Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 1.
Published in final edited form as: Hepatology. 2021 Aug 30;74(5):2745–2758. doi: 10.1002/hep.32010

Gender-Specific Regulation of Interferon-γ Cytotoxicity in Mouse Liver by Autophagy

Yang Shen 1, Francesca Cingolani 1, Shoaib Ahmad Malik 2, Jing Wen 1, Yunshan Liu 1, Mark J Czaja 1
PMCID: PMC8542567  NIHMSID: NIHMS1714613  PMID: 34118081

Abstract

Background & Aims:

Interferon-γ (IFNγ) is a central activator of immune responses in the liver and other organs. IFNγ triggers tissue injury and inflammation in immune diseases which occur predominantly in females for unknown reasons. Recent findings that autophagy regulates hepatotoxicity from proinflammatory cytokines led to an examination of whether defective hepatocyte autophagy underlies gender-specific liver injury and inflammation induced by IFNγ.

Approach & Results:

A lentiviral Atg5 knockdown to decrease autophagy sensitized AML12 hepatocytes to death from IFNγ in combination with IL-1β or TNF. Death was necrosis due to impaired energy homeostasis and ATP depletion. Male mice with decreased autophagy from a tamoxifen-inducible, hepatocyte-specific Atg5 knockout were resistant to IFNγ hepatotoxicity whereas female knockout mice developed liver injury and inflammation. Female mice had increased IFNγ-induced signal transducer and activator of transcription 1 (STAT1) levels compared to males. Blocking STAT1 but not interferon regulatory factor 1 signaling prevented IFNγ-induced hepatocyte death in autophagy-deficient AML12 cells and female mice. The mechanism of death is STAT1-induced overexpression of nitric oxide synthase 2 (NOS2) as in vitro hepatocyte death and in vivo liver injury were blocked by NOS2 inhibition.

Conclusions:

Decreased hepatocyte autophagy sensitizes mice to IFNγ-induced liver injury and inflammation through overactivation of STAT1 signaling that causes NOS2 overexpression. Hepatotoxicity is restricted to female mice suggesting that gender-specific effects of defective autophagy may underlie the increased susceptibility of females to IFNγ-mediated immune diseases.

Keywords: female, gender, interferon regulatory factor 1, necrosis, nitric oxide synthase 2, signal transducer and activator of transcription 1


Interferon-γ (IFNγ) plays a critical role in host defense through immunoregulatory functions that promote macrophage activation and tissue inflammation. Additional direct effects of IFNγ on nonimmune cells such as the induction of cell cycle arrest or apoptosis in hepatocytes have been reported.(1) These functions of IFNγ make its dysregulation an important pathophysiological mechanism of diseases of the liver and other organs. Overactive IFNγ signaling contributes to the development of immune-mediated tissue injury in the liver,(2) and in systemic diseases such as systemic lupus erythematosus.(3) A perplexing and largely unexplained feature of immune diseases is their female gender predominance. The liver diseases autoimmune hepatitis and primary biliary cholangitis, and nonhepatic diseases such as systemic lupus erythematosus, are all much more prevalent in females than males.(4)

Injurious overactivation of the immune-mediated inflammatory response is a central mechanism of liver disease.(5) Transient inflammation is beneficial in tissue repair, but excessive inflammation drives hepatocyte injury and death that leads to chronic inflammation, liver fibrosis and carcinogenesis. A potential mechanism of the detrimental effects of sustained inflammation is the direct cytotoxicity of cytokines such as TNF, IL-1β and IFNγ on the liver. Except for well-defined apoptotic and necrotic TNF death pathways, the mechanisms by which cytokines trigger hepatocyte injury and death have not been well defined. The failure of anti-TNF therapy in human liver disease makes it imperative to better understand the cytotoxic and proinflammatory effects of other cytokines such as IFNγ.

Recent studies have begun to define functions of the lysosomal degradative pathway of macroautophagy (hereafter referred to as autophagy) in regulating hepatocyte death after liver injury.(6,7) The importance of impaired autophagy in liver disease is suggested by the finding that multiple factors that sensitize the liver to injury, including excessive lipid accumulation, toxins, and aging, also decrease levels of hepatocyte and/or macrophage autophagy.(8-12) Our recent findings have demonstrated that decreased autophagy promotes liver injury and inflammation from the immune activator lipopolysaccharide (LPS) and the LPS-inducible cytokines TNF and IL-1β.(8,1315) The present study examined the function of hepatocyte autophagy in the liver’s gender-specific response to IFNγ.

Decreased autophagy sensitized AML12 hepatocytes to IFNγ-induced necrosis from ATP depletion. Mice with a hepatocyte-specific decrease in autophagy developed liver injury and inflammation from IFNγ, but toxicity was restricted to female mice. Compared to male mice, IFNγ-treated females developed increased signal transducer and activator of transcription 1 (STAT1) signaling causing nitric oxide synthase 2 (NOS2) overexpression that triggered hepatocyte death. These findings are the first to define a critical function for autophagy in hepatocyte resistance to the toxic effects of IFNγ and identify a novel cellular mechanism to explain gender differences in IFNγ-mediated tissue injury.

Experimental Procedures

Studies were conducted in AML12 control Vec cells stably infected with a lentivirus containing empty vector and siAtg5 cells infected with a lentivirus expressing an shRNA to the autophagy gene Atg5.(15) In vivo studies were performed in Atg5Δhep mice with a tamoxifen-inducible, albumin promoter-driven Atg5 knockout.(15) Cre expression was induced by subcutaneous injection of 3 mg of tamoxifen (MilliporeSigma, St. Louis, MO). Tamoxifen-injected littermate Atg5F/F mice lacking the Cre transgene served as controls for all experiments.

Supporting Information provides additional detailed methods.

Results

Decreased autophagy sensitizes AML12 hepatocytes to death from IFNγ

The function of autophagy in hepatocyte IFNγ cytotoxicity was examined in siAtg5 AML12 hepatocytes with decreased autophagy from a stable lentiviral knockdown of the autophagy gene Atg5.(15) By trypan blue staining at 24 h, IFNγ was nontoxic to both control Vec cells infected with empty lentiviral vector and siAtg5 cells (Fig. 1A,B). As previously reported,(15) siAtg5 but not Vec cells had a modest but significant increase in death from IL-1β (Fig. 1A,B). Combined IFNγ/IL-1β treatment led to significant death in both cell types, but cell death was 7-fold greater in knockdown cells (Fig. 1A,B). Increased death in IFNγ/IL-1β-treated siAtg5 cells was confirmed by a second cell death assay of propidium iodide staining (Fig. 1C,D). Decreasing autophagy by pharmacological inhibition of lysosomal function with chloroquine also sensitized AML12 cells to death from IFNγ/IL-1β (Fig. 1E).

FIG. 1.

FIG. 1.

Decreased autophagy sensitizes hepatocytes to death from IFNγ. (A) Representative images of trypan blue-stained Vec and siAtg5 cells untreated (Con) or treated with the indicated cytokines for 24 h (200X). (B) Numbers of IFNγ and/or IL-1β treated trypan blue-positive cells (*P<0.001 compared to untreated Vec cells; #P<0.00001 compared to Vec cells with the same treatment; n=9). (C) Representative images of propidium iodide-stained cells (200X). (D) Numbers of propidium iodide-positive cells untreated (Con) or treated with IFNγ and/or IL-1β (*P<0.00001 compared to untreated Vec cells; #P<0.000001 compared to Vec cells with the same treatment; n=6). (E) Numbers of trypan blue-positive wild-type AML12 cells untreated (Con) or treated with chloroquine (CQ) and/or IFNγ/IL-1β (I/I) for 24 h (*P<0.000001 compared to untreated cells; #P<0.000001 compared to cells treated with IFNγ/IL-1β; n=8). (F) Numbers of trypan blue-positive cells treated with IFNγ and/or TNF (*P<0.001 compared to untreated Vec cells; #P<0.00001 compared to Vec cells with the same treatment; n=6-9). (G) Immunoblots of wild-type AML12 cells untreated or IFNγ/IL-1β treated for 6 h with some cells pretreated with bafilomycin (Baf). Arrows indicate molecular weights in KD and the LC3-I and -II forms. (H) Ratios of LC3-II with bafilomycin treatment to that without bafilomycin in cells untreated (Con) or treated with IFNγ/IL-1β quantitated by densitometric scanning of immunoblots from three independent experiments.

To exclude the possibility that IFNγ amplifies death from IL-1β, toxicity of IFNγ in combination with TNF was determined. By trypan blue staining, TNF was nontoxic to control and knockdown cells (Fig. 1A,F). IFNγ/TNF in combination triggered death in both cell types which was 14-fold greater in siAtg5 cells (Fig. 1A,F). Decreased autophagy sensitizes AML12 hepatocytes to IFNγ-dependent death in combination with other proinflammatory cytokines.

Constitutive levels of autophagy may be sufficient for resistance to IFNγ cytotoxicity, or IFNγ may increase autophagy to protective levels. To distinguish between these two possibilities, the effect of IFNγ on autophagy was assessed. IFNγ/IL-1β treatment did not alter autophagic flux as indicated by immunoblot findings of equivalent LC3-II levels in untreated and bafilomycin-treated wild-type AML12 cells (Fig. 1G). Quantification of autophagic flux by a densitometric determination of the ratio of LC3-II levels in cells treated with bafilomycin to that in cells without bafilomycin revealed no change from IFNγ/IL-1β (Fig. 1H). The failure of IFNγ to increase autophagy indicates that constitutive levels of autophagy provide hepatocyte resistance to IFNγ cytotoxicity.

IFNγ/IL-1β death is necrosis secondary to impaired energy homeostasis

The fact that IFNγ receptor levels can be altered by liver injury,(1) led to an examination of the effects of decreased autophagy on cytokine receptors as a possible mechanism of siAtg5 cell death from IFNγ. Vec and siAtg5 cells have equivalent mRNA levels for the IFNγ and IL-1 receptor genes that were unaffected by IFNγ/IL-1β treatment (Fig. 2A), indicating that death was not the result of altered cytokine receptor expression.

FIG. 2.

FIG. 2.

IFNγ induces necrosis due to ATP depletion. (A) Relative mRNA levels for IFNγ and IL-1β receptor genes in Vec and siAtg5 cells untreated (0h) and treated for the indicated hours with IFNγ/IL-1β (n=6). (B) Numbers of trypan blue-positive siAtg5 cells treated with DMSO vehicle, or IFNγ/IL-1β after pretreatment with DMSO, Q-VD-OPh (QVD) or necrostatin-1s (Nec) (n=10). (C) Relative ATP levels in untreated control (Con) and 24 h IFNγ and/or IL-1β treated cells (*P<0.05, **P<0.01 compared to untreated Vec cells; #P<0.00001 compared to Vec cells with the same treatment; n=6). (D) ATP content in untreated Vec cells (Vec) and siAtg5 cells (Con), and siAtg5 cells treated for 24 h with IFNγ/IL-1β alone (I/I) or together with a 24 and 1 h pretreatment of oleate (I/I/O) or pyruvate (I/I/P) (*P<0.01, **P<0.001 compared to untreated Vec cells; #P<0.002 compared to IFNγ/IL-1β-treated cells; n=6). (E) Numbers of trypan blue-positive siAtg5 cells treated with IFNγ/IL-1β alone (I/I) or with a 24 and/or 1 h pretreatment of oleate (I/I + Ol) or pyruvate (I/I + Pyr) (*P<0.00001 compared to I/I cells; n=6). (F) Numbers of trypan blue-positive siAtg5 cells that received DMSO vehicle alone or DMSO, Cathepsin L Inhibitor III (CI3), or Cathepsin L Inhibitor IV (CI4) for 1 h prior to IFNγ/IL-1β for 24 h (n=6).

Trypan blue and propidium iodide positivity indicate necrosis which can be primary or secondary to another form of cell death. To exclude secondary necrosis, the effects of caspase and necroptosis inhibitors on siAtg5 cell death were examined. IFNγ/IL-1β death was not prevented by the caspase inhibitor Q-VD-OPh or the necroptosis inhibitor necrostatin-1s (Fig. 2B), demonstrating that IFNγ-induced death in siAtg5 cells is a primary necrosis.

Necrosis often results from compromised energy homeostasis. In agreement with previous findings,(15) decreased autophagy resulted in a modest but significant decrease in ATP in siAtg5 cells (Fig. 2C). Treatment with IFNγ or IL-1β alone did not reduce ATP content in either cell type (Fig. 2C). Combined cytokine treatment significantly decreased ATP levels in both cell types, but levels in siAtg5 cells were 44% less than in Vec cells (Fig. 2C). To determine whether a decreased supply of substrates caused ATP depletion and death, siAtg5 cells were supplemented with the critical autophagy-generated energy substrate oleate.(9) Oleate failed to reverse ATP depletion in IFNγ/IL-1β-treated siAtg5 cells (Fig. 2D), or prevent death (Fig. 2E). Death was the result of impaired energy homeostasis, however, as the glycolytic pathway metabolite pyruvate, which restored ATP levels (Fig. 2D), also blocked cell death (Fig. 2E). Hepatocytes with defective autophagy develop impaired energy homeostasis when stressed with IFNγ/IL-1β that triggers necrosis, but the mechanism of ATP depletion and death is not the loss of autophagy-supplied energy substrates.

Our prior studies demonstrated that death from IL-1β/TNF in hepatocytes with decreased autophagy results from lysosomal permeabilization and cathepsin-dependent necrosis.(15) Cathepsin inhibitors that blocked death from IL-1β/TNF(15) had no effect on IFNγ/IL-1β death (Fig. 2F). IFNγ-induced necrosis in the setting of impaired autophagy occurs by a mechanism distinct from that for IL-1β toxicity.

Female Atg5Δhep mice are sensitized to liver injury from IFNγ

The effect of decreased autophagy on IFNγ hepatotoxicity was examined in Atg5Δhep mice with a tamoxifen-inducible, hepatocyte-specific Atg5 knockout.(15) Littermate control Cre and knockout Cre+ mice were tamoxifen-injected and administered IFNγ 4 days later, a time both sufficient for a gene knockout and prior to the onset of spontaneous liver injury from impaired autophagy.(15) Male mice were resistant to IFNγ toxicity as indicated by minimal increases in serum alanine aminotransferase (ALT) levels at 12 and 24 h that were equivalent in control and knockout mice (Fig. 3A). Similarly female control mice had mild increases in ALT 12 and 24 h after IFNγ treatment (Fig. 3B). In contrast, female Atg5Δhep knockout mice had markedly increased ALTs at all times after IFNγ injection (Fig. 3B). Male knockout mice had a slight increase in terminal deoxynucleotide transferase-mediated deoxyuridine triphosphate nick end-labeling (TUNEL)-positive cells at 6 and 12 h, but the numbers were not significantly different from control mice (Fig. 3C,S1). Numbers of TUNEL-positive cells were markedly increased in IFNγ-treated female knockout mice as compared to controls (Fig. 3D,S1). Histological evidence of liver injury and inflammation was restricted to IFNγ-treated female knockout mice (Fig. 3E). The effect of decreased autophagy on IFNγ hepatotoxicity was further examined through inhibition of lysosomal function by leupeptin which has been demonstrated previously to effectively decrease hepatic autophagy.(8) Leupeptin sensitized female mice to toxicity from IFNγ as indicated by increased ALTs and TUNEL staining (Fig. S2). Decreased autophagy therefore specifically sensitized female but not male mice to hepatocyte injury and death from IFNγ.

FIG. 3.

FIG. 3.

Female Atg5Δhep mice develop IFNγ hepatotoxicity. (A) Serum ALT levels in male control (Con) and Atg5Δhep knockout (KO) mice not injected (NI) or at the indicated hours after IFNγ injection (*P<0.05, **P<0.01 compared to untreated control mice; n=10). (B) Serum ALTs in female mice (*P<0.001 compared to untreated control mice; #P<0.00001 compared to control mice with the same treatment; n=11-13). (C) Numbers of TUNEL-positive cells per high power field (HPF) in male mice (*P<0.001 compared to untreated control mice; n=8). (D) TUNEL positivity in female mice (*P<0.01, **P<0.000001 compared to untreated control mice; #P<0.01, ##P<0.00001 compared to treated control mice; n=9-10). (E) Representative hematoxylin and eosin stained images of livers from female mice (200X). (F) Immunoblots of liver mitochondrial and cytosolic proteins from control and Atg5Δhep mice untreated or treated with IFNγ for 24 h, and wild-type (WT) mice treated with GalN/LPS for 6 h. Blots were probed for cytochrome c (Cyto c), BID, truncated BID (tBID), caspase 3 (Casp 3), caspase 7 (Casp 7), cytochrome oxidase (Cyto ox) as a mitochondrial protein loading and purity control, and tubulin as a cytosolic protein loading and purity control. Molecular weights in KD are indicted by the arrows. Immunoblots are representative of three independent experiments.

Hepatocyte TUNEL positivity can reflect either necrosis or apoptosis.(16) To exclude death by apoptosis, mitochondrial death pathway and caspase activation in IFNγ-treated female mice was examined by immunoblots. Mitochondrial death pathway activation as determined by BID cleavage into tBID and release of cytochrome c into the cytosol was not present in IFNγ-injured knockout livers but detected in mice treated with GalN/LPS (Fig. 3F). Downstream caspase 3 and 7 cleavage occurred with GalN/LPS but not IFNγ treatment (Fig. 3F). Thus, consistent with findings in AML12 cells IFNγ hepatocyte death in vivo was necrotic and not apoptotic.

Gender differences can result from the hormonal effects of estrogen. To determine whether male resistance to IFNγ hepatotoxicity is due to a lack of estrogen, male Atg5Δhep mice were implanted with slow-release 17β-estradiol pellets and challenged with IFNγ 60 days later. Mice implanted with vehicle- or 17β-estradiol-containing pellets had equivalent levels of IFNγ-induced injury by serum ALTs and TUNEL staining (Fig. S3). Female mouse sensitivity to IFNγ therefore cannot be explained by increased estrogen levels.

Female knockout mice have increased inflammation

Hepatocyte injury and death trigger a liver immune response and inflammation. Female control mice had no change in hepatic mRNA levels with IFNγ treatment for the macrophage marker gene Cd68 and a modest 6 h increase in the neutrophil marker Ly6g, whereas Atg5Δhep mice had significantly increased levels of both markers at 6 and 12 h (Fig. 4A). This increased inflammatory cell infiltration in knockout livers resulted in a greater induction of a variety of proinflammatory genes including the cytokines Tnf and Il1b (Fig. 4B), chemokine C-C motif chemokine ligand 2 (Ccl2) (Fig. 4C), and the inflammatory marker Nos2 (Fig. 4D). Male knockout mice had no increase in inflammatory cell recruitment (Fig. 4E) or proinflammatory genes in response to IFNγ (Fig. 4F), demonstrating that only female knockout mice developed hepatic inflammation from IFNγ.

FIG. 4.

FIG. 4.

Female but not male Atg5Δhep knockout mice have increased hepatic inflammation. (A-D) Relative hepatic mRNA levels of the indicated genes in female control (Con) and Atg5Δhep knockout (KO) mice not injected (NI) or administered IFNγ for 6 or 12 h (*P<0.05, **P<0.01 compared to control mice with no injection; #P<0.05, ##P<0.01 compared to control mice with the same treatment; n=6-7). (E,F) mRNA levels in male mice (n=6).

IFNγ-induced liver injury is IL-1β independent

The in vitro and in vivo findings differ somewhat in that toxicity in AML12 hepatocytes requires IL-1β in addition to IFNγ whereas IFNγ alone is sufficient in mice. Findings of increased Il1b mRNA in female mice suggested that this difference might reflect the presence in mice of IFNγ-induced IL-1β leading to combined toxicity from exogenous IFNγ and endogenous IL-1β. To exclude this possibility, effects of the IL-1R antagonist anakinra on mouse liver injury from IFNγ were examined. IL-1 neutralization had no effect on IFNγ-induced hepatic injury at 6 h as determined by ALT levels (Fig. 5A), TUNEL staining (Fig. 5B, S4A), immune cells markers (Fig. 5C), and proinflammatory cytokine and gene expression (Fig. 5D). Anakinra was effective in neutralizing hepatic IL-1β activity as anakinra administration significantly reduced IL-1β-dependent liver injury from GalN/LPS (Fig. S4B).

FIG. 5.

FIG. 5.

Mouse IFNγ hepatotoxicity is IL-1β independent. (A) Serum ALTs in control (Con) and Atg5Δhep knockout (KO) mice not injected (NI), or treated with normal saline vehicle (NS) or anakinra (AN) and IFNγ (n=5-8). (B) Numbers of TUNEL-positive cells per high power field (HPF) in the livers of the mice (n=5). (C) Relative hepatic Cd68 and Ly6g mRNA levels (n=5). (D) Relative mRNA levels for proinflammatory genes (n=5).

STAT1 is overactivated in female knockout mice

Cellular responses to IFNγ result from up regulation of gene expression by the transcription factors STAT1 and IRF1.(17) Hepatic Stat1 and Irf1 mRNA levels were induced by IFNγ to significantly higher levels in female knockout than control mice (Fig. 6A), whereas induction was equivalent in male control and knockout mouse livers (Fig. 6B). Nuclear and cytoplasmic levels of total and phosphorylated STAT1 were increased to higher levels in female knockout mice than control mice (Fig. 6C). In contrast IRF1 protein induction was equivalent in control and knockout female mice despite differences in mRNA content (Fig. 6C). In contrast, STAT1 as well as IRF1 protein levels were equivalent in control and knockout male mice (Fig. S5).

FIG. 6.

FIG. 6.

IFNγ-induced STAT1 overactivation occurs in female Atg5Δhep mice. (A) Relative Stat1 and Irf1 hepatic mRNA levels in female control (Con) and Atg5Δhep knockout (KO) mice not injected (NI) or injected with IFNγ for the indicated hours (*P<0.05, **P<0.01 compared to control mice not injected; #P<0.05, ##P<0.0002 compared to control mice with the same treatment; n=6). (B) mRNA levels in male mice (*P<0.01 compared to control mice not injected; n=6). (C) Immunoblots of liver nuclear and cytosolic protein fractions from control and Atg5Δhep mice treated with IFNγ for the indicated hours and probed with the antibodies shown. Lamin A/C and tubulin are nuclear and cytosolic protein loading/purity controls, respectively. Molecular weights in KD are indicted by arrows. Immunoblots are representative of four independent experiments. (D) Relative mRNA levels for STAT1-dependent genes in female mice (*P<0.05, **P<0.01 compared to control mice with no injection; #P<0.05, ##P<0.01 compared to control mice with the same treatment; n=6). (E) mRNA levels for the same genes in male mice (*P<0.05 compared to control mice with no injection; n=6).

STAT1-dependent gene expression confirmed transcriptional overactivation in autophagy-deficient female mice. Hepatic mRNA levels of the IFNγ target genes C-X-C motif chemokine ligand (Cxcl) 10 and 16, leukemia inhibitory factor (Lif) and matrix metallopeptidase 9 (Mmp9) were increased at 6 h after IFNγ treatment in female control mice, but female Atg5Δhep mice had significantly greater mRNA increases at 6 and 12 h for all four genes (Fig. 6D). Gene expression was not induced in male mice except for a modest Mmp9 increase at 6 h (Fig. 6E). Female autophagy-deficient mice have increased STAT1 signaling suggesting a possible mechanism for the sensitivity of female mice to IFNγ-induced liver injury and inflammation.

IFNγ hepatotoxicity is STAT1 dependent

The mechanistic involvement of STAT1 overactivation in IFNγ-induced hepatocyte cytotoxicity was tested initially in Vec and siAtg5 cells infected with a lentiviral Stat1 shRNA to generate control Vec-siStat1 cells and siAtg5-siStat1 cells (Fig. 7A). Knockdown of Stat1 in siAtg5 cells reduced death from IFNγ/IL-1β to the level in Vec cells by trypan blue (Fig. 7B,C) and propidium iodide staining (Fig. S6A,B). Death from IFNγ/TNF was similarly blocked (Fig. S6A,C,D,E). In contrast, a knockdown of Irf1 failed to block IFNγ-induced death (Fig. S7, S8).

FIG. 7.

FIG. 7.

Hepatocyte IFNγ death is STAT1 mediated. (A) Western blot images of Vec and siAtg5 AML12 cells untreated or treated with IFNγ/IL-1β for 2 h. Blots were probed for the proteins listed with the molecular weights in KD indicated by arrows. The ATG5 band represents the ATG5-ATG12 conjugate form. (B) Representative trypan blue-stained images of cells untreated (Con) or treated with IFNγ/IL-1β (I/I) for 24 h. (C) Quantification of the numbers of trypan blue-positive cells (*P<0.000001 compared to Vec control; #P<0.000001 compared to siAtg5 cells with the same treatment; n=8). (D) ALTs in control (Con) and Atg5Δhep knockout (KO) mice not injected (NI) or pretreated with DMSO vehicle or fludarabine (FLU) and IFNγ for 6 h (*P<0.000001 compared to Atg5Δhep mice DMSO/IFNγ treated; n=7-8). (E) Numbers of TUNEL-positive cells per high power field (HPF) in the livers of the same mice (*P<0.0002 compared to Atg5Δhep mice DMSO/IFNγ treated; n=7-8).

STAT1 overactivation was further examined as the mechanism of in vivo IFNγ toxicity in female mice. The STAT1 inhibitor fludarabine(18) significantly reduced levels of ALT and TUNEL staining after IFNγ administration (Fig. 7D,E and S9). Increased STAT1 signaling that occurs in female mice from decreased autophagy leads to IFNγ hepatotoxicity.

STAT1-dependent Nos2 overexpression mediates liver injury

STAT1 up regulates Nos2 expression,(19) and excessive NOS2 triggers cell death.(20) IFNγ induction of Nos2 was amplified in both autophagy-deficient AML12 cells (Fig. 8A) and female mice (Fig. 4D). NOS2 protein levels were also increased in IFNγ-treated siAtg5 cells (Fig. 8B) and female knockout mouse livers (Fig. 8C). The NOS inhibitors 1400W and L-NIL decreased IFNγ/IL-1β-induced siAtg5 cell death by over 90% by trypan blue (Fig. 8D,E) and propidium iodide staining (Fig. S10). NOS2 inhibition reversed the decrease in ATP in IFNγ/IL-1β-treated siAtg5 cells (Fig. 8F), indicating that STAT1-mediated NOS2 overexpression is the mechanism of ATP depletion and necrosis.

FIG. 8.

FIG. 8.

Increased NOS2 mediates death from IFNγ in vitro and in vivo. (A) Relative Nos2 mRNA levels in Vec and siAtg5 AML12 cells untreated or treated with IFNγ/IL-1β for the indicated hours (*P<0.01, **P<0.0001 compared to Vec cells with the same treatment; n=6). (B) Immunoblot images of total protein from the same cells probed for NOS2 and tubulin. (C) Immunoblots of total liver protein from female control (Con) and Atg5Δhep knockout (KO) mice treated with IFNγ for the hours shown. Molecular weights in KD are indicated by arrows. Immunoblots are representative of three independent experiments. (D) Representative trypan blue images of siAtg5 cells untreated (Con) or 24 h after the indicated treatments. (E) Numbers of trypan blue-positive siAtg5 cells untreated or treated with IFNγ/IL-1β after no pretreatment (Ø) or 1400W or L-NIL (*P< 0.0000001 compared to IFNγ/IL-1β; n=10). (F) ATP content in siAtg5 cells untreated (Con) or treated with 1400W and/or IFNγ/IL-1β(I/I)(*P<0.03, **P<0.000001 compared to Con; #P<0.0002 compared to IFNγ/IL-1β; n=6). (G) Serum ALTs at 6 h in control (Con) and Atg5Δhep knockout (KO) mice not injected (NI) or injected with PBS vehicle or 1400W and IFNγ (*P< 0.00001 compared to Atg5Δhep mice treated with IFNγ; n=8). (H) Numbers of TUNEL-positive cells per high power field (HPF) in the same livers (*P<0.001 compared to Atg5Δhep mice IFNγ treated; n=7-8).

The mechanistic role of NOS2 in hepatocyte death from IFNγ was confirmed in vivo. Consistent with findings in AML12 cells, 1400W significantly decreased serum ALT by 73% (Fig. 8G), and TUNEL positivity by 61% (Fig. 8I, S11) in IFNγ-treated female knockout mice. IFNγ death in the setting of decreased autophagy results from the increase in NOS2 due to STAT1 overactivation.

Discussion

Overactivation of the hepatic immune response underlies the development of many liver diseases in part by generating proinflammatory cytokines that trigger hepatocyte injury and death. Cytokine hepatotoxicity may result directly from initiation of a hepatocyte death pathway, or indirectly through further amplification of the inflammatory response to generate secondary injurious factors such reactive oxygen species. Investigations of cytokine hepatotoxicity have focused largely on TNF and IL-1β. IFNγ has not been well investigated despite its induction in many human liver diseases.(21) In addition to antiviral and immunoregulatory functions, IFNγ can be cytotoxic to hepatocytes.(1) IFNγ has been described to induce a delayed apoptosis in cultured hepatocytes,(22) and concanavalin A-induced apoptosis in mice results from IFNγ/TNF co-toxicity.(23) The present study demonstrates a novel form of IFNγ-induced necrosis in hepatocytes resulting from STAT1-mediated Nos2 overexpression in the setting of decreased autophagy (Fig. S12).

Recently we demonstrated that decreased autophagy sensitizes hepatocytes to death from TNF and IL-1β.(8,15) These studies together with the current findings for IFNγ establish autophagy as a central, critical hepatocyte resistance pathway against cytokine-induced hepatocyte injury and death. Although decreased autophagy sensitizes to death from all three cytokines, the type and mechanism of death are distinct for each cytokine. Autophagy-deficient hepatocytes are resistant to death from TNF alone but are sensitized to hepatotoxin-dependent apoptosis due to increased caspase 8 activation.(8) Hepatocyte necrosis from IL-1β occurs only in combination with TNF through the mechanism of lysosomal permeabilization.(15) In contrast to TNF and IL-1β, decreased autophagy is sufficient by itself to sensitize to hepatocyte death from IFNγ in mice, although AML12 hepatocytes require IL-1β or TNF co-treatment. Death from IFNγ/IL-1β is mechanistically distinct from that of IL-1β/TNF. IL-1β/TNF death resulted from a lack of autophagy-supplied substrates such as oleate required to maintain ATP levels after lysosomal death pathway activation.(15) Death from IFNγ/IL-1β was not blocked by oleate indicating that a deficiency of autophagy-generated energy substrates is not the mechanism of death. Death did result from reduced ATP as supplementation with the glycolytic pathway product pyruvate restored ATP content and prevented death. With IFNγ/IL-1β NOS2 overexpression induced ATP depletion and death as NOS2 inhibition blocked both events further demonstrating that the decrease in ATP was the direct effect of excessive NOS2 and not a loss of energy substrates normally generated by autophagy. Thus, despite the commonality of decreased ATP in death from IFNγ and IL-1β, distinct mechanisms of ATP depletion and necrosis are activated in hepatotoxicity from these two cytokines.

Death from IFNγ occurs from increased expression of its downstream transcriptional activator STAT1. Female mice exhibited increased Stat1 and Irf1 gene expression compared to male mice that was further amplified by decreased autophagy. IFNγ-induced death is STAT1 and not IRF1 dependent as: (1) IFNγ death in AML12 cells was prevented by a Stat1 but not an Irf1 knockdown; and (2) IFNγ hepatotoxicity in autophagy-deficient female mice was associated with selective overexpression of STAT1 and not IRF1 and prevented by STAT1 inhibition. These findings differ from a previous report of increased IRF1 in autophagy-deficient macrophages,(24) suggesting that the effects of decreased autophagy on IFNγ signaling may be cell type specific.

Our studies further identified NOS2 as the downstream effector of IFNγ-induced necrosis. Autophagy-deficient AML12 cells and female mouse livers had increased Nos2 gene expression and NOS2 protein levels after IFNγ stimulation. Death in both IFNγ/IL-1β-treated AML12 cells and IFNγ-injected mice with decreased autophagy was blocked by NOS2 inhibition. Excess nitric oxide can impair mitochondrial energy generation by several mechanisms,(20) and lead to hepatocyte ATP depletion and necrosis.(25) NOS2 overactivation is the upstream driver of the IFNγ necrotic pathway as NOS2 inhibition prevented ATP depletion as well as death. The mechanistic involvement of NOS2 differentiates necrotic IFNγ hepatotoxicity in the setting of decreased autophagy from previously reported IFNγ-induced apoptosis in hepatocytes or concanavalin A-treated mouse livers which are both NOS2 independent.(22,23) Our findings demonstrate that impaired autophagy sensitizes to death from three proinflammatory cytokines through distinct mechanisms, although it remains possible that an additional as yet undefined common function of autophagy regulates cytokine death as well.

The most striking distinction in cytokine death in the setting of defective autophagy is the gender selective effect for only IFNγ. Female mice had a modest amount of liver injury from IFNγ that was markedly amplified with impaired autophagy whereas male mice were completely resistant to IFNγ hepatotoxicity irrespective of autophagy levels. An increased female prevalence exists in a number of human liver diseases for unknown reasons.(26) Recently a gender-specific protective effect in acetaminophen hepatotoxicity was attributed to decreased expression of the mitochondrial membrane protein SAB.(27) This effect is unrelated to autophagy and the opposite of our findings as differences in this protein led to greater liver injury in males rather than females. No gender differences were detected in mouse liver SAB levels at baseline, with a decrease in autophagy or after IFNγ stimulation (data not shown). IFNγ hepatotoxicity occurred from overactivation of the STAT1-NOS2 pathway which only occurred in female mice.

The human diseases linked strongly to IFNγ are immune disorders.(17) Transgenic mice with chronic IFNγ overexpression develop liver disease similar to human primary biliary cholangitis.(2) Immune diseases ranging from the liver disease primary biliary cholangitis to the systemic disease lupus erythematosus exhibit a female predominance.(4) The existence of gender differences in immune responses including IFNγ have been reported and suggest the presence of estrogen and/or androgen regulation. Estrogen enhances IFNγ and IL-10 secretion from human female T-cell clones,(28) whereas androgen inhibits IFNγ secretion from murine T cells.(29) Long-term estrogen administration by an established protocol(30) failed to sensitize male knockout mice to IFNγ suggesting an estrogen-independent gender effect, however 17β-estradiol supplementation may have failed to fully replicate hormonal differences present in female mouse livers. Additional studies will be needed to examine the mechanism of gender specificity for hepatic IFNγ toxicity.

Immune diseases are marked by increased levels of IFNγ and the destruction of self-tissues, but how IFNγ promotes organ injury is unclear. Our findings demonstrate a new necrotic death pathway by which decreased autophagy may promote cytotoxicity from IFNγ. Decreased autophagy sensitizes to hepatic injury from IFNγ only in female mice. Levels of autophagy are decreased in many liver diseases,(810,12) therefore impairment of this pathway may explain the increased prevalence and severity of liver diseases such as autoimmune hepatitis and primary biliary cholangitis in women. IFNγ is a major regulator of immune diseases in all organs, and the ability of defective autophagy to alter the effects of IFNγ through altered downstream signaling may be important in the generalized female preponderance of immune diseases. Attempts to increase autophagy may be an effective therapeutic strategy in these diseases.

Supplementary Material

supinfo

Acknowledgements

We thank Norboru Mizushima for the Atg5F/F mice, Pierre Chambon for the ERt-Alb-Cre mice, Xiao-Ming Yin for the BID antibody, and Pradeep Kumar for his technical help.

Supported by NIH grant R01DK044234 (MJC).

Abbreviations:

ALT

alanine aminotransferase

CCL2

chemokine C-C motif chemokine ligand 2

GalN

D-galactosamine

GAPDH

glyceraldehyde 3-phosphate dehydrogenase

IFNγ

interferon-γ

IRF1

interferon regulatory factor 1

LC3

microtubule associated protein 1 light chain 3

LPS

lipopolysaccharide

NOS2

nitric oxide synthase 2

STAT1

signal transducer and activator of transcription 1

TUNEL

terminal deoxynucleotide transferase-mediated deoxyuridine triphosphate nick end-labeling

Footnotes

Conflict of Interest

The Authors have nothing to disclose.

REFERENCES

  • 1).Horras CJ, Lamb CL, Mitchell KA. Regulation of hepatocyte fate by interferon-γ. Cytokine Growth Factor Rev 2011;22:35–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2).Bae HR, Leung PS, Tsuneyama K, Valencia JC, Hodge DL, Kim S, et al. Chronic expression of interferon-gamma leads to murine autoimmune cholangitis with a female predominance. HEPATOLOGY 2016;64:1189–1201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3).Baechler EC, Batliwalla FM, Karypis G, Gaffney PM, Ortmann WA, Espe KJ, et al. Interferon-inducible gene expression signature in peripheral blood cells of patients with severe lupus. Proc Natl Acad Sci U S A 2003;100:2610–2615. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4).Ngo ST, Steyn FJ, McCombe PA. Gender differences in autoimmune disease. Front Neuroendocrinol 2014;35:347–369. [DOI] [PubMed] [Google Scholar]
  • 5).Brenner C, Galluzzi L, Kepp O, Kroemer G. Decoding cell death signals in liver inflammation. J Hepatol 2013;59:583–594. [DOI] [PubMed] [Google Scholar]
  • 6).Czaja MJ. Functions of autophagy in hepatic and pancreatic physiology and disease. Gastroenterology 2011;140:1895–1908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7).Czaja MJ, Ding WX, Donohue TM, Friedman SL, Kim JS, Komatsu M, et al. Functions of autophagy in normal and diseased liver. Autophagy 2013;9:1131–1158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8).Amir M, Zhao E, Fontana L, Rosenberg H, Tanaka K, Gao G, et al. Inhibition of hepatocyte autophagy increases tumor necrosis factor-dependent liver injury by promoting caspase-8 activation. Cell Death Differ 2013;20:878–887. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9).Singh R, Kaushik S, Wang Y, Xiang Y, Novak I, Komatsu M, et al. Autophagy regulates lipid metabolism. Nature 2009;458:1131–1135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10).Yang L, Li P, Fu S, Calay ES, Hotamisligil GS. Defective hepatic autophagy in obesity promotes ER stress and causes insulin resistance. Cell Metab 2010;11:467–478. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11).Cuervo AM, Bergamini E, Brunk UT, Droge W, Ffrench M, Terman A. Autophagy and aging: the importance of maintaining “clean” cells. Autophagy 2005;1:131–140. [DOI] [PubMed] [Google Scholar]
  • 12).Liu K, Zhao E, Ilyas G, Lalazar G, Lin Y, Haseeb M, et al. Impaired macrophage autophagy increases the immune response in obese mice by promoting proinflammatory macrophage polarization. Autophagy 2015;11:271–284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13).Lalazar G, Ilyas G, Malik SA, Liu K, Zhao E, Amir M, et al. Autophagy confers resistance to lipopolysaccharide-induced mouse hepatocyte injury. Am J Physiol Gastrointest Liver Physiol 2016;311:G377–G386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14).Ilyas G, Zhao E, Liu K, Lin Y, Tesfa L, Tanaka KE, et al. Macrophage autophagy limits acute toxic liver injury in mice through down regulation of interleukin-1β. J Hepatol 2016;64:118–127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15).Shen Y, Malik SA, Amir M, Kumar P, Cingolani F, Wen J, et al. Decreased hepatocyte autophagy leads to synergistic IL-1β and TNF mouse liver injury and inflammation. HEPATOLOGY 2020;72:595–608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16).Grasl-Kraupp B, Ruttkay-Nedecky B, Koudelka H, Bukowska K, Bursch W, Schulte-Hermann R. In situ detection of fragmented DNA (TUNEL assay) fails to discriminate among apoptosis, necrosis, and autolytic cell death: a cautionary note. HEPATOLOGY 1995;21:1465–1468. [DOI] [PubMed] [Google Scholar]
  • 17).Green DS, Young HA, Valencia JC. Current prospects of type II interferon γ signaling and autoimmunity. J Biol Chem 2017;292:13925–13933. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18).Frank DA, Mahajan S, Ritz J. Fludarabine-induced immunosuppression is associated with inhibition of STAT1 signaling. Nat Med 1999;5:444–447. [DOI] [PubMed] [Google Scholar]
  • 19).Jiang DS, Li L, Huang L, Gong J, Xia H, Liu X, et al. Interferon regulatory factor 1 is required for cardiac remodeling in response to pressure overload. Hypertension 2014;64:77–86. [DOI] [PubMed] [Google Scholar]
  • 20).Ghasemi M, Mayasi Y, Hannoun A, Eslami SM, Carandang R. Nitric oxide and mitochondrial function in neurological diseases. Neuroscience 2018;376:48–71. [DOI] [PubMed] [Google Scholar]
  • 21).Tilg H, Wilmer A, Vogel W, Herold M, Nolchen B, Judmaier G, et al. Serum levels of cytokines in chronic liver diseases. Gastroenterology 1992;103:264–274. [DOI] [PubMed] [Google Scholar]
  • 22).Kano A, Watanabe Y, Takeda N, Aizawa S, Akaike T. Analysis of IFN-γ-induced cell cycle arrest and cell death in hepatocytes. J Biochem 1997;121:677–683. [DOI] [PubMed] [Google Scholar]
  • 23).Kusters S, Gantner F, Kunstle G, Tiegs G. Interferon gamma plays a critical role in T cell-dependent liver injury in mice initiated by concanavalin A. Gastroenterology 1996;111:462–471. [DOI] [PubMed] [Google Scholar]
  • 24).Liang S, Zhong Z, Kim SY, Uchiyama R, Roh YS, Matsushita H, et al. Murine macrophage autophagy protects against alcohol-induced liver injury by degrading interferon regulatory factor 1 (IRF1) and removing damaged mitochondria. J Biol Chem 2019;294:12359–12369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25).Kim PK, Zuckerbraun BS, Otterbein LE, Vodovotz Y, Billiar TR. ’Til cell death do us part: nitric oxide and mechanisms of hepatotoxicity. Biol Chem 2004;385:11–15. [DOI] [PubMed] [Google Scholar]
  • 26).Durazzo M, Belci P, Collo A, Prandi V, Pistone E, Martorana M, et al. Gender specific medicine in liver diseases: a point of view. World J Gastroenterol 2014;20:2127–2135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27).Win S, Min RW, Chen CQ, Zhang J, Chen Y, Li M, et al. Expression of mitochondrial membrane-linked SAB determines severity of sex-dependent acute liver injury. J Clin Invest 2019;129:5278–5293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28).Gilmore W, Weiner LP, Correale J. Effect of estradiol on cytokine secretion by proteolipid protein-specific T cell clones isolated from multiple sclerosis patients and normal control subjects. J Immunol 1997;158:446–451. [PubMed] [Google Scholar]
  • 29).Araneo BA, Dowell T, Diegel M, Daynes RA. Dihydrotestosterone exerts a depressive influence on the production of interleukin-4 (IL-4), IL-5, and gamma-interferon, but not IL-2 by activated murine T cells. Blood 1991;78:688–699. [PubMed] [Google Scholar]
  • 30).Bhardwaj P, Du B, Zhou XK, Sue E, Giri D, Harbus MD, et al. Estrogen protects against obesity-induced mammary gland inflammation in mice. Cancer Prev Res (Phila) 2015;8:751–759. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supinfo

RESOURCES