïve mice (Figure 5). This was evident in immunofluorescence analysis of crypts in the small intestine, which showed that Paneth cells were detected in WT mice but not in PC-Mtor or PC-Raptor KO mice (Figure 5A). Similarly, unbiased analysis of Paneth cell-specific transcripts revealed that both mTOR and Raptor1 were required for the presence of Paneth cells and functioned in a cell-intrinsic manner for Paneth cell maintenance (Figure 5B, Figure 5—source data 1). In addition to the genetic models, we observed that treatment of mice with the mTORC1 inhibitor rapamycin resulted in the rapid loss of Paneth cells (Figure 5C), which was also evidenced by reduced expression of Paneth cell-specific transcripts (Figure 5D, Figure 5—source data 1). When combined, our experiments reveal that Paneth cell-intrinsic mTORC1 activity is essential for the presence of Paneth cells.

Our experiments with IFN-γ and rapamycin treatments suggested that during T. gondii infection, IFN-γ functions upstream of mTORC1 and causes rapid Paneth cell loss due to its potent inhibitory effect on mTORC1. To formally examine this hypothesis, WT mice were infected with T. gondii and treated with rapamycin during the course of infection. As expected, rapamycin treatment prevented the induction of proinflammatory cytokines, including IFN-γ, in T. gondii-infected mice due to immunosuppressive functions (Figure 5—figure supplement 1). Most importantly, despite completely inhibiting IFN-γ responses in T. gondii-infected mice, rapamycin treatment resulted in a loss of Paneth cells that was preceded by a lack of mTOR activity in these cells measured by evaluating the level of the phosphorylated ribosomal S6 protein (Figure 5—figure supplement 1A). These results placed inhibition of mTORC1 response downstream of the IFN-γresponse in Paneth cells in vivo (Figure 5—figure supplement 1).

The kinase mTOR integrates inputs from growth factor receptors, nutrient availability, and mitochondrial function (Saxton and Sabatini, 2017). In an effort to investigate whether mitochondrial alterations contribute to Paneth cell mTORC1 inhibition in response to IFN-γ, we first analyzed mitochondrial integrity in Paneth cells in vivo during T. gondii infection. Flow cytometric analysis revealed that by day 5 post infection, when Paneth cell loss is starting to become noticeable (Raetz et al., 2013), the remaining Paneth cells while retaining MitoTracker positivity, which allows identification of all mitochondria independent of their membrane potential (Figure 6A and B, Figure 6—source data 1), were significantly compromised by day 5 post infection in an IFN-γ-dependent manner, as assessed by MitoStatus staining, which detects functionally polarized mitochondria (Figure 6A,C). These results were in the full agreement with the previous electron microscopy experiments that revealed extensive damage and vacuolization of the mitochondria in Paneth cells during T. gondii infection (Raetz et al., 2013).

To further investigate the effects of IFN-γ on mitochondrial functions in Paneth cells, intestinal organoids were generated from WT mice and treated with recombinant IFN-γ. This in vitro system revealed that IFN-γ administration resulted in rapid mitochondrial depolarization, which was followed by the loss of Paneth cells (Figure 6D, F). These data suggest that IFN-γ is responsible for the compromised mitochondrial integrity in Paneth cells that leads to mTORC1 inactivation. To formally examine whether the mitochondrial changes downstream of IFN-γ are sufficient to trigger mTORC1 inhibition in Paneth cells, intestinal organoids were treated with FCCP, a potent uncoupler of oxidative phosphorylation in the mitochondria that disrupts ATP synthesis. We established that, similar to IFN-γ and rapamycin treatment, FCCP addition led to a rapid loss of ribosomal S6 phosphorylation (Figure 6G). These experiments revealed that disruption of mitochondrial function is sufficient to induce the loss of TORC1 activity and subsequent Paneth cell death (Figure 6G and H, Figure 6—source data 1). Importantly, rapamycin treatment of intestinal organoids did not result in loss of the inner mitochondrial membrane potential, indicating that mTORC1 functions downstream of mitochondrial functions in Paneth cells (Figure 6I).

The intestinal microbiota has a profound effect on intestinal homeostasis (Hooper and Macpherson, 2010; Garrett et al., 2010; Belkaid and Harrison, 2017). We have previously observed that the presence of the microbiota is required for Paneth cell loss during acute infection with T. gondii (Raetz et al., 2013), even though the precise understanding of microbiota-mediated effects on Paneth cell death remains incomplete. To examine a potential role for the intestinal microbiota in modulating the IFN-γ-mediated inhibition of mTORC1 during acute T. gondii infection, we first analyzed mTOR activity in naïve microbiologically sterile (germ-free [GF]) mice. We observed that naïve germ-free mice exhibited Paneth cell mTOR activity similar to that in conventional (CONV) mice, as was evident from comparable staining for ribosomal protein S6 phosphorylation in the small intestine between CONV and GF mice (Figure 7, Figure 7—source data 1). Thus, the presence of the microbiota is not required for basal mTOR activity in Paneth cells. In contrast, while parasitic infection resulted in the rapid loss of Paneth cell mTOR activity in CONV mice, GF mice retained phosphorylated ribosomal protein S6 at all examined time points post infection (Figure 7B and C, and data not shown), strongly suggesting that the presence of the intestinal microbiota is required for the loss of mTOR activity in Paneth cells. As a result, Paneth cells were largely retained in GF mice, and only a mild loss of Paneth cells was observed in the GF mice compared with CONV mice (Figure 7).

To test if microbiota-dependent loss of Paneth cells was due to the contribution of the intestinal bacteria to the potent IFN-γ responses caused by the parasitic infection (Figure 7—figure supplement 1A), GF mice were injected with the recombinant IFN-γ and Paneth cells were next analyzed in those mice. We observed that GF mice treated with rIFN-γ (Figure 7—figure supplement 1B) lost Paneth cell specific transcripts similarly to CV mice (Figure 7—figure supplement 1C), ruling out an intrinsic resistance of Paneth cells to IFN-γ in the absence of the intestinal bacteria. Thus, the reduced levels of IFN-γ seen in GF mice infected with T. gondii (Figure 7—figure supplement 1A) are responsible for the survival of Paneth cells in the absence of microbiota. Altogether, these data show that the microbiota is required for IFN-γ-mediated mTORC1 dysfunction and the subsequent death of Paneth cells during acute gastrointestinal infection.

C57BL/6 mice were originally purchased from the Jackson Laboratories and maintained in the pathogen-free animal facility at the University of Rochester of Medicine and Dentistry. Ifngr2^flox/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 (Tcyganov et al., 2021).

Rptor^fl/fl (B6.Cg-Rptortm1.1Dmsa/J) and Mtor ^fl/fl (B6.129S4-Mtortm1.2Koz/J) mice were purchased from Jackson Laboratories and also crossed to Defa4 IRES-Cre mice (Burger et al., 2018) to generate PC-Raptor or PC-Mtor conditional knockouts. Mice for all experiments were age- and sex-matched. This study included both male and female mice, and the data derived from male and female mice identified no sex-specific differences in the performed experiments. GF C5BL/6 mice were bred and maintained at the University of Rochester School of Medicine and Dentistry. 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).

Me49 strain T. gondii tissue cyst (bradyzoite) stages were maintained through serial passage in Swiss Webster mice. For infections, brains of chronically infected mice were mechanically homogenized by passage through a series of 18-gauge, 20-gauge, and 22-gauge needles. Experimental mice were orally infected with 20 T. gondii brain cysts (ME49 strain). Portions of small intestine were fixed in Carnoy’s fixative, embedded in paraffin, and stained with Alcian blue/Periodic Acid Schiff (PAS). Paneth cells were identified based on their morphology of large granule-containing cells and a basolateral nucleus at the base of the intestinal crypt. Quantification of Paneth cells per intestinal crypt was performed in a double blinded manner.

For immunofluorescence images, small intestines were divided into 16 equal parts from duodenum to distal illeum and segment 8 was flushed and flash-frozen in OCT compound (Tissue Tek). 8 µm intestine sections were fixed with Foxp3/Transcription Factor Staining Buffer Set (eBioscience, Cat# 00-5521-00) for 15 min, permeabilized with 0.25% Triton-X in phosphate-buffered saline (PBS), and blocked with 0.1% Triton-X containing 5% Normal Goat Serum in PBS. Sections were stained with primary antibodies against Lysozyme 1 (rabbit polyclonal; Dako) or Phospho-S6 Ribosomal Protein (Ser235/236) (Rabbit mAb#4858; Cell Signaling Technology) in 0.1% Triton-X/PBS with 5% Normal Goat Serum overnight at 4°C. Slides were washed and incubated with donkey anti-rabbit Alexa 568 (Molecular Probes), goat anti-rabbit Alexa 488 (Molecular Probes), and UEA for 1 hr in PBS at room temperature. Slides were mounted in ProLong Gold (Molecular Probes). Specimens were imaged with a Leica SPE system (Leica DMi8) fitted with a Leica 63× objective NA 1.4. or in an inverted Olympus IX81 60×/1.42 Oil objective.

For electron microscopy analysis, the small intestines were fixed in 2.5% glutaraldehyde in 0.1 M sodium cacodylate, followed by 1% osmium tetroxide in 0.1 M sodium cacodylate. The samples were embedded in epoxy resin (University of Rochester Electron Microscopy Core) and polymerized at 70°C. Ultrathin sections were cut at 70 nm and stained with uranyl acetate and lead citrate. Sections were examined at 120 kV with a Hitachi 7650 Analytical TEM with an Erlangshen 11 MP digital camera and Gatan software for imaging and morphometric analysis.

In some experiments, mice were injected i.p. with 200 μg anti-IFN-γ (clone XMG1.2, BioXCell) on days 0, 2, 4, and 6 post T. gondii infection or 50 μg of recombinant mouse IFN-γalone every 8 hr for 3 consecutive days. In some experiments, mice were intravenously injected with PI on day 6 post T. gondii infection for 20 min prior analysis. Quantification of PI positive puncta was performed in a double-blinded manner. For rapamycin experiments, rapamycin (Sigma-Aldrich) was administered orally daily for a max of 7 days at 7.5 mg/kg in 5% PEG-400/5% Tween-80 and reconstituted in 100% EtOH at 10 mg/ml. Necrostatin was administered at (4.5 mg/kg) IP twice daily during T. gondii infection.

Intestinal organoids were generated from the small intestine as previously described (Sato et al., 2009). Briefly, the proximal small intestine was dissected, rinsed with PBS, minced, and incubated in 2.5 mM EDTA in PBS for 30 min at 4°C with gentle agitation. This mixture was passed through a 70 mm cell strainer and pelleted. Isolated crypts were resuspended in Matrigel and plated in 50 μl drops on pre-warmed wells in a 24-well plate. After polymerization, 500 ml of complete growth medium (Advanced DMEM/F-12 supplemented with EGF [Peprotech], Noggin [Peprotech], N2 [Life Technologies], B27 [Life Technologies], Glutamax [Life Technologies], Penicillin/Streptomycin [Life Technologies], HEPES [Life Technologies], N-Acetylcysteine [Sigma-Aldrich], and R-Spondin) was added and refreshed every 2–3 days. Organoids were stimulated by addition of recombinant murine IFN-γ (200 ng/ml; R&D Systems), Carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (50 μM; Sigma=Aldrich), or rapamycin (1 μM; Sigma-Aldrich) for 2, 6, or 24 hr. For immunofluorescence images, organoids were either imaged live after the addition of Hoesch (Sigma-Aldrich) and Mitostatus red (BD Biosciences) or underwent intracellular staining after brief fixation with 4% PFA and permeabilization with 0.25% Triton-X. For intracellular staining, samples were incubated with primary antibody against Phospho-S6 Ribosomal Protein (Ser235/236) (Rabbit mAb#4858; Cell Signaling Technology) overnight at 4°C, washed, and incubated for 1 hr with goat anti-rabbit Alexa 488 (Molecular Probes) and DAPI (Molecular Probes). Images of crypt organoids were taken using a Leica DMi8 using the long range 40× objective.

Total RNA was isolated from the small intestines or organoid well using TRIzol (Life Technologies) and subjected to first-strand cDNA synthesis using iScript Reverse Transcription Supermix for RT-qPCR (Bio-Rad). Real-time PCR was performed using Ssofast EvaGreen Supermix (Bio-Rad) as previously described. The relative expression of each sample was determined after normalization to housekeeping gene HPRT using the ddCt method. The following primers were used for analysis of the gene expression: IFN-γ 5′-ACTGGCAAAAGGATGGTGAC-3′ and 5′-TGAGCTCATTGAATGCTTGG-3′; TNF 5′-GCCTCTTCTCATTCCTGCTTGT-3′ and 5′- GGCCATTTGGGAACTTCTCAT-3′; TNFR2 5′-CAGGTTGTCTTGACACCCTAC-3′ and 5′-GCACAGCACATCTGAGCCT-3′; CXCL10 5′- GACGGTCCGCTGCAACTG-3′ and 5′- GCTTCCCTATGGCCCTCATT-3′; IRGM1 5′- AGCCAACGAGTCCTTGAAGA-3′ and 5′-GCACATGTCATCAGCCTCAG-3′; IRGM3 5′- CTGGAGGCAGCTGTCAGCTCCGAG-3′ and 5′-GTCCTTTAGAGCTTTCCTCAGGGAGGTCTTG-3′; CXCL10 5′- GACGGTCCGCTGCAACTG-3′ and 5′- GCTTCCCTATGGCCCTCATT-3′; IDO1 5′-GTGGGCAGCTTTTCAACTTC-3′ and 5′- GGGCTTTGCTCTACCACATC-3′, CCL2 5′- CAGGTCCCTGTCATGCTTCT-3′ and 5′- GAGTGGGGCGTTAACTGCA-3′; SDHA 5′- GCTCCTGCCTCTGTGGTTGA-3′ and 5′-AGCAACACCGATGAGCCTG-3′; and IL-1b 5′- GCCCATCCTCTGTGACTCAT-3′ and 5′- AGGCCACAGGTATTTTGTCG-3′.

Paneth cell transcripts were additionally validated with the following primers that allowed for the selective analysis of Defa23 (5′-TCTGGTATGCTATTGTAGAAC-3′ and 5′-GACAGCAGAGCGTGTATA-3′) and Defa24 (5′-GATCTGGTATGCTATTGTAGAG-3′ and 5′-GACAGCAGAGCATGTACAA-3′) as recently described (Castillo et al., 2019).

Small intestine segments were washed and incubated in HBSS with 30 mM EDTA for 20 min. Epithelial cells were dissociated mechanically through vigorous shaking followed by incubation in Triple-X (Thermo Fisher Scientific) supplemented with Y-27632 (10 mM, Sigma-Aldrich) and 0.5 mM N-acetyl-L-cysteine at 37°C. Dissociated enterocytes were filtered and pelleted before re-suspended in complete advanced DMEM/F12 media containing 10 mM Y-27632 and 0.5 mM N-acetylcysteine. Cells were stained with fluorochrome-conjugated surface antibodies according to the manufacturer’s instructions. For intracellular staining, cells were fixed and permeabilized for 1 hr at 4°C with the Foxp3/ Transcription Factor Staining Buffer Set according to the manufacturer’s instructions (eBioscience). The following antibodies were used for surface or intracellular staining: CD45 BUV393 (BD Cat#565967), Epcam BV421 (BioLegend, Cat# 324219), p-S6 (S235/236) PerCP e710 (eBioscience, Cat# 46900742), p-mTOR eFluor 660 9S24480 (eBioscience, Cat# 50971842), CD4 BV786 (BD, Cat#563727), IFN-γ BV421 (BD, 563376), Lysozyme 1 (rabbit polyclonal; Dako), and goat anti-rabbit Alexa 488 (Molecular Probes). Cell fluorescence was measured using an LSRII flow cytometer, and data were analyzed using FlowJo software (Tree Star).

IFN-γ, TNF, CCL2, and IL-12p40 concentrations in the sera were analyzed by standard sandwich ELISA Kit according to manufacturer’s instructions (eBioscience).

All data were analyzed with Prism (version 6; GraphPad). These data were considered statistically significant when p-values were less than 0.05 by two-tailed t-test or two-way ANOVA using Tukey’s multiple comparisons tests or one-tailed t-tests. Image data sets were processed using Leica Advanced Fluorescence software (Leica) and ImageJ software (National Institutes of Health).