Tet2 deficiency drives liver microbiome dysbiosis triggering Tc1 cell autoimmune hepatitis

SUMMARY

The triggers that drive interferon-γ (IFNγ)-producing CD8 T cell (Tc1 cell)-mediated autoimmune hepatitis (AIH) remain obscure. Here, we show that lack of hematopoietic Tet methylcytosine dioxygenase 2 (Tet2), an epigenetic regulator associated with autoimmunity, results in the development of microbiota-dependent AIH-like pathology, accompanied by hepatic enrichment of aryl hydrocarbon receptor (AhR) ligand-producing pathobionts and rampant Tc1 cell immunity. We report that AIH-like disease development is dependent on both IFNγ and AhR signaling, as blocking either reverts ongoing AIH-like pathology. Illustrating the critical role of AhR-ligand-producing pathobionts in this condition, hepatic translocation of the AhR ligand indole-3-aldehyde (I3A)-releasing Lactobacillus reuteri is sufficient to trigger AIH-like pathology. Finally, we demonstrate that I3A is required for L. reuteri-induced Tc1 cell differentiation in vitro and AIH-like pathology in vivo, both of which are restrained by Tet2 within CD8 T cells. This AIH-disease model may contribute to the development of therapeutics to alleviate AIH.

Graphical Abstract

In brief

In a mouse model of autoimmune hepatitis (AIH), Pandey et al. find that liver microbiome dysbiosis, characterized by an expansion of aryl hydrocarbon receptor (AhR) agonist-producing Lactobacilli, promotes CD8 T cell-dependent AIH. The epigenetic enzyme Tet methylcytosine dioxygenase 2 (Tet2) within CD8 T cells blocks AhR-agonist-induced AIH.

INTRODUCTION

Autoimmune hepatitis (AIH) is an autoinflammatory liver disorder that often becomes refractory to immunosuppressants—the sole therapeutic option for AIH patients (Manns et al., 2010)—and progresses to end-stage liver disease in the absence of treatment (Bataller and Brenner, 2005). Despite previous work suggesting genetic (Czaja and Donaldson, 2000; Krawitt, 2006) and environmental factors (Czaja, 2019) play a role in AIH development, mechanisms underlying disease initiation remain enigmatic due to the lack of suitable murine models. Hepatic inflammation, mediated by interferon-γ (IFNγ)-producing TCRβ CD4 and cytotoxic CD8 T cells (T helper [Th1] cells or Tc1 cells, respectively), is the main driver of hepatic tissue damage and sets the stage for overt AIH (Assis, 2020). In particular, Tc1 cells are considered to play an indispensable role in driving preclinical AIH-like pathology (Ikeda et al., 2014; Kido et al., 2008; Longhi et al., 2007; Watanabe et al., 1996; Zhang et al., 2019), and a histological hallmark of human AIH is global lymphocyte infiltration, dominated by cytotoxic TCRβ CD8 T cells (Assis, 2020; Eggink et al., 1982; Hashimoto et al., 1993; Ichiki et al., 2005; Renand et al., 2018; Tordjmann et al., 1998). However, the underlying mechanisms that trigger rampant hepatic Tc1 cell effector function and consequently drive AIH remain enigmatic.

Hematopoietic deficiency of the key epigenetic regulator Tet methylcytosine dioxygenase 2 (TET2), which occurs in around 10% of healthy individuals over 50 years of age, increases the risk of developing cancer (Busque et al., 2012; Delhommeau et al., 2009; Kosmider et al., 2009) and autoinflammatory disorders such as rheumatoid arthritis (Savola et al., 2018). Highlighting a key role for TET2 in autoimmune dysfunction, germline deficiency of TET2 has been associated with the development of autoinflammatory pathology (Stremenova Spegarova et al., 2020), and lack of Tet2 in TCRβ T cells increases the susceptibility to develop preclinical multiple sclerosis-like disease (Ichiyama et al., 2015). In addition, deficiency of Tet2 and Tet3 in either B cells or CD4 T cells promoted the development of lupus-like pathology (Tanaka et al., 2020) and fueled autoinflammatory CD4 T cell responses (Yue et al., 2016, 2019), respectively. Although the vast majority of previous studies describe Tet2 as a critical modulator in CD4 T cell function (Ichiyama et al., 2015; Nakatsukasa et al., 2019; Yue et al., 2016, 2019), emerging evidence highlights a key role of TET2 in shaping human (Fraietta et al., 2018) and murine (Carty et al., 2018) cytotoxic CD8 T cell fate via suppressing effector cytokine responses. Furthermore, it has been shown by us (McPherson et al., 2021; Meisel et al., 2018) and others (Fuster et al., 2017; Zhang et al., 2015) that loss of hematopoietic Tet2 invokes a poised cellular state—in a cell-autologous manner—that is highly sensitive to environmental, including microbial, signals and ultimately drives context-dependent inflammatory or autoinflammatory immune responses.

Systemic expansion of aryl hydrocarbon receptor (AhR) ligand-producing pathobionts is found to be associated with autoimmune inflammation in mice (Choi et al., 2020; Vieira et al., 2018; Sonner et al., 2019) and AIH in humans (Vieira et al., 2018). These observations together with the finding that experimental AIH requires the microbiota (Wei et al., 2016) illustrate the urgency to define if and how AhR ligand-releasing pathobionts impact the etiopathogenesis of AIH, given that such mechanisms potentially provide therapeutic targets for prevention or intervention in the development of AIH.

The AhR, a ligand-activated transcription factor, senses both endogenous and exogenous ligands, including microbial-derived ligands, and has been found to play a critical role in orchestrating CD4 T cell differentiation (Cervantes-Barragan et al., 2017; Quintana et al., 2008; Stockinger et al., 2014). However, whether and how microbial-induced AhR signaling affects CD8 T cell effector function remains to be defined.

In this study, we uncover a critical role for the hepatic microbiota in facilitating AhR-ligand-mediated Tc1 cell-driven AIH-like pathology development in an AIH-like disease model. Collectively, this work highlights the interplay between the liver microbiota, Tc1 cell immunity, and autoimmunity in the context of AIH.

RESULTS

Tet2 deficiency in hematopoietic cells drives spontaneous hepatic pathology

We have recently shown that loss of hematopoietic Tet2 (Tet2^ΔVAV) is associated with the microbial-mediated systemic expansion of CD11b⁺Gr1⁺ myeloid cells in 50%–70% of >20-week-old mice (Meisel et al., 2018). This systemic myelo-expansion is a shared pathological feature of Tet2-deficiency (Abdel-Wahab and Levine, 2013; Delhommeau et al., 2009; Meisel et al., 2018; Moran-Crusio et al., 2011; Quivoron et al., 2011) and inflammatory liver disease (Cripps et al., 2010; Diao et al., 2014; Li et al., 2015) in mice and humans. These observations, in conjunction with the finding that myelo-expansion in humans with TET2 deficiency is associated with signs of hepatic pathology such as hepatomegaly and hepatic injury (Barraco et al., 2017; Pereira et al., 1988; Tefferi, 2016; Traina et al., 2012), led us to hypothesize that myelo-expansion in Tet2^ΔVAV mice is linked to hepatic pathology. Assessment of hepatic injury in aged (20 weeks old) Tet2^ΔVAV mice with (CD11b⁺Gr1⁺ frequency >16% in peripheral blood [PB]) or without (CD11b⁺Gr1⁺ frequency ≤ 16% in PB; referred to as symptom-free) myelo-expansion and Tet2 cre⁻ littermates (Figure S1A; Meisel et al., 2018) revealed a significant correlation between myelo-expansion and the activity of plasma alanine aminotransferase (ALT), a key marker of hepatocyte injury (Fouts et al., 2012; Figure 1A). Unbiased transcriptomic profiling of livers from aged Tet2^ΔVAV mice and Tet2 cre⁻ littermates demonstrated that compared with WT littermates, Tet2^ΔVAV mice with, but not those without, myelo-expansion displayed global alterations of transcription (Figures 1B, S1B, and S1C; Tables S1 and S2), including a substantial upregulation of genes indicative of hepatic fibrosis and inflammation (Figure 1C). Additionally, Tet2^ΔVAV mice with myelo-expansion displayed cardinal signs of chronic hepatic pathology characterized by hepatomegaly (Figure 1D), hepatic fibrosis (Figures 1E and 1F), hepatic inflammation and immune cell infiltration (Figures 1F and 1G), hepatocyte injury (Figure 1H), and cell death (Figure 1I) relative to symptom-free Tet2^ΔVAV mice and WT littermates.

Figure 1. Tet2^ΔVAV mice develop spontaneous hepatic pathology with age.

>20-week-old Tet2^ΔVAV mice and Tet2 cre⁻ littermates were used.

(A) Correlation between CD11b⁺Gr1⁺ cell frequency and ALT activity. Pearson correlation test of Tet2^ΔVAV mice, p value shown.

(B and C) RNA-seq of whole liver.

(B) Venn diagram.

(C) Heatmap of selected DEGs, average is shown.

(D) Liver weight.

(E) Representative trichrome staining of livers (left) and quantification of area of fibrosis (right). Scale bars, 100 μm.

(F) Hepatic gene expression.

(G) Number of CD45⁺ lymphocytes/g liver tissue quantified by flow cytometry.

(H) ALT activity.

(I) Representative images (left) and quantification (right) of TUNEL⁺ cells in livers; FOV, field of view. Scale bars, 20 μm.

(B and C) n = 4/5 mice/group. (D–I) Mean ± SEM, one-way ANOVA, Sidak’s post hoc test. Data represent three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Taken together, our data demonstrate that a fraction of Tet2^ΔVAV mice spontaneously develop age-related hepatic pathology and that myelo-expansion is a predictive marker of ongoing hepatic pathology in Tet2^ΔVAV mice.

Hepatic disease in Tet2^ΔVAV mice resembles autoimmune hepatitis-like pathology

Further computational assessment of hepatic transcriptomes of Tet2^ΔVAV mice with myelo-expansion relative to WT littermates uncovered a highly significant enhancement of gene-sets associated with systemic autoimmunity (p = 1.54e–120) (Figure S1D; Table S3) and revealed a significant upregulation of several pathways associated with AIH (Sucher et al., 2019), such as T cell-mediated cytotoxicity (p = 1.13e–118) and positive regulation of IFNγ production (p = 8.56e–47) (Figures S1E and S1F; Table S4). This observation, together with the finding that Tet2^ΔVAV mice spontaneously develop hepatic pathology in the absence of an overt pathological trigger, led us to hypothesize that there might be an autoimmune component in their hepatic disease etiology. In addition to common signs of hepatic pathology, AIH in mice (Cripps et al., 2010; Diao et al., 2014) and humans (Krawitt, 2006; Li et al., 2015) is characterized by elevated plasma nonorgan- and organ-specific autoantibodies, including antinuclear antibodies (ANAs) (Hennes et al., 2008). Interestingly, Tet2^ΔVAV mice with myelo-expansion, in contrast to symptom-free Tet2^ΔVAV mice and WT littermates, displayed significantly elevated blood plasma ANA levels (Figure 2A). Similar to observations in patients with AIH (Li et al., 2015), myelo-expansion in Tet2^ΔVAV mice significantly correlated with both ANA plasma levels (Figure 2B) as well as liver injury (Figure 1A). Characterization of blood plasma ANA profiles by using ANA-reactive HEp2 cells (Tilstra et al., 2020) revealed that ANA profiles were significantly altered in Tet2^ΔVAV mice with myelo-expansion (Figure 2C), mirroring ANA alterations typically documented in autoimmune-mouse models (Christensen et al., 2005; Nickerson et al., 2010; Tilstra et al., 2020). Of note, both symptom-free Tet2^ΔVAV mice and WT littermates did not show aberrant ANA profiles (Figure 2C). Other autoantibodies found to be upregulated in plasma of preclinical and clinical AIH are antiliver-kidney microsomal (anti-LKM1) antibodies (Abs), antisoluble liver antigen (anti-SLA) Abs, and immunoglobulin G (IgG) (Hennes et al., 2008). Remarkably, anti-LKM1 Abs (Figure 2D), anti-SLA Abs (Figure 2E), and IgG levels (Figure 2F) were significantly increased within the plasma of Tet2^ΔVAV mice with myelo-expansion relative to symptom-free Tet2^ΔVAV mice and WT littermates. Assessment of our model for characteristic histological features of human AIH (Hennes et al., 2008) revealed that Tet2^ΔVAV mice with myelo-expansion display key histological features of human AIH including interface hepatitis, lymphoplasmacytic infiltration, rosette formation, and emperipolesis, highlighting the translational relevancy of our model (Figures 2G and 2H). Furthermore, confirmation of hepatic plasma cell (PC) infiltration by flow cytometry revealed a significant expansion of CD138⁺ PCs in our AIH-like disease model (Figure 2I). Characterization of the hepatic immune response uncovered that Tet2^ΔVAV mice with myelo-expansion, in stark contrast to symptom-free Tet2^ΔVAV mice and WT littermates, displayed significantly elevated hepatic Tc1 and Th1 cell frequencies (Figure 2J). Notably, we failed to detect gross differences in hepatic IFNγ production of CD11b⁺Gr1⁺ cells, CD11b⁺ F4/80⁺ Kupffer cells, natural killer (NK) cells, NK T cells, or γδ T cells across groups, implying a TCRβ T cell specific type 1 immune response in our AIH-like disease model (Figure S2A). Although elevated Th17 cell numbers have been found in the PB of AIH patients relative to healthy controls (Zhao et al., 2011), the exact role of Th17 cells in AIH-like disease remains ill-defined as contrasting results have been described (Nagata et al., 2008; Zenewicz et al., 2007). In our AIH-like disease model, Th17 cell effector function remained unaffected (Figure S2B). In line with findings in clinical (Peiseler et al., 2012; Taubert et al., 2014) and preclinical AIH (Bochtler et al., 2008), hepatic Foxp3⁺ regulatory T cell (Treg) frequencies were significantly increased in AIH-like diseased Tet2^ΔVAV mice relative to symptom-free Tet2^ΔVAV mice and WT littermate controls (Figure S2C). Concordant with the dominant type 1 T cell immune response on a transcriptional level, we observed a significant upregulation of type 1 immune promoting transcription factors, such as T box transcription factor 21 (Tbet), PR domain 1 (Blimp1), Interferon regulatory factor 4 (Irf4), Signal transducer and activator of transcription 1 (Stat1) and Stat4, as well as type 1 immune effector cytokines Ifng and granzyme B (Gzmb), in livers of Tet2^ΔVAV mice with AIH-like pathology relative to symptom-free Tet2^ΔVAV mice and littermate controls (Figure S2D). Although interleukin-12 (IL-12)-producing antigen presenting cells (APCs) are key drivers of type 1 T cell immunity (Murphy et al., 1995), we found that the elevated type 1 T cell effector response in AIH-like diseased Tet2^ΔVAV mice could not be explained by an increase in IL-12 production by hepatic APCs (Figure S2E). In accordance with the increased probability of developing AIH with age (Peng et al., 2014; Schramm and Lohse, 2014; Tanaka, 2020), young (6 weeks old) Tet2^ΔVAV mice did not display signs of AIH (Figures S2F-S2N). To further define the AIH-specificity of our model, we interrogated the susceptibility of Tet2^ΔVAV mice to AIH-specific and non-AIH-specific hepatic injury. We found that young Tet2^ΔVAV mice displayed increased susceptibility to concanavalin A (ConA)-induced hepatitis, a well-accepted AIH model (Küsters et al., 1996; Ye et al., 2018; Figures 2K-2N), yet failed to display increased susceptibility to CCl₄-mediated hepatic injury (Scholten et al., 2015; Figures S2O-S2R), demonstrating that Tet2^ΔVAV mice display a selective susceptibility to AIH-like disease development, rather than a general susceptibility to hepatic injury.

Figure 2. Tet2^ΔVAV mice develop AIH-like pathology.

(A–J) Mice from Figure 1 cohort were used. (A) ANA levels.

(B) Correlation between CD11b⁺Gr1⁺ cell frequency and ANA levels. Pearson correlation test of Tet2^ΔVAV mice, p value shown.

(C) Representative images (top, left), scores (top, right), and pie charts of plasma ANA reactivity by Hep2 cells.

(D–F) Plasma levels of (D) LKM1, (E) SLA, and (F) total IgG.

(G) Representative H&E staining of livers. Scale bars, 100 μm.

(H) AIH-histology score. Each dot represents an independent mouse.

(I and J) Representative flow cytometry plots (left) and quantification (right) of hepatic (I) CD138⁺ PCs and (J) IFNγ-producing CD8 T cells (top) and CD4 T cells (bottom).

(K–N) 6-week-old Tet2^ΔVAV and Tet2 cre⁻ mice treated with ConA or PBS.

(K) ALT activity.

(L) Representative trichrome staining of livers. Scale bars, 100 μm.

(M and N) Hepatic gene expression.

(A, D–F, I–K, M, and N) Mean ± SEM, one-way ANOVA, Sidak’s post hoc test. (C and H) Kruskal-Wallis test, Dunn’s multiple comparisons test. Data represent at least two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Taken together, these results demonstrate that Tet2^ΔVAV mice spontaneously develop AIH-like pathology characterized by dominant type 1 T cell immunity.

IFNγ drives AIH-like pathology in Tet2^ΔVAV mice

IFNγ is a key player in experimental AIH models (Gorham et al., 2001; Küsters et al., 1996; Nicoletti et al., 2000) and human AIH (Bovensiepen et al., 2019). To test the causal role of IFNγ in our AIH-like disease model, we treated Tet2^ΔVAV mice with ongoing AIH, displayed by myelo-expansion, and WT littermates with a neutralizing IFNγ Ab (Figures S3A and S3B). Although isotype control Ab-treated Tet2^ΔVAV mice showed overt AIH-like pathology (Figures 3A-3G), neutralizing IFNγ reversed AIH-like pathology in Tet2^ΔVAV mice, as displayed by the absence of hepatic fibrosis, hepatic immune cell infiltration and inflammation (Figures 3A and 3B), hepatocyte injury (Figure 3C), restoration of Ifng mRNA levels (Figure 3D), hepatic Th1 and Tc1 cell frequencies (Figure 3E) and ANA plasma levels (Figure 3F), and the absence of hepatocyte cell death (Figure 3G). In accordance, aged Tet2^−/−Ifng^−/− mice, similar to Tet2^+/+Ifng^−/− littermates, did not develop AIH-like pathology (Figures S3C-S3H). Taken together, these data demonstrate that IFNγ is required for the development of AIH-like pathology in Tet2^ΔVAV mice.

Figure 3. AIH-like pathology development in Tet2^ΔVAV mice requires IFNγ.

>20-week-old Tet2^ΔVAV mice with AIH-like disease and Tet2 cre⁻ littermates treated with anti-IFNγ (see Figures S3A and S3B).

(A) Representative H&E (top) and trichrome (bottom) liver stain (left) and quantification of fibrosis area (right). Scale bars, 100 μm.

(B) Hepatic gene expression.

(C) ALT activity.

(D) Hepatic gene expression.

(E) Representative flow cytometry plots and quantification of hepatic IFNγ-producing CD4 T cells (left) and CD8 T cells (right).

(F) ANA levels.

(G) Representative image (left) and quantification of TUNEL⁺ cells in livers (right); FOV, field of view. Scale bars, 20 μm.(A-G) Mean ± SEM is shown. One-way ANOVA, Sidak’s post hoc test. Data represent two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Deficiency of Tet2 in T cells alone is sufficient to drive AIH-like pathology development

We found the absence of Tet2 in myeloid cells, through the use of Tet2^fl/fl LysM cre⁺ mice, was insufficient to drive AIH-like disease (Figures 4A-4E), and although reduced expression of TET2 in hepatocytes is associated with hepatic fibrosis (Wang et al., 2019; Xu et al., 2020), Tet2^fl/fl Alb cre⁺ mice failed to develop signs of AIH-like pathology with age (Figures 4F-4J). Concordant with the key role of T cells in AIH (Senaldi et al., 1992; Tiegs et al., 1992; Zhang et al., 2019), we identified that aged Tet2^fl/fl CD2 cre⁺ mice, in contrast to littermate controls, developed AIH-like pathology, displayed by substantial hepatic fibrosis and immune cell infiltration (Figure 4K), increased ALT levels (p = 0.055) (Figure 4L), significantly elevated ANA plasma levels (Figure 4M) and aberrant ANA profiles (Figure 4N), and significantly increased plasma anti-SLA Abs (Figure 4O) and IgG levels (Figure 4P). These key features of AIH were accompanied by a significantly increased frequency of hepatic Tc1, but not Th1, cells (Figure 4Q), suggesting that Tc1 cells are key drivers of AIH-like pathology. Further confirming the crucial role of T cells in our AIH model, although aged Tet2^−/−Rag^−/− mice exhibited significant myelo-expansio, they failed to develop AIH-like pathology (Figures S4A-S4F). These data reveal that Tet2 deficiency in T cells, but neither in myeloid cells nor hepatocytes, is sufficient to drive AIH-like pathology.

Figure 4. Deficiency of Tet2 in T cells, but neither in myeloid cells nor hepatocytes, is sufficient to drive the development of AIH-like pathology.

>20-week-old (A–E) Tet2^fl/fl LysM cre⁺, (F–J) Tet2^fl/fl Alb cre⁺, or (K–Q) Tet2^fl/fl CD2 cre⁺ mice and corresponding WT littermates were used.

(A, F, and K) Representative H&E (top) and trichrome (bottom) liver stain (left) and quantification of fibrosis area (right). Scale bars, 100 μm.

(B, G, and Q) Quantification of IFNγ-producing CD4 T cells (left) and CD8 T cells (right).

(C, H, and L) ALT activity.

(D, I, and M) ANA levels.

(E, J, and P) Total IgG levels.

(N) Representative images (top, left), scores (top, right), and pie charts of plasma ANA reactivity by Hep2 cells.

(O) SLA levels.

(A–M and O–Q) Mean ± SEM is shown.

(K–M and O–Q) Two-tailed unpaired t test.

(N) Mann-Whitney test.

(A-Q) Data represent at least two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

The microbiota is required, and liver microbiome dysbiosis is linked to the development of AIH-like pathology in Tet2^ΔVAV mice

Given that livers of Tet2^ΔVAV mice with AIH-like disease harbor a transcriptional profile indicative of defense responses against microbial stimuli (Figure S1E; Table S4) and the reported requirement of the microbiota for the development of experimental AIH (Wei et al., 2016), we tested the requirement of the microbiota in our AIH model. Accordingly, aged germ-free (GF) Tet2^−/− mice, similar to WT littermates, failed to display signs of AIH-like pathology (Figures 5A-5H). Liver microbiome dysbiosis has been identified in AIH patients relative to healthy controls (Vieira et al., 2018). Thus, we assessed the liver microbiome composition of symptomatic Tet2^ΔVAV mice (myelo-expansion and elevated ANA plasma) compared with age-matched symptom-free Tet2^ΔVAV mice and WT littermates using two independent but complementary approaches (Figure S5A). First, we sterilely cultured livers and performed 16S ribosomal RNA sequencing (16S rRNA-seq) of single colonies; second, we performed 16S rRNA-seq of hepatic microbial DNA. Intriguingly, profiling of the liver microbiota revealed a significant increase in microbial diversity in AIH-like diseased Tet2^ΔVAV mice in both independent approaches (Figures 6A-6C and S5B; Table S5). Livers of AIH-like diseased Tet2^ΔVAV mice harbored an increased diversity in translocated bacteria, such as Lactobacillus reuteri (L. reuteri), L. johnsonii, Enterococcus faecalis, and Escherichia coli when compared with symptom-free Tet2^ΔVAV mice and WT littermate controls (Figure 6A). The increased microbial alpha-diversity identified by 16S rRNA-seq of hepatic microbial DNA (Figure 6B), predominately driven by the genera Lactobacillus, Sutterella, Lachnospiraceae, and Prevotella (Figure 6C), significantly correlated with both ANA plasma levels (Figure 6D) and hepatic IFNγ-producing T cell frequencies in Tet2^ΔVAV mice (Figure 6E), indicating that liver dysbiosis may drive AIH-like disease. The significantly increased hepatic bacterial load in AIH-like diseased Tet2^ΔVAV mice relative to WT littermates (Figure 6F) suggests that expansion of pathobionts is implicated in AIH-like disease. Of note, Tet2^ΔVAV mice with both AIH-like disease and liver microbiome dysbiosis failed to display gross differences in their small and large intestinal microbial composition and bacterial load relative to symptom-free Tet2^ΔVAV mice and WT littermates (Figures S5C-S5F), implying that gut dysbiosis does not drive AIH-like pathology in Tet2^ΔVAV mice. Taken together, these data demonstrate that the microbiota is required for and that liver microbiome dysbiosis is linked to AIH-like disease development in Tet2^ΔVAV mice.

Figure 5. The microbiota is required for the development of AIH-like pathology in Tet2-deficient mice.

25-week-old germ-free (GF) Tet2^−/− and Tet2^+/+ littermates were used.

(A) Liver weight.

(B) Representative H&E (top) and trichrome (bottom) liver stain (left) and quantification of fibrosis area (right). Scale bars, 100 μm.

(C) Hepatic gene expression.

(D) ALT activity.

(E) Hepatic gene expression.

(F) Quantification of hepatic IFNγ-producing CD4 T cells (left) and CD8 T cells (right).

(G) ANA levels.

(H) Total IgG levels.

(A–H) Mean ± SEM is shown. Data represent two independent experiments.

Figure 6. AIH-like pathology development in Tet2^ΔVAV mice is linked to liver microbiome dysbiosis and is triggered upon L. reuteri treatment.

>20-week-old (A–G and N) and 6-week-old (H–M) Tet2^ΔVAV mice and WT littermates were used.

(A) Live commensal bacteria identified in livers.

(B) Shannon diversity index of hepatic microbial DNA of species. p value compares AIH-like Tet2^ΔVAV mice with WT littermates.

(C) Hierarchical clustering heatmap of hepatic genera identified from cohort defined in Figure S5A.

(D and E) Correlation of Shannon diversity indices (of B) with (D) ANA levels and (E) frequency of hepatic IFNγ-producing T cells. Pearson correlation test of Tet2^ΔVAV mice, p value shown.

(F) Hepatic bacterial 16S rRNA gene expression.

(G) AhR activity of liver homogenate.

(H–M) Young Tet2^ΔVAV mice and WT littermates were orally treated with 10⁹ CFU L. reuteri consecutively (see STAR Methods for details).

(H) ANA levels.

(I) Representative flow cytometry plots (left) and quantification (right) of hepatic IFNγ-producing CD8 T cells.

(J) ALT activity.

(K) Presence of viable L. reuteri detected in livers of PBS (top) or L. reuteri-treated (bottom) mice.

(L and M) AhR activity of (L) recultured hepatic bacteria and (M) liver homogenate.

(N) I3A levels in liver homogenate.

(A) Data represent two independent experiments, n = 8 mice/group. (B) One-way ANOVA, Sidak’s post hoc test. Median, first and third quartiles are shown. (C) n = 4 mice/group, average is shown. (G, H–J, and N) Data represent two independent experiments. (F and H–J) Mean ± SEM, one-way ANOVA, Sidak’s post hoc test. (G and N) Kruskal-Wallis test, Dunn’s multiple comparisons test. Median, first, and third quartiles are shown. (L and M) Mean ± SEM, two-tailed unpaired t test. *p < 0.05, **p < 0.01, ***p < 0.001.

Indole-3-aldehyde-producing L. reuteri translocates to the liver and triggers AIH-like pathology in Tet2^ΔVAV mice

To investigate whether the functional activities of intrahepatic bacteria are linked to AIH-like pathology in our model, we used the phylogenetic investigation of communities by reconstruction of unobserved states (PICRUSt) tool (Langille et al., 2013). Here, we identified that the hepatic microbiome of AIH-like diseased Tet2^ΔVAV mice determines a differential enrichment of metabolic functional pathways in comparison with symptom-free Tet2^ΔVAV mice and WT littermates (Figure S5G; Table S6). AIH-like diseased Tet2^ΔVAV mice exhibited enriched pathways related to metabolism of amino acids and mono- and oligo-saccharides in addition to other cellular functions. Most intriguingly, there was a dominance of microbial genes that encode the Phenylalanine, tyrosine, and tryptophan biosynthesis pathway, a metabolic pathway that enables bacteria to catabolize dietary tryptophan (Trp) into indole derivatives, which have the ability to activate the AhR in mammalian cells (Zelante et al., 2013). Concordantly, we found several key AhR target genes to be significantly upregulated in AIH-like diseased livers of Tet2^ΔVAV mice relative to WT littermates and symptom-free Tet2^ΔVAV mice, such as Cyp1a1 (encodes cytochrome p450 family 1 subfamily a member 1), Cyp1b1, Ahrr (encodes aryl-hydrocarbon receptor repressor), Igf1r (encodes insulin like growth factor 1 receptor), and Nfe2l2 (encodes nfe2 like bzip transcription factor 2) (Figure S5H). Furthermore, we observed via an AhR reporter cell line that liver homogenate of AIH-like diseased Tet2^ΔVAV mice displayed a significant increase in the ability to induce AhR activation relative to symptom-free Tet2^ΔVAV mice and WT littermates (Figure 6G), indicating livers of Tet2^ΔVAV mice harbor increased levels of AhR ligands.

Due to the observed enrichment of Lactobacillus in livers of AIH-like diseased Tet2^ΔVAV mice and because L. reuteri is known to have high AhR ligand-producing abilities (Zelante et al., 2013), we set out to test whether L. reuteri is sufficient to induce AIH-like pathology. Indeed, we identified that oral administration of L. reuteri triggered significantly elevated ANA levels (Figure 6H), robust hepatic expansion of Tc1, but not Th1 cell frequencies (Figures 6I and S5I), and substantial hepatic injury (Figure 6J) in young Tet2^ΔVAV mice relative to WT littermates and vehicle-treated Tet2^ΔVAV mice. To better understand how L. reuteri induces AIH-like pathology, we set out to test whether orally gavaged L. reuteri translocates to and releases AhR ligands within the liver. We found that hepatic L. reuteri translocation was independent of Tet2 and occurs in L. reuteri-treated, but not PBS-treated, mice (Figure 6K), and confirmed that translocated L. reuteri maintained the ability to catabolize Trp and subsequently produce AhR ligands ex vivo (Figure 6L). Accordingly, L. reuteri-treated mice displayed significantly increased AhR activity in liver homogenate when compared with PBS-treated mice (Figure 6M), implying that intrahepatic L. reuteri releases AhR ligands within the liver. These findings prompted us to assess the hepatic abundance of indole-3-aldehyde (I3A), the primary AhR agonist of several Lactobacillus spp (Zelante et al., 2013), including L. reuteri, within the livers of untreated AIH-like diseased Tet2^ΔVAV mice. Strikingly, we found a significant increase in hepatic I3A levels of AIH-like diseased Tet2^ΔVAV mice relative to symptom-free Tet2^ΔVAV mice and WT littermate controls via quantitative high-resolution mass spectrometry (Figure 6N). Furthermore, L. johnsonii, a phylogenetic relative of L. reuteri that lacks the ability to produce I3A (Cervantes-Barragan et al., 2017; Zelante et al., 2013), was insufficient to induce AIH-like pathology in young Tet2^ΔVAV mice (Figures S6A-S6C), in stark contrast to L. reuteri (Figures 6H-6J).

Taken together, these findings imply that intrahepatic I3A-releasing pathobionts play a critical role in the induction of AIH in our model.

The development of AIH-like disease in Tet2^ΔVAV mice is dependent on AhR signaling

Next, we assessed whether L. reuteri-induced AIH-like disease in Tet2^ΔVAV mice is dependent on AhR signaling. As expected, L. reuteri treatment triggered AIH-like pathology in young Tet2^ΔVAV mice, characterized by elevated ANA and ALT levels relative to vehicle-treated Tet2^ΔVAV mice and L. reuteri-treated WT littermates, yet remarkably, AhR blockade efficiently suppressed the development of L. reuteri-mediated AIH-like pathology in Tet2^ΔVAV mice (Figures 7A and 7B). L. reuteri translocated to a similar proportion of livers independent of AhR inhibitor treatment (Figure 7C). These findings prompted us to assess whether blocking AhR signaling reverts ongoing AIH-like pathology in Tet2^ΔVAV mice. Strikingly, we observed that blocking AhR signaling efficiently suppressed ongoing AIH-like pathology in aged Tet2^ΔVAV mice, demonstrated by a significant reduction in plasma ANA and ALT levels (Figures 7D and 7E) and hepatic Tc1 cell frequencies (Figure 7F). Of note, AhR inhibitor-treated, compared with vehicle-treated, WT littermates remained grossly unaffected (Figures 7D-7F). Taken together, these data demonstrate that AhR signaling is critical in the etiopathogenesis of Tc1 cell-dependent AIH-like pathology in the context of hematopoietic Tet2 deficiency.

Figure 7. I3A is required for, and the absence of Tet2 in CD8 T cells is sufficient to succumb to, L. reuteri-mediated AIH-like pathology.

(A–C) 6-week-old Tet2^ΔVAV mice and WT littermates treated with L. reuteri or PBS ± AhR antagonist (CH223191; intraperitoneal [i.p.] injection) or vehicle control.

(A) ANA levels.

(B) ALT activity.

(C) Presence of viable L. reuteri in the livers of L. reuteri-treated mice.

(D–F) >20-week-old Tet2^ΔVAV AIH-like diseased mice and WT littermates treated with AhR antagonist (CH223191; i.p.) or vehicle control.

(D) ANA levels; D, day.

(E) ALT activity.

(F) Representative flow cytometry plots (left) and quantification (right) of hepatic IFNγ-producing CD8 T cells.

(G) IFNγ production by naive CD8 T cells following stimulation with anti-CD3 and anti-CD28 ± L. reuteri supernatant ± ΔArAT L. reuteri supernatant ± AhR antagonist.

(H) RNA-seq heatmap of naive CD8 T cells following stimulation with anti-CD3 and anti-CD28 ± I3A ± AhR antagonist; WT, Tet2 cre⁻ mice; KO, Tet2^ΔVAV mice.

(I–K) 6-week-old Tet2^ΔCD8 mice and WT littermates treated with L. reuteri or ΔArAT L. reuteri.

(I) ANA levels.

(J) ALT activity.

(K) Presence of viable L. reuteri in livers of L. reuteri or ΔArAT L. reuteri-treated mice.

(L and M) 6-week-old Tet2^ΔVAV mice and WT littermates were used.

(L) Representative histogram plot of mean fluorescent intensity (MFI) (left) and quantification (right) of IFNγR1 expression on hepatic naive CD8 T cells.

(M) Representative histogram plot of mean fluorescent intensity (MFI) (left) and quantification (right) of pSTAT1 activation in naive CD8 T cells following stimulation with anti-CD3 and anti-CD28 ± recombinant IFNγ; dashed lines, PBS treated; solid lines, recombinant IFNγ treated.

(A–C) Data represent 3–7 mice/group.

(A and B) Mean ± SEM. (A) d0 to d35 comparison, paired t test, treatment comparison at D35, two-way ANOVA, Sidak’s multiple comparisons test. (B) One-way ANOVA, Sidak’s post hoc test.

(D–F) Data represent two independent experiments, mean ± SEM n = 4–6 mice/group.

(D) D0 to D35 comparison, paired t test, treatment comparison at D35, two-way ANOVA, Sidak’s multiple comparisons test.

(E and F) One-way ANOVA, Sidak’s post hoc test.

(G) Data represent 5 mice/group, mean ± SEM, two-way ANOVA, Sidak’s multiple comparisons test.

(H) Data represent n = 3–4 mice/group, average is shown.

(I and J) Mean ± SEM, two-way ANOVA, Sidak’s multiple comparisons test.

(L) Mean ± SEM, two-tailed unpaired t test.

(M) Data represent two independent experiments, mean ± S.E.M., paired t test, two-way ANOVA, Sidak’s multiple comparisons test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

I3A is required for L. reuteri-induced Tc1 cell immunity in vitro and AIH-pathology in vivo, both of which are restrained by Tet2 within CD8 T cells

Finally, we interrogated whether L. reuteri-released I3A directly acts on CD8 T cells to promote effector function and the mechanism of how the absence of Tet2 in CD8 T cells elevates effector responses and subsequently accelerates the development of AIH in our model. To this end, we deployed a reductionist approach and cultured splenic naive CD8 T cells with supernatant derived from either I3A-releasing L. reuteri or a genetically engineered L. reuteri strain lacking the aromatic amino acid aminotransferase class I/II (ArAT) and consequently the ability to release the AhR agonist I3A (ΔArAT L. reuteri) (Cervantes-Barragan et al., 2017). We observed that L. reuteri-derived supernatant substantially promoted IFNγ production in WT CD8 T cells, which was significantly amplified in the absence of Tet2 (Figure 7G). This increase in Tc1 cell effector function was dependent on both AhR signaling and the presence of I3A, since both the addition of the AhR antagonist CH223191 and the supernatant of ΔArAT L. reuteri either efficiently suppressed or failed to induce IFNγ production in CD8 T cells, respectively (Figure 7G). In line with the ability of L. reuteri to selectively induce Tc1, but not Th1, cell effector function in vivo (Figures 6I and S5I), L. reuteri-derived supernatant failed to induce IFNγ production independent of Tet2 in CD4 T cells in vitro (Figure S6D), indicating that L. reuteri-derived I3A induces a type 1 immune profile selectively in CD8 T cells in vitro and in vivo. Assessment of the impact of I3A on CD8 T cells and its antagonism by Tet2 at a genome-wide level uncovered that I3A induced key type 1 immune transcription factors such as Tbet, Eomes, and Irf4, as well as effector cytokines such as Ifng, Gzmb, and perforin (Figure 7H; Table S7), and this effect was AhR dependent and significantly enhanced in the absence of Tet2 (Figure 7H).

Next, we interrogated whether I3A is required for, and absence of Tet2 within CD8 T cells is sufficient to succumb to, L. reuteri-induced AIH-like pathology. Remarkably, in contrast to PBS-treated young Tet2^ΔCD8 mice and littermate controls that remained unaffected, L. reuteri-treated Tet2^ΔCD8 mice, but not WT littermates, developed AIH-like pathology, displayed by significantly elevated ANA and ALT plasma levels, whereas ΔArAT L. reuteri-treated Tet2^ΔCD8 mice failed to develop AIH-like disease (Figures 7I and 7J). Notably, both viable L. reuteri strains were present in a similar proportion of livers independent of genotype (Figure 7K). These findings demonstrate the absence of Tet2 in CD8 T cells is sufficient to render mice susceptible to L. reuteri-mediated AIH-like pathology, dependent on the production of I3A.

To better understand the mechanism of how the L. reuteri-derived I3A-induced Tc1 cell immune response is exacerbated in CD8 T cells deficient of Tet2, we assessed whether Tet2^−/− CD8 T cells display a heightened sensitivity to IL-12, given that IL-12 is a key cytokine for inducing a type 1 effector program (Xiao et al., 2009). However, we found that Tet2^+/+ and Tet2^−/− CD8 T cells, which express comparable surface IL-12 receptor β2 levels, produced a comparable amount of IFNγ when stimulated with IL-12 ex vivo (Figures S6E and S6F). IFNγ-mediated engagement of IFNγR1 is described to promote a type 1 effector program (Whitmire et al., 2005). Interestingly, we found IFNγR1 to be significantly upregulated on hepatic and splenic naive CD8 T cells of young Tet2^ΔVAV mice relative to WT littermates (Figures 7L and S6G). As expected, IFNγ stimulation of splenic Tet2^+/+ and Tet2^−/− naive CD8 T cells ex vivo led to an increase in STAT1 activation in both Tet2^+/+ and Tet2^−/− CD8 T cells; however, strikingly, IFNγ-induced STAT1 activation was significantly higher in Tet2^−/− compared with Tet2^+/+ CD8 T cells (Figure 7M). This finding suggests that given their elevated expression of IFNγR1, naive Tet2^−/− CD8 T cells display enhanced sensitivity to extrinsic IFNγ leading to increased activation of STAT1 that maintains a type 1 immune program.

Taken together, these data suggest that mechanistically, Tet2 within CD8 T cells antagonizes L. reuteri-derived I3A-induction of Tc1 cell differentiation and function that ultimately drives the development of AIH-like pathology, possibly via restricting increased sensitivity to exogenous IFNγ.

DISCUSSION

Our defined AIH model suggests that the hepatic expansion of AhR ligand-releasing pathobionts triggers AIH-like pathology in an IFNγ and AhR-dependent manner. We postulate that given the enhanced sensitivity of naive Tet2^−/− CD8 T cells to extrinsic IFNγ, I3A-induced IFNγ production promotes a vicious feedforward cycle that leads to AIH-like pathology (Figure S6H). Supporting this notion, genetic Ifng ablation prevented the development of (Figures S3C-S3H), and neutralization of IFNγ reverted ongoing, AIH-like pathology in our model (Figures 3A-3G).

AIH-like disease development in Tet2^ΔVAV mice is linked to both elevated hepatic Tc1, as well as Th1 cell responses; thus, we postulate that other bacterial taxa, besides L. reuteri, promote hepatic Th1 cell effector function in our model. In line with this hypothesis, livers of AIH-like diseased Tet2^ΔVAV mice displayed an expansion of other genera besides Lactobacillus spp., including Sutterella spp., Lachnospiraceae spp., and Prevotella spp.

Given that GF mice are absent of microbial-derived AhR ligands but are exposed to endogenous AhR ligands, our finding that GF Tet2^−/− mice fail to develop AIH-like disease confirms that endogenous AhR ligands do not play a major role in our AIH-like disease model. Furthermore, although a recent report revealed alterations in AhR signaling in purified Th17 and Treg cells obtained from AIH patients (Vuerich et al., 2021), these alterations were ligand-specific and did not address the impact of microbial-derived AhR ligands. Thus, the impact of microbial-derived AhR ligands such as I3A on Th17 and Treg cells from AIH patients warrants further investigation.

Although our study shows that L. reuteri is sufficient to induce AIH-like pathology in Tet2^ΔVAV and Tet2^ΔCD8 mice, it is likely not required for the development of AIH in our model; rather, we hypothesize that an expansion of hepatic AhR ligand-producing pathobionts is required. This is supported by our observation that other pathobionts capable of producing AhR ligands translocate to livers of AIH-like Tet2^ΔVAV mice, such as L. murinus, E. faecalis, and E. coli (Lamas et al., 2018).

Despite the defined role of Tet2 as critical modulator of CD4 T cell function (Ichiyama et al., 2015; Yue et al., 2019), the impact of Tet2 on CD8 T cell function remains poorly understood. In a mouse model of TCRβ Tet2 T cell deficiency, it was shown that the loss of Tet2 was associated with enhanced Tc1 cell effector function in acute lymphocytic choriomeningitis virus infection (Carty et al., 2018). Another study highlighted that TET2 acts as an epigenetic modulator of human CD8 T cell fate and showed that spontaneous somatic TET2 mutation (which led to bi-allelic deletion of TET2) within CD19-targeted CD8 T cells promoted their antitumor activity mediated by increased IFNγ and Gzmb release (Fraietta et al., 2018). Thus, these findings, in accordance with our own, suggest Tet2 antagonizes cytotoxic CD8 T cell effector function.

Our study shows that the liver microbiome, in particular translocated I3A-producing pathobionts, plays a critical role in mediating Tc1 cell effector function that drives AIH-like disease. One exciting area of future investigations will be to define the reactivity and antigen dependence of hepatic CD8 T cells against expanded AhR ligand-producing pathobionts, including L. reuteri, and/or hepatocytes in the context of Tet2 deficiency. Previous studies have shown increased abundance and effector function of autoantigen-specific CD4 T cells in AIH-like disease models and human AIH (Ma et al., 2006; Mix et al., 2008; Preti et al., 2021; Umeshappa et al., 2021) and that adoptive transfer of autoantigen-specific CD8 T cells is sufficient to promote AIH-like disease (Zierden et al., 2010), implicating both autoantigen-specific CD4 and CD8 T cells in AIH.

Collectively, we identified a AIH-like mouse model that will aid in a better understanding of the underlying mechanisms that drive microbiota- and Tc1 cell-dependent AIH and will help in developing innovative treatment regimens to alleviate AIH in humans.

Limitations of the study

Limitations to our study include identifying whether L. reuteri secretes I3A within the liver in situ and if an expansion of Trp catabolizing gut-pathobionts specifically within the liver is required to trigger AIH in our model. Furthermore, the cell type that directly responds to AhR ligands in our AIH model in vivo remains unaddressed. The mechanism that drives liver dysbiosis in Tet2^ΔVAV mice remains unanswered and the genomic mechanism by which I3A induces Tc1 cell differentiation remains to be addressed. Future studies will need to test whether AIH-like pathology spontaneously develops in Tet2^ΔCD8 mice and if so, the role of translocated AhR-ligand-producing pathobionts will need to be defined.

STAR★METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Marlies Meisel (marlies@pitt.edu).

Materials availability

Mouse lines generated in this study are available from the lead contact upon request.

Data and code availability

Raw and analyzed 16S rRNA-Seq data are available via the NCBI Sequence Read Archive (SRA: PRJNA830441). Raw and analyzed hepatic RNA-seq data are available at NCBI: GSE201459. Remaining data needed to support the conclusion of this manuscript are included in the main text and supplementary materials. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Mice

All knockout mice used in these studies are on a C57BL/6 background. Age-matched >20-week (wk)-old mice (age range 20-25 weeks) were used for all experiments, unless indicated otherwise. Both female and male mice were used for experiments at equal levels; no notable sex-dependent differences were found for the reported experiments. Tet2^−/− mice were previously described (Moran-Crusio et al., 2011). Tet2^fl (B6;129S-Tet2^tm1.1Iaai/J) mice were crossed with the hematopoietic-specific Vav-cre line (B6.Cg-Tg(Vav1-icre)A2Kio/J), the myeloid cell lineage-specific LysM-cre line (Lyz2tm1(cre)Ifo)), the T cell specific CD2-cre line (C57BL/6-Tg(CD2-cre)1Lov/J), the hepatocyte specific Albumin-cre line (B6.Cg-Speer6-ps1^{Tg(Alb-cre)21Mgn}/J), and the CD8 T cell specific CD8α-cre line (C57BL/6-Tg(Cd8a-cre)1Itan/J) all from the Jackson Laboratory. Full Tet2^−/− mice were crossed with Ifng^−/− mice (B6.129S7-Ifng^tm1Ts/J) to generate Tet2^−/−Ifng^−/− and Tet2^+/+Ifng^−/− littermate controls. Tet2^−/− mice were crossed with Rag^−/− mice (B6(Cg)-Rag2^tm1.1Cgn/J) to generate Tet2^−/−Rag^−/− and Tet2^+/+Rag^−/−littermate controls. Rag^−/− mice were kindly provided by Dr. M. Shlomchik, University of Pittsburgh. Mice were housed at the University of Pittsburgh animal facilities under specific pathogen-free (SPF) conditions, where cages were changed on a weekly basis; ventilated cages, bedding, food, and water (non-acidified) were autoclaved before use, ambient temperature maintained at 23 °C, and 5% Clidox-S was used as a disinfectant. Experimental breeding cages were randomly housed on two different racks in vivarium, and all cages were kept on automatic 12-h light/dark cycles. Germ-free (GF) Tet2^−/− mice (Meisel et al., 2018) were bred to GF C57BL/6 WT mice to generate littermate controls and maintained in flexible film isolators in University of Pittsburgh Axenic Gnotobiotic Units. Animal husbandry for both SPF and GF facilities, and experimental procedures, were performed in accordance with Public Health Service policy and approved by the University of Pittsburgh Institutional Animal Care and Use Committees and the University of Chicago Animal Care Committee.

Gnotobiotic animal husbandry

Food, bedding, and water (non-acidified) were autoclaved before transfer into sterile isolators. Cages within isolators were changed weekly, and all cages in the vivarium were kept on 12-h light/dark cycles. Microbiology testing of fecal (experimental mice), or of cecum samples (sentinel mice; aerobic and anaerobic culture, 16S qPCR) was performed every other week to confirm GF status.

Bacterial cultures

The following bacterial strains were used for in vitro and/or in vivo studies: Lactobacillus reuteri (L. reuteri; ATCC), L. johnsonii (ATCC) and ΔArAT L. reuteri (lacking the gene coding for aminotransferase class I/II (Cervantes-Barragan et al., 2017) generously provided by Dr. L. Cervantes-Barragan, Emory University). L. reuteri strains were grown in MRS broth (Gibco) at 37 °C under anaerobic conditions.

Culture of livers at steady state

Similar to Vieira et al. (2018), livers were aseptically removed and homogenized by mashing through a 100μm cell strainer in sterile water containing 0.05% NP-40. Post 3-4 h of incubation at 37 °C, an aliquot of the tissue homogenate was added to BBL Mycoflask Thioglycollate (Fluid) Prepared Media (BD Diagnostic Systems) and incubated for 72 h under aerobic or anaerobic conditions at 37 °C. Serial dilutions of samples were prepared and 12 μl of cell suspension were dropped, in duplicates, onto BD BBL^™ Brucella Agar with 5% Sheep Blood, Hemin and Vitamin K1 agar plates (Fisher Scientific) under anaerobic (anaerobic chamber (Coy Systems)) or aerobic conditions. Plates were sealed with parafilm and incubated upside down for 48 h (aerobic) or 72 h (anaerobic).

METHOD DETAILS

Tissue harvest and cell purification

Peripheral blood (PB) was collected by cheek bleeding (post sterilizing cheeks with 70% ethanol wipes) into EDTA-containing tubes. Livers were harvested with autoclaved tools under sterile conditions and liver weight was recorded. Liver mononuclear cells (LMNCs) were isolated by using previously described protocol with the following modifications (Castillo-Dela Cruz et al., 2019). Briefly, livers were mashed through a 100μm cell strainer. LMNCs were purified from liver single cell suspensions by a 40% percoll (Fisher Scientific) centrifugation. Erylysis of PB and livers were performed using the Mouse Erythrocyte Lysing Kit (R&D Systems).

Flow cytometry

Single cell suspensions were prepared as described and either not stimulated (for assessment of myeloid cells in PB) or stimulated in RPMI containing PMA (0.1 μg/ml), ionomycin (1μg/ml), Golgi Stop (1.95μM), and Golgi Plug (1μg/ml). Post 3 h of stimulation, cells were washed and resuspended in FACS buffer (PBS, 2% FBS) for immunostaining and subsequent FACS analysis. All cell suspensions were incubated with Fc Block (BD Biosciences) followed with surface marker Ab stain for 20 min (min) at 4 °C. For dead cell exclusion, cells were stained with Zombie NIR^™ Fixable (BioLegend) or LIVE/DEAD Fixable Aqua dye (Invitrogen) and washed with FACS buffer. For intracellular cytokine and transcription factor staining, surface Ab-stained cells were first fixed and permeabilized using the FoxP3 Transcription Factor Staining Buffer kit (eBioscience^™) following manufacturer’s instructions. Cells were further stained with Abs for intracellular proteins for 30 min at 4 °C. For IL-12p40 staining, LMNCs were incubated in RPMI medium (Thermo Fisher Scientific) for 6 h in the presence of 1 μl/ml Golgi Plug (BD Biosciences). Surface staining was performed as described above and cells were permeabilized with Cytofix/Cytoperm (BD Biosciences) for intracellular cytokine staining. Samples were FSC-A/SSC-A gated to exclude debris and gated to exclude dead cells. Samples were run on an Aurora (Cytek) flow cytometer and analyzed with FlowJo 10 (Tree Star). All fluorophore conjugated Abs used are listed under key resources table. Representative gating strategies of all cell populations displayed in this study can be found in Figure S7.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
anti-CD16 and anti-CD32 (clone 2.4G2)	BD Biosciences	Cat# 553142; RRID:AB_394657
anti-Mouse CD4 monoclonal antibody (Rat, Clone GK1.5), Brilliant Violet 650 conjugated	BD Biosciences	Cat# 563232; RRID:AB_2738083
anti-Mouse CD8 monoclonal antibody (Rat, Clone 53-6.7), Brilliant Violet 570 conjugated	Biolegend	Cat# 100740; RRID:AB_2563055
anti-Mouse IFN-gamma monoclonal antibody (Rat, Clone XMG1.2), Brilliant Violet 605 conjugated	BioLegend	Cat# 505839; RRID:AB_2561438
anti-Mouse CD45 monoclonal antibody (Rat, Clone 30-F11), Brilliant Violet 480 conjugated	BD Biosciences	Cat# 566095; RRID:AB_2739499
anti-Mouse TCR β monoclonal antibody (armenian hamster, Clone H57-597), Alexa Fluor® 700 conjugated	BD Biosciences	Cat# 560705; RRID:AB_1727573
anti-Mouse FOXP3 monoclonal antibody (Rat, Clone FJK-16s), FITC conjugated	Thermo Fisher Scientific	Cat# 11-5773-82; RRID:AB_465243
anti-Mouse CD45 monoclonal antibody (Rat, 30-F11), APC-Cy™7 conjugated	BD Biosciences	Cat# 557659; RRID:AB_396774
anti-Mouse CD4 monoclonal antibody (Rat, GK1.5), BUV395 conjugated	BD Biosciences	Cat# 563790; RRID:AB_2738426
anti-Mouse CD8 monoclonal antibody (Rat, 53-6.7), BUV737 conjugated	BD Biosciences	Cat# 612759; RRID:AB_2870090
anti-Mouse CD19 monoclonal antibody (Rat, Clone 1D3/CD19), PerCP/Cyanine5.5 conjugated	BioLegend	Cat# 152406; RRID:AB_2629815
anti-Mouse F4/80 monoclonal antibody (Rat, Clone BM8), PE-Cyanine5 conjugated	eBioscience™	Cat# 15-4801-82; RRID:AB_468798
anti-Mouse CD11c monoclonal antibody (Armenian hamster, Clone N418), Brilliant Violet 421 conjugated	BD Biosciences	Cat# 565451; RRID:AB_2744278
anti-Mouse IL-17A monoclonal antibody (Rat, Clone TC11-18H10.1), Alexa Fluor(R) 647 conjugated	BioLegend	Cat# 506912; RRID:AB_536014
anti-Mouse NK-1.1 monoclonal antibody (Mouse, Clone PK136), Brilliant Violet 786 conjugated	BD Biosciences	Cat# 740853; RRID:AB_2740506
anti-Mouse CD44 monoclonal antibody (Rat, IM7), APC conjugated	BD Biosciences	Cat# 559250; RRID:AB_398661
anti-Mouse CD62L monoclonal antibody (Rat, Clone MEL-14), PE conjugated	Thermo Fisher Scientific	Cat# 12-0621-81; RRID:AB_465720
anti-Mouse NK1.1 monoclonal antibody (Rat, Clone PK136), FITC conjugated	BioLegend	Cat# 108706; RRID:AB_313393
anti-Mouse MHC II monoclonal antibody (Rat, Clone M5/114.15.2), FITC conjugated	Thermo Fisher Scientific	Cat# 11-5321-82; RRID:AB_465232
anti-Mouse TCR γ/δ monoclonal antibody (Rat, Clone UC7-13D5), FITC conjugated	BioLegend	Cat# 107504; RRID:AB_313313
anti-Mouse CD11b monoclonal antibody (Rat, Clone M1/70), APC-eFluor 780 conjugated	Thermo Fisher Scientific	Cat# 47-0112-82; RRID:AB_1603193
anti-Mouse Ly-6G/Ly-6C (Gr-1) monoclonal antibody (Rat, Clone RB6-8C5), Alexa Fluor(R) 700 conjugated	BioLegend	Cat# 108422; RRID:AB_2137487
anti-Mouse Ly-6G/Ly-6C monoclonal Antibody (Rat, Clone RB6-8C5), PerCP-Cyanine5.5 conjugated	Thermo Fisher Scientific	Cat# 45-5931-80; RRID:AB_906247
anti-Mouse CD11b Monoclonal Antibody (Rat, Clone M1/70), PE-Cyanine7 conjugated	Thermo Fisher Scientific	Cat# 25-0112-82; RRID:AB_469588
anti-Mouse IL-12/IL-23 p40 Monoclonal Antibody (Rat, Clone C17.8), PE conjugated	eBioscience™	Cat# 12-7123-82; RRID:AB_466185
anti-Mouse IFNγ monoclonal Antibody (Rat, Clone XMG1.2) unconjugated	Bio X Cell	Cat# BE0055; RRID:AB_1107694
IgG1 isotype control (anti-HRP) monoclonal antibody (Rat, Clone, HRPN) unconjugated	Bio X Cell	Cat# BE0088; RRID:AB_1107775
anti-Mouse CD3e monoclonal antibody (Armenian hamster, 145-2C11) unconjugated	BD Biosciences	Cat# 553057; RRID:AB_394590
anti-Mouse CD28 monoclonal antibody (Syrian hamster, 37.51) unconjugated	BD Biosciences	Cat# 553294; RRID:AB_394763
anti-Mouse CD138 Monoclonal Antibody (Rat, Clone 281-2), PE conjugated	BioLegend	Cat# 142504; RRID:AB_10916119
anti-Mouse IFN gamma Receptor 1 Monoclonal Antibody (Armenian hamster, Clone 2E2), PE conjugated	Thermo Fisher Scientific	Cat# 12-1191-82; RRID:AB_1210730
anti-Human/Mouse IL-12 R beta 2 Monoclonal Antibody (Mouse IgG1, Clone 305719), APC conjugated	R & D Systems	Cat# FAB1959A; RRID: AB_2124049
anti-Mouse STAT1 Monoclonal Antibody (Mouse, Clone A15158B), PE conjugated	BioLegend	Cat# 686404; RRID:AB_2616939
Goat Anti-Mouse IgG, Human ads-UNLB antibody unconjugated	SouthernBiotech	Cat# 1030-01, RRID:AB_2794290
goat anti-Mouse IgG, Human ads-HRP antibody unconjugated	SouthernBiotech	Cat# 1030-05, RRID:AB_2619742
goat anti-Mouse IgG, Human ads-FITC antibody unconjugated	SouthernBiotech	Cat# 1030-02, RRID:AB_2794291
anti-mouse IFN-gamma Monoclonal Antibody (Rat, Clone AN 18), unconjugated	BD Biosciences	Cat# 551309, RRID:AB_394145
anti-mouse IFN-gamma Monoclonal Antibody (Rat, Clone XMG1.2), biotin conjugated	BD Biosciences	Cat# 554410, RRID:AB_395374
Peroxidase-Streptavidin antibody	Jackson ImmunoResearch	Cat# 016-030-084, RRID:AB_2337238
Chemicals, Peptides, and Recombinant Proteins
Recombinant Mouse IFN-gamma	R & D Systems	Cat# 485-MI-100
Zombie NIR™ Fixable Viability Kit	BioLegend	Cat# 423106
LIVE/DEAD™ Fixable Aqua Dead Cell Stain Kit	Invitrogen	Cat# L34957
Super Bright Complete Staining Buffer	eBioscience	Cat# SB-4401-75
Proteinase K Solution	Invitrogen	Cat# 25530049
ExoSAP-IT™ PCR Product Cleanup Reagent	Applied Biosystems	Cat# 78201.1.ML
iTaq™ Universal SYBR® Green Supermix	Bio-Rad	Cat# 1725125
GolgiStop™	BD Biosciences	Cat# 554724
GolgiPlug™	BD Biosciences	Cat# 555029
PMA	Sigma-Aldrich	Cat# P1585
Ionomycin calcium salt	Sigma-Aldrich	Cat# I0634
Mouse Erythrocyte Lysing Kit	R & D Systems	Cat# WL2000
Percoll™	Thermo Fisher Scientific	Cat# 45-001-747
RPMI 1640	Thermo Fisher Scientific	Cat# MT10040CV
DMEM	Gibco™	Cat# 11-995-065
Penicillin-Streptomycin-Glutamine	Gibco™	Cat# 10-378-016
L-Tryptophan	Sigma-Aldrich	Cat# T0254
Peptone	Sigma	Cat# 70173
Tryptone	Sigma	Cat# T7293
Sodium chloride	Sigma	Cat# S3014
Indole-3-aldehyde (carboxyaldehyde)	Sigma-Aldrich	Cat# 129445
AHR inhibitor (CH-223191)	Sigma-Aldrich	Cat# C8124
DNase I	Roche	Cat# 10104159001
BBL™ Fluid Thioglycollate	BD Biosciences	Cat# BB21196
BBL™ Brucella Agar with 5% Sheep Blood, Hemin and Vitamin K1 Prepared Media	Midland Scientific	Cat# 297848
Lactobacillus spp. MRS Broth	BD Biosciences	Cat# DF0881-17-5
Vancomycin	Sigma-Aldrich	Cat# V2002
Kanamycin	Sigma-Aldrich	Cat# 60615
Metronidazole	Sigma-Aldrich	Cat# M3761
Colistin	Sigma-Aldrich	Cat# C4461
Gentamycin	Sigma-Aldrich	Cat# G1914
ALT /GPT Reagent; Infinity	Fisher Scientific	Cat# TR71121
Normocin	InvivoGen	ant-nr-1
Zeocin	InvivoGen	ant-zn-05
QUANTI-Luc™	InvivoGen	rep-qlc1
Recombinant Mouse IL-12 (p70)	BioLegend	Cat#577002
Concanavalin A	Sigma-Aldrich	Cat# C5275
Carbon tetrachloride	Sigma-Aldrich	Cat# 319961
Critical Commercial Assays
FoxP3 Transcription Factor Staining Kit	eBioscience	Cat# 00-5523-00
Cytofix/Cytoperm™ Fixation/Permeablization Kit	BD Biosciences	Cat# 554714
In Situ Cell Death Detection Kit, TMR red	Sigma-Aldrich	Cat# 12156792910
EasySep™ Mouse Naïve CD8+ T Cell Isolation Kit	STEMCELL Technologies	Cat# 19858
EasySep™ Mouse Naïve CD4+ T Cell Isolation Kit	STEMCELL Technologies	Cat# 19765
Mouse Anti-nuclear antibody (ANA) ELISA Kit	MyBioSource	Cat# MBS261480
TMB Substrate Kit	Thermo Fisher Scientific	Cat# 34021
iScript™ cDNA Synthesis Kit	Bio-Rad	Cat# #1708891BUN
SuperScript™ VILO™ cDNA Synthesis Kit	Invitrogen	Cat# 11754050
RNeasy Plus Mini Kit	Qiagen	Cat# 74136
DNA Microbiome Kit	Qiagen	Cat# 51704
RNeasy Plus Micro Kit	Qiagen	Cat# 74034
QIAamp Fast DNA Stool Mini Kit	Qiagen	Cat# 51604
Phosflow Lyse/Fix Buffer	BD Biosciences	Cat# 558049
True-Phos™ Perm Buffer	BioLegend	Cat#425401
Mouse Liver Kidney Microsome Autoantibody (Anti-LKM) ELISA Kit	MyBioSource	Cat# MBS3807100
Mouse soluble liverantigen antibody (SLA) ELISA Kit	MyBioSource	Cat# MBS266348
Kallestad HEp-2 Slides	Bio-Rad	Cat# 26103
Invitrogen™ Fluoromount-G™ Mounting Medium	Fisher Scientific	Cat# 00-4958-02
Deposited Data
Raw and analyzed 16S rRNA-Seq data	This paper	NCBI Sequence Read Archive, SRA: PRJNA830441
Raw and analyzed hepatic RNA-seq data	This paper	NCBI: GSE201459
Experimental Models: Organisms/Strains
Mouse: B6.Tet2−/−	Meisel et al. (2018)	N/A
Mouse: B6. Tet2f (B6;129S-Tet2tm1.1Iaai/J)	Meisel et al. (2018)	N/A
Mouse: B6. Vav-cre line (B6.Cg-Tg(Vav1-icre)A2Kio/J)	Jackson Laboratory	Stock No: 008610
Mouse: B6. LysM-cre line (Lyz2tm1(cre)Ifo))	Jackson Laboratory	Stock No: 004781
Mouse: B6. Tg(CD2-cre)1Lov/J	Jackson Laboratory	Stock No: 027406
Mouse: B6. Albumin-cre (B6.Cg-Speer6-ps1Tg(Alb-cre)21Mgn/J)	Jackson Laboratory	Stock no.: 003574
Mouse: B6. Ifng−/− (B6.129S7-Ifngtm1Ts/J)	Jackson Laboratory	Stock no.: 002287
CD8α-Cre (C57BL/6-Tg(Cd8a-cre)1Itan/J)	Jackson Laboratory	Stock no.: 008766
Mouse: B6. Rag−/−	Provided by M. Shlomchik	N/A
Germ free B6. Tet2−/− and B6.Tet2+/+	Meisel et al. (2018)	N/A
L. reuteri	ATCC	Cat# BAA-2837
L. johnsonii	ATCC	Cat# BAA-3147
ΔArAT L. reuteri	Dr. Cervantes-Barragan	N/A
AhR reporter cell line	Invivogen	Cat# ht2l-ahr
Blautia producta	ATCC	27340D-5
Oligonucleotides
(mouse) see Table S8
Software and Algorithms
GraphPad Prism 9	GraphPad Software	N/A
FlowJo 10.7.1	Tree Star	https://www.flowjo.com/solutions/flowjo/downloads
ImageJ	Open source	https://imagej.nih.gov/ij/index.html
BioRender	BioRender Company	N/A

RNA processing and RT-PCR

A piece of the left liver lobe (~3x3mm) was incubated in RNAlater^™ (Qiagen) at 4 °C for 48 h and stored at −80 °C until further analysis. For RNA extraction a Tissue-Tearor Homogenizer (Biospec) was used. RNA was extracted using the RNeasy Mini Kit (Qiagen). cDNA synthesis was performed using iScript^™ cDNA Synthesis Kit (Bio-Rad) according to manufacturer’s instructions. Expression analysis was performed in duplicate via real-time PCR on a BioRad CFX384 Touch^™ Real-Time PCR Detection System using iTaq^™ Universal SYBR (Bio-Rad). Expression levels were quantified and normalized to Gapdh expression.

RNA-sequencing processing and data analysis

RNA of liver tissue was extracted (see above for details). RNA-seq libraries were sequenced with single-end 50bp reads on an Illumina HiSeq4000. Single end Illumina reads were checked for quality using FastQC (v0.11.5) and they were trimmed for quality and adapters using Cutadapt (v1.12). These quality trimmed reads were later mapped against the Ensembl Mus Musculus reference transcriptome (GRCm38) using Salmon mapper (v0.9.1). edgeR (v3.24.1), a bioconductor R (v3.8) package, was used to analyze differentially expressed genes (DEG).

We considered a gene as differentially expressed (DEG) if statistically supported at a Q-value false discovery rate (FDR) < 0.1 and a ∣ log₂(fold change)∣ > 0.5. The top 2000 DEGs with FDR < 0.05 and an absolute value of log₂(Fold change) > 1 were uploaded to Cytoscape (v3.7.0) and to the Ingenuity pathway analysis platform (IPA). Analysis was performed by using IPA tools 7.0. Using the ClueGO (v2.5.3) (Backes et al., 2007) app extension, a visual representation of GO terms was created. The size of the nodes shows the enrichment significance after Bonferroni correction statistical analysis. Analysis was done using the following parameters: FDR < 0.01; terms > 10 genes; kappa 0.4, (see Table S4 for list of GO terms). Heat maps were made from select GO terms using the R package gplots (v3.0.1) or IPA (see Table S3 for list of GO terms). Venn Diagram was made from DEG list using the R package VennDiagram (1.6.20) (see Table S1 for list of genes displayed in Venn diagram and Table S2 for entire list of DEGs of RNA-seq analysis). For CD8 T cell RNA-sequencing libraries were prepared using the Illumina TruSeq protocol and sequenced with paired-end 75 bp reads on an Illumina HiSeq. FASTQ sequences with Phred scores less than 20 were trimmed from 3 prime end and reads smaller than 25 bp were removed. The remaining clean reads were aligned to the mm10 protein-coding reference library from Ensembl using STAR. Noise reduction filter was applied to quantified transcript counts to exclude features with maximum counts less than 10 and FPKMs were generated (see Table S7).

Measurement of intra-cellular staining pattern of plasma anti-nuclear antibodies

HEp-2 immunofluorescence assay (measures reactivity to plasma ANA) and HEp-2 -scoring were performed as previously described (Tilstra et al., 2020). Briefly, 1:50 PBS-diluted mouse plasma samples were incubated on HEp-2 cell slides according to manufacturer’s instructions (Bio-rad). After washing in PBS, cells were incubated with fluorescein isothiocyanate–conjugated goat anti-mouse IgG and mounted with Fluoromount-G. Images were captured on a fluorescent microscopy (Olympus, IX83) and processed in Adobe Photoshop. Slides (blinded) were imaged at 20x and scored 0 – 3 for intensity of Kallestad HEp-2 staining as either nuclear homogenous, nuclear speckled, or cytoplasmic staining patterns. Representative images were captured at 20x and with a constant exposure time of 250 ms.

ALT measurement in blood plasma

Plasma samples and controls were mixed with Infinity ALT reagent according to manufacturer’s instructions, and the absorptions read at 340 nm on a SuperMax i3 plate reader every 28 seconds for 10 min with a 1-minute initial lag time. ALT activity was determined by the slope of NADH degradation in a coupled enzyme reaction, using Softmax Pro 6.3 analytical software.

Total IgG ELISA

High-binding ELISA 96-well plates (Fisher Scientific) were coated with anti-mouse IgG (SouthernBiotech) in PBS overnight at 4 °C. Plates were washed three times with ELISA wash buffer (1x PBS and 0.05% Tween 20) and blocked with blocking buffer (2% FBS in wash buffer) for 2 h at room temperature. Plasma samples were diluted 1:50,0000 in PBS and the plates were incubated overnight at 4 °C. Anti-mouse IgG-horseradish peroxidase (HRP) (SouthernBiotech) in blocking buffer was added to plates and incubated for 1 h at room temperature. Plates were washed five times with wash buffer. HRP substrate TMB was added, and the reaction stopped by the addition of 2 N H2SO4. Absorbance was read at 450 nm. Levels of total IgG were expressed in OD values.

In vivo anti-IFNγ Ab treatment

PB of >20-wk-old Tet2^ΔVAV mice and littermate controls were analyzed for myelo-expansion by flow cytometry prior to treatment. Tet2^ΔVAV mice with myelo-expansion (frequency of CD11b⁺ Gr1⁺ >16% PB [Meisel et al., 2018]; see Figure S3B) and littermate controls were subjected to weekly intraperitoneal injections of 1 mg anti-IFNγ (XMG1.2, Bio X Cell) or IgG1 isotype control (HRPN, Bio X Cell), for a total period of three weeks. In vivo IFNγ depletion was confirmed by efficient suppression of plasma IFNγ levels in Tet2^ΔVAV mice (data not shown).

Bacterial colony identification

Bacterial colonies were identified as described in Meisel et al. (2018). In brief, grown colonies were picked with sterile pipette tips and stored at −80 °C until analysis. At day of analysis, picked bacterial colonies were thawed at room temperature (RT), resuspended with 20 μL of sterile water and lysed at 95 °C for 10 min. Samples were subsequently cooled down to 4 °C and then the DNA (2 μl) was used as template DNA in PCR reactions amplifying the 16S rRNA gene using universal bacterial 16S rRNA primers (27F, 5′-AGAGTTTGATCMTGGCTCAG-3′ and 1525R, 5′-AAGGAGGTGATCCAGCC-3′) with reaction conditions: 95 °C for 5 min followed by 35 cycles of 95 °C for 30 sec, 55 °C for 30 sec, 72 °C for 2 min and then 72 °C for 20 min. The amplification product (8 μl) was incubated with 2 μl ExoSAP-IT^™ (ThermoFisher) for 37 °C for 15 min, followed by 80 °C for 15 min. As recently described (Meisel et al., 2018), amplicons were sequenced by capillary sequencing, and the resulting sequences were analyzed using BLASTN and the 16S ribosomal RNA sequences NCBI database for species identification.

Oral gavage of Lactobacillus spp. to test for ability to induce AIH-like pathology

L. spp. were grown in MRS broth for 12-15 h. Bacteria were harvested by centrifugation (5000 x g for 10 min), washed and resuspended in PBS at a density of 5x10⁹ CFU/mL (CFU was quantified by plating dilutions on MRS agar). Naïve SPF mice were gavaged with 10⁹ CFU of L. spp. in 200μl PBS. To assess impact of L. spp. on hepatic type 1 T cell effector function, as well as PB plasma levels of ANA, SPF mice were, similar to Meisel et al. (2018), pre-treated with broad-spectrum antibiotics for 10 days (100μl oral gavage), followed by oral gavage with 10⁹ CFU four times (day 0, 1, 4 and 5). End point analysis (EPA) was performed on day 6. To define the impact of L. reuteri on hepatocyte damage (ALT in plasma), mice were treated for an additional week. In Figures 7A-7C and 7I-7K mice were pre-treated with broad-spectrum antibiotics for 10 days (100μl oral gavage) as previously described (Meisel et al., 2018), followed by oral gavage with 10⁹ CFU with L. reuteri or ΔArAT L. reuteri up to 5 wk (five days in a week). PB plasma ANA and ALT level were determined on day 0 (before L. reuteri gavage) and EPA on day 35.

Assessment of bacterial translocation efficiency and ability to produce AhR ligands

L. reuteri: 6-8-wk-old SPF Tet2^−/− and Tet2^+/+ mice were orally gavaged (200μl) with 10⁹ CFU L. reuteri for 6 consecutive days. EPA was performed 24 h after final oral gavage, where livers were aseptically removed, weighed, and homogenized by mashing through a 100 μm cell strainer in 3 mL sterile water containing 0.05% NP-40. Bacterial expansion was performed under anaerobic conditions at 37 °C as follows. After 1 h incubation, 3 mL MRS broth was added to each sample followed by an additional 23 h of incubation, after which, 500 μL of each sample was expanded in 5 mL MRS broth. Following an incubation of 24 h, the expansion and subsequent serial dilutions were plated on MRS agar and plates were incubated for 24 h. Single colonies were selected for 16s sequencing to confirm translocation of L. reuteri.

AhR activity measurement in translocated bacteria in the liver

For assessments of AhR activity, MRS broth-expanded cultures were pelleted at 5000 x g for 10 min, resuspended in Peptone Tryptone Tryptophan (PTT) media, and cultured anaerobically at 37 °C, shaking at 250 rpm, for 14 h. Bacteria were again pelleted and supernatant was collected and measured for AhR activity.

AhR activity measurement in liver homogenate

Pieces of liver (~75 mg) from the left lobe were removed and mashed through 100 μm cell strainers in PBS, followed by centrifugation at 1800 rpm for 5 min to pellet cells. Supernatant was collected and measured for AhR activity.

AhR activity assay

HT29-Lucia^™ AhR cells were purchased from InvivoGen (ht2l-ahr). Cells were cultured in DMEM (Gibco) supplemented with 10% FBS, 1x Penicillin-Streptomycin-Glutamine (Gibco), 100 ug/mL Normocin (InvivoGen), and 100 ug/mL selective antibiotic Zeocin (InvivoGen). Briefly, 20 μL of sample was incubated with ~50,000 HT29-Lucia^™ AhR Cells for 24 h. Following incubation, 20 uL of supernatant was transferred into a 96-Well Clear Bottom Black Microplate (Corning) and 50uL QUANTI-Luc^™ (InvivoGen) was added. Samples were immediately read for luminescence via a SpectraMax^® i3x using SoftMax Pro 3.0.7 Software under the following settings: read type endpoint at all wavelengths, integration time 100ms, read height 2mm.

AhR agonist containing supernatant

Similar to previously described (Cervantes-Barragan et al., 2017), to generate AhR agonist-containing supernatants, L. reuteri or ΔArAT L. reuteri were grown in MRS broth overnight, harvested by centrifugation, washed with PBS and ~1x10¹⁰ CFU were inoculated into 10 mL of peptone-tryptone water (10 g/L peptone and 10 g/L tryptone, 5 g/L NaCl) supplemented with 0.6 mM L-tryptophan. After a 14 h incubation at 37 °C under anaerobic conditions, cells were centrifuged (5000 x g, 10 min), supernatant collected and filter sterilized (0.2-μm pore diameter cellulose acetate filter (VWR)) and stored at −80 °C until further use. For in vitro experiments, supernatants were added to cells at a final concentration of 10% vol/vol in culture media (Cervantes-Barragan et al., 2017).

DNA extraction from liver tissue and intestinal contents

Microbial DNA was extracted from a piece of the left liver lobe (~30mg), harvested under sterile conditions, using the Microbiome Kit (Qiagen). The Fast DNA Stool Mini Kit (Qiagen) was used to extract total DNA from intestinal contents and intestinal scrapings. Quantitative PCR (qPCR) of 16S rRNA-encoding gene was performed as recently described (Meisel et al., 2017). Briefly, qPCR was performed on a BioRad CFX384 (BioRad) using the primers (340F, 5’-ACTCCTACGGGAGGCAGCAGT-3’) and (514R, 5’-ATTACCGCGGCTGCTGGC-3’). Reactions were run at 95 °C for 3 min, followed by 40 cycles of 95 °C for 15 min and 63 °C for 60 sec. Specific amplification of targets was quantified using dilution curves of a purified pCR4-TOPO vector (Invitrogen) containing a cloned 16S rRNA-encoding gene from Blautia producta (Prevot) (ATCC^®, 27340D-5^™) as a standard (Huse et al., 2014) for liver and intestinal contents. Standards ranging in concentration from 10⁸ – 10⁰ plasmid copies/μl were run in parallel with our samples during each qPCR run. Using those results, a standard curve was generated to quantify the number of copy numbers within the samples. To determine the bacterial load in the intestinal content/scraping samples, the results were normalized to their corresponding weight. Bacterial 16S rRNA gene RT-PCR from livers was normalized to the host murine Infb1 gene (forward: 5’-CCATCCAAGAGATGCTCCAG-3’; reverse 5’- GTGGAGAGCAGTTGAGGACA-3’).

16S rRNA amplicon library preparation, sequencing and data analysis

16S rRNA amplicon library preparation and sequencing of microbial hepatic DNA, intestinal luminal DNA, and fecal DNA (see above for extraction techniques) was performed as previously described (Meisel et al., 2017). Generated libraries underwent V3-V4 16S amplicon sequencing using the NextGen Illumina MiSeq platform (University of Pittsburgh Genomics Research Core). Raw sequencing data were processed in QIIME^™ 1.9.1 (Caporaso et al., 2010b). Sequences were demultiplexed, denoised, and clustered into operational taxonomic units (OTUs) and aligned via PyNAST (Caporaso et al., 2010a). Taxonomy was assigned using the RDP Classifier (Wang et al., 2007) using GreenGenes (v13.8). Processed data were then imported into Calypso 8.84 for data analysis and visualization (Zakrzewski et al., 2017). Samples were rarified to a depth of 5000 reads per sample. For liver, samples were analyzed with and without inclusion of g_Burholderia, where the latter were rarified to a depth of 500 due to the dominance of this excluded taxa in tissue samples. To assess alpha diversity, the Shannon, Simpson’s and Richness diversity indices were calculated and significant differences were identified using ANOVA. Beta diversity were assessed using principal coordinates analysis (PCoA) plots of Bray-Curtis dissimilarity distances and Redundancy analysis (RDA) (Anderson, 2006). Statistical significance of beta diversity clustering was assessed using permutational multivariate analysis of variance (PERMANOVA) with 999 permutations followed by permutational analysis of multivariate dispersions (PERMDISP2) to assess the homogeneity of group variance (distance to centroid). Phylogenetic investigation of communities by reconstruction of unobserved states (PICRUSt) (Langille et al., 2013), a technique which uses evolutionary modeling, was used to predict metagenomes from 16S data and reference genome databases (Kanehisa and Goto, 2000). Heatmaps were generated using Spearman’s rank correlation coefficient with dendrograms illustrating relatedness of sample groups and taxa abundance.

In vitro naïve CD4 and CD8 T cell stimulation with L. reuteri-derived AhR ligand

Naïve CD8 or CD4 T cells were purified with respective purification kits (Stem cell Technologies). Purity test by flow cytometry revealed 95%–98% enriched TCRβ⁺ CD44⁻ CD62L⁺ cells (data not shown). 5x10⁵ naïve CD4 or CD8 T cells were stimulated with plate coated anti-CD3 and soluble anti-CD28 (each 1 μg/mL) in the presence of 10% L. reuteri- or 10% ΔArAT L. reuteri-supernatant with or without 30 μM AhR antagonist (CH-223191, Sigma-Aldrich). Amount of produced IFNγ was assessed after 72 h by ELISA.

Enzyme-linked immunosorbent assay to measure IFNγ

In vitro T cell supernatants were used to measure IFNγ according to manufacturer’s instructions. Briefly, ELISA plates were coated overnight at 4 °C with 0.5 μg/ml IFNγ (BD Biosciences). Plates were washed once (0.05% Tween 20 in PBS) and blocked for 1 h with blocking buffer (5% FBS in PBS). Standards and samples were incubated overnight at 4 °C. Plates were washed and incubated with biotin-conjugated IFNγ detection Ab (0.5 μg/ml, BD Biosciences) for 1 h at RT. Plates were then washed and incubated with Horseradish Peroxidase-conjugated streptavidin (Jackson ImmunoResearch) for 15 min followed by washing and development with TMB substrate (Thermo Fisher Scientific). Reactions were stopped by the addition of 2N H₂SO₄, and absorbance was measured at 450 nm on a SpectraMax i3 plate reader (Molecular Devices). IFNγ concentration in supernatants was determined from a standard curve generated by serial dilution of IFNγ.

Measurement of plasma ANA, LKM1 and SLA

For ANA and SLA, plasma samples were diluted 1:10 and for LKM1 1:5 in sample diluent. Plasma anti-nuclear Abs (ANA), soluble liver antigen Abs (SLA), and liver kidney microsome Abs (LKM1) levels were determined by using the mouse specific Anti-nuclear antibody (ANA), soluble liver antigen antibody (SLA) and, liver kidney microsome autoantibody (Anti-LKM) ELISA Kit (MyBioSource) according to manufacturer’s instructions. Absorbance was measured at 450nm on a SpectraMax i3 plate reader (Molecular Devices). ANA, SLA, and LKM1 concentrations in plasma were determined from a standard curve generated by serial dilution of ANA, SLA, and LKM1.

Hematoxylin/eosin and Trichrome staining

A piece of the left liver lobe (5 x 5 mm) was fixed in 10% formalin for 24 h at RT, transferred to 70% ethanol, and processed for embedding in paraffin. Hematoxylin and eosin (H&E) and trichrome staining was performed on 4 μm paraffin-embedded liver sections. Staining was performed by the Clinical Biospecimen Repository and Processing Core (Pittsburgh Liver Research Center). The slides were imaged by LAS V4- Leica microscope. Fibrotic area was quantified using ImageJ Software by averaging the percent of quantified fibrotic area/ field of view (FOV) of >3 FOV/liver.

Histological grading for key AIH features

Using hematoxylin and eosin-stained slides, liver histological features that are either compatible with AIH or typical of AIH, such as interface hepatitis, lymphoplasmacytic infiltrate, rosette formation, and emperipolesis were scored in a blinded fashion by a gastrointestinal and hepatopancreatobiliary trained pathologist (ADS) on a scale from 0 to 3. Score: 0, negative; 1, mild; 1.5, mild-moderate: 2; moderate; 3, marked. Each dot represents an individual mouse.

TUNEL staining

Liver (left lobe) tissue was fixed in 10% formalin and processed for paraffin sectioning. 4 μm sections were dewaxed by immersion in xylene and hydrated by serial immersion in ethanol and PBS. Antigen retrieval was performed by incubating sections in Proteinase K (1 μg/ml, Invitrogen) at 37 °C for 30 min. Sections were washed with PBS (4x for 2 min) and blocking buffer (0.1M Tris-HCl containing 3% BSA and 20% normal bovine serum) was added for 30 min. TUNEL reaction mixture was added to liver sections according to the manufacturer’s instructions (Roche Applied Science). Sections were stained with DAPI (1 μg/ml, Invitrogen) and mounted with Fluoromount-G (Beckman Coulter). At least three images were taken randomly from each section by fluorescent microscopy (Olympus, IX83). 3-4 FOV were randomly selected from each section, 100 DAPI⁺ cells were counted for each FOV and double positive (DAPI⁺ fluorescent red nuclei) were considered TUNEL⁺ cells. Data are shown as the number of TUNEL⁺ cells per FOV.

In vivo AhR antagonist treatment

Test for requirement of AhR signaling in L. reuteri-induced AIH-like pathology: 6-wk-old Tet2^ΔVAV mice and Tet2 cre⁻ littermate controls were used and treated with 10*9 CFU (OG, oral gavage) along with either vehicle (corn oil) or AhR antagonist (CH223191; 300 μg/mouse) via intraperitoneal (i.p.) injection for 5 wk (five days in a week), and end point analysis (EPA) performed on day 35. Test for requirement of AhR signaling in AIH-like diseased Tet2^ΔVAV mice: > 20-wk-old AIH-like diseased Tet2^ΔVAV mice were treated with i.p. injections of AhR antagonist (CH-223191, Sigma-Aldrich) or vehicle control (corn oil) for 5 wk (five days in a week) and EPA performed on day 35.

Carbon tetrachloride and Concanavalin A treatment

6-wk-old Tet2^ΔVAV and Tet2 cre- mice were treated twice per week for 6 weeks with CCl₄ (0.5ml CCl₄/kg body weight) or vehicle (corn oil) via intraperitoneal injection (Scholten et al., 2015). To assess impact of ConA-treatment on liver injury and fibrosis, 6-wk-old Tet2^ΔVAV and Tet2 cre- mice were treated once with ConA (5 mg/kg body weight) or vehicle (PBS) iv (retroorbital injection) and analyzed after 42 h of administration.

In vitro naïve CD8 T cell stimulation for RNA-seq analysis

Naïve CD8 T cells were purified by a naïve CD8 purification Kit (Stem cell Technologies) and TCRβ⁺ NK1.1⁻ TCRγδ⁻ MHCII⁻ CD44⁻ CD62L⁺ cells were further enriched by FACS sorting. 3x10⁵ T cells were added per well and stimulated with plate bound anti-CD3 and soluble anti-CD28 (each 1μg/ml). Cells were treated with vehicle (DMSO), 1μM I3A (Sigma-Aldrich) with or without 30μM AhR antagonist (CH-223191, Sigma-Aldrich). Cells were lysed for RNA sequencing analysis after 8 h.

Sample preparation for quantitative LC-HRMS

Metabolic quenching and polar metabolite pool extraction was performed by adding 360 μL ice cold methanol containing 10 μM ¹³C₁-creatine (Sigma-Aldrich, Boston MA) to 90 μL sample. After 3 min of vortexing, the supernatant was cleared of protein by centrifugation at 16,000 x g. 450 μL cleared supernatant was dried to completion under nitrogen gas and resuspended in 90 μL water. 2 μL of resuspended sample was subjected to online LC-MS analysis. Molar quantities were calculated using a calibration curve using indole-3-aldehyde (I3A, Sigma-Aldrich) in a 1:3 series dilution from 25 μM to 11 nM.

LC-HRMS Analysis

Briefly, samples were injected via a Thermo Vanquish UHPLC and separated over a reversed phase Phenomenex Kinetix C18 column (2.1 × 150mm, 1.7 μm particle size) maintained at 55 °C. For the 10 min LC gradient, the mobile phase consisted of the following: solvent A (water/5 mM ammonium formate/0.1% formic acid) and solvent B (methanol/0.1% formic acid). The gradient was the following: 0 - 0.3 min 3% B, increase to 30 % B over 0.5 min, continue increasing to 60% B over 1 min, hold at 60 % B for 1.3 min, increase to 95% B over 0.5 min, hold at 95% B for 1.4 min, equilibrate at 3% B for 4.5 min. The Thermo ID-X tribrid mass spectrometer was operated in positive ion mode, scanning in ddMS² mode (2 μscans)from 70 to 800 m/z at 120,000 resolution with an AGC target of 2e5 for full scan, 2e4 for MS² scans using HCD fragmentation at stepped 15,35,50 collision energies. Source ionization setting was 3.0 kV spray voltage for positive mode. Source gas parameters were 35 sheath gas, 12 auxiliary gas at 320 °C, and 8 sweep gas. Calibration was performed prior to analysis using the PierceTM FlexMix Ion Calibration Solutions (Thermo Fisher Scientific). Integrated peak areas were then extracted manually using Quan Browser (Thermo Fisher Xcalibur ver. 2.7). I3A was quantified using the calibration curve created with the I3A standard and ¹³C-₁-creatine internal standard.

In vitro naïve CD8 T cell stimulation with mouse recombinant IL-12

Naïve CD8 cells were purified with a naïve CD8 purification Kit (Stem cell Technologies). Purity test by flow cytometry revealed 95%–98% enriched TCRb⁺ CD44⁻ CD62L⁺ cells (data not shown). 5x10⁵ naïve CD8T cells were stimulated with different concentrations of plate coated anti-CD3 (BD Biosciences; 5 μg/ml) and soluble anti-CD28 (BD Biosciences; 2 μg/ml) respectively, with different doses of recombinant IL-12 (BioLegend; 0.1, 1.0 and, 10 ng/ml). Amount of produced IFNγ was assessed after 72 h by ELISA.

Assessment of STAT1 phosphorylation and IFNγR1 expression in CD8 T cells

Naïve CD8 T cells were purified by a naïve CD8 purification Kit (Stem cell Technologies). 10⁶ naïve CD8 T cells were added per well into a 96-well flat-bottom plate and stimulated with plate bound anti-CD3 and soluble anti-CD28 (each 1 μg/ml) for 8 h. Cells were either treated with mouse recombinant IFNγ, 0.2 μg/ml (R & D Systems) for 40 min or left untreated. Fixable live/dead viability stain (Invitrogen) was added 10 min prior to fixation. Cells were fixed with BD Phosflow Lyse/Fix Buffer (BD Biosciences) and subsequently incubated for 15 min at 37 °C in a water bath. Next, samples were permeabilized with ice-cold pre-chilled True-Phos^™ Perm Buffer (BioLegend) and incubated at −20 °C for 1 h. After washing, cells were incubated with Abs for CD45, CD4, CD8, TCRβ and anti-phospho-STAT1 (BioLegend) in FACS buffer for 30 min at RT. After washing twice, samples were acquired on an Aurora (Cytek) flow cytometer and analyzed with FlowJo 10 (Tree Star). To determine the surface expression of IFNγR1 on splenic CD8 T cells, naïve CD8 T cells were stimulated with anti-CD3 and anti-CD28 (both 1 μg/ml) for 48h followed by flow cytometry analysis for surface IFNγR1 expression (Figure S6G).

Statistical analysis

The majority of experiments were repeated at least two times to obtain data for indicated statistical analyses. Mice were allocated to experimental groups on the basis of their genotype and randomized within the given sex- and age-matched group. Given that our mice were inbred and matched for age and sex, we assumed similar variance between the different experimental groups. We did not perform a priori sample size estimation but always used as many mice per group as possible in an attempt to minimize type I and type II errors. Investigators were not blinded during experiments and outcome assessment, except for microscopic analysis of fluorescent immunostaining and HEp-2 ANA intensity analysis, which has been performed blinded. All experimental and control animals were littermates and none were excluded from the analysis at the time of harvest. All quantitative data are presented as mean ± standard error of the mean (sem), unless otherwise indicated. Data were analyzed using an unpaired two-tailed Student’s t-test for single comparisons, and one-way or 2-way ANOVA for multiple comparisons. ANOVA analysis was followed by a Sidak’s post-hoc test. Correlations were calculated using the Pearson correlation. Figures and statistical analysis were generated using GraphPad Prism 9 (GraphPad Software). The statistical test used, and P values are indicated in each figure legend. P values of < 0.05 were considered statistically significant. *P < 0.05, **P <0.01, ***P <0.001 and ****P <0.0001.

Highlights.

• Lack of hematopoietic Tet2 drives microbiota- and IFNγ-dependent AIH-like disease in mice

• AIH-like pathology associates with hepatic AhR agonist-producing pathobiont expansion

• Liver translocated L. reuteri releases I3A driving AIH-like disease

• Tet2 within CD8 T cells antagonizes I3A-induced AIH-like pathology

INCLUSION AND DIVERSITY

We worked to ensure sex balance in the selection of nonhuman subjects. One or more of the authors of this paper self-identifies as an underrepresented ethnic minority in science. One or more of the authors of this paper self-identifies as a member of the LGBTQ+ community. One or more of the authors of this paper received support from a program designed to increase minority representation in science.