Opposing Effects of Prednisolone Treatment on T/NKT Cell- and Hepatotoxin-mediated Hepatitis in Mice

Mice

C57BL/6N mice were purchased from the NCI (Frederick, MD). Myeloid-specific glucocorticoid receptor knockout (LysCre⁺GR^flox/flox)(mGR KO) mice were kindly provided by Dr. Louis J. Muglia (University of Cincinnati).³² LysCre^−/− GR^flox/flox mice were used as littermate wild-type (WT) controls. All animal experiments were approved by the National Institute on Alcohol Abuse and Alcoholism Animal Care and Use Committee.

Mouse models

For the acute liver injury model, 8- to 10-week old male mice were injected with ConA (20 mg/kg body weight, i.v.)(Sigma-Aldrich, St Louis, MO), α-GalCer (100 mg/kg body weight, i.p.)(Enzo Life Science, Farmingdale, NY), or CCl₄ (0.2 ml/kg body weight, i.p.)(Sigma-Aldrich). Subsequently, prednisolone (10 mg/kg body weight in 5% ethanol, i.p; Sigma-Aldrich) or vehicle (5% ethanol) was administered to mice, as described in supporting Fig. 1. In some experiments, mice were injected i.p. with 50 mg/kg body weight 5-bromo-2′-deoxy-uridine (BrdU; Sigma-Aldrich) 2h before sacrifice for the examination of liver regeneration.

The chronic plus binge ethanol model was described previously.²⁹ For ethanol plus CCl₄-induced liver fibrosis model, mice were fed Liber-DeCarli liquid diet for 8 weeks, and CCl₄ (0.1 mg/kg) were administrated twice a week during the 8-week feeding. Prednisolone was injected i.p. daily for the last two weeks. For acetaminophen (APAP)-induced liver injury, mice were fasting for 24h and subsequently injected with APAP (200 mg/kg, i.p), and prednisolone or vehicle was administrated as described in the figure legends.

Other materials and methods are included in supporting document.

Co-treatment with prednisolone ameliorates acute T/NKT cell hepatitis induced by ConA or α-GalCer

To investigate the role of prednisolone in experimental T cell hepatitis, mice were co-treated with ConA plus prednisolone or vehicle (supporting Fig. 1A). Treatment of mice with prednisolone alone did not induce elevation of serum ALT and no obvious liver necrosis was observed in mice after prednisolone treatment alone (supporting Fig. 2). ConA injection induced an elevation of serum ALT levels and massive necrosis in the liver, which were markedly decreased in mice co-treated with prednisolone (supporting Figs. 3A-B). Prednisolone treatment also markedly prevented ConA-induced elevation of serum levels of various cytokines (supporting Fig. 3C). Similarly, prednisolone treatment prevented the NKT activator α-GalCer-induced systemic inflammatory responses, liver inflammation, and injury (supporting Fig. 4).

Co-treatment with prednisolone delays liver repair in CCl₄-induced acute hepatitis

CCl₄ is a hepatotoxic agent that has been widely used to induce liver injury in a hepatotoxin-mediated hepatitis model that allows for the evaluation of both necrosis and subsequent inflammation.²⁷ To determine the role of prednisolone in toxin-induced acute liver injury, mice were injected with CCl₄ and co-treated with prednisolone or vehicle (supporting Fig. 1C). Liver histology analyses in Figs. 1A-B shows that in CCl₄+vehicle group, injection of a single dose of CCl₄ induced massive liver necrosis 24, 48, and 72h post injection, but such necrosis was markedly recovered 96h post injection; surprisingly, in CCl₄+prednisolone group, massive liver necrosis was not improved 96h post injection and still remained until 120h. In agreement with liver histology data, prednisolone co-treatment induced higher levels of serum ALT levels 48h post CCl₄ injection (Fig. 1C). Treatment with prednisolone alone did not induce obvious liver necrosis (Fig. 1B) nor affected serum ALT levels (Fig. 1C). Moreover, the serum levels of several pro-inflammatory cytokines (e.g., TNF-α and IL-6), which are important for liver regeneration, ^{33, 34} were elevated after CCl₄ injection with vehicle co-treatment. These effects were completely attenuated in prednisolone co-treated mice (Fig. 1D). Finally, treatment of mice with prednisolone did not affect hepatic expression of p450CYP2E1, a key enzyme to metabolize CCl₄ (supporting Fig. 5). This suggests that prednisolone exacerbation of CCl₄-induced liver injury is not mediated via the regulation of CCl₄ metabolism.

In addition, we also examined the effects of prednisolone treatment on APAP–induced liver injury (supporting Fig. 6). Liver histology analyses shows a single injection of APAP induced massive liver necrosis 48h post injection, but such necrosis was still remained until 72h in APAP+prednisolone group (supporting Fig. 6C). In addition, prednisolone co-treatment markedly attenuated BrdU incorporation in hepatocytes 48h after APAP injection (supporting Figs. 6C-D).

Co-treatment with prednisolone exacerbates ethanol or ethanol plus CCl₄-induced chronic liver injury and fibrosis

To investigate the effect of prednisolone on alcohol-induced liver injury, mice were fed an ethanol diet for 10 days, followed by a single gavage of ethanol plus prednisolone or vehicle treatment (supporting Fig. 7A). Chronic-binge ethanol feeding elevated serum ALT levels, which were markedly increased in mice co-treated with prednisolone (supporting Fig. 7B). In addition, H&E and Oil Red O staining revealed higher degree of steatosis in mice co-treated with prednisolone than in mice co-treated with vehicle (supporting Fig. 7C).

To further investigate the effect of prednisolone in chronic liver injury and fibrosis, mice were treated with ethanol plus CCl₄ and co-injected with prednisolone or vehicle (supporting Fig. 8A). As illustrated in supporting Fig. 8B, prednisolone co-treatment induced higher levels of serum ALT levels, greater levels of Sirius red and α-SMA staining, and higher hepatic expression levels of α-SMA compared with vehicle co-treated groups (supporting Figs. 8E-F).

Prednisolone administration inhibits acute CCl₄-induced macrophage and neutrophil recruitment in the liver

CCl₄-induced liver injury is associated with activation of Kupffer cells and trigger migration of macrophages into hepatic cords. As shown in Fig. 2A, administration of CCl₄ plus vehicle co-treatment induced accumulation of mononuclear cells (MNCs) and Gr-1^intCD11b⁺ macrophages in the liver 24h post treatment, but such accumulation was attenuated in prednisolone co-treated mice. This reduction of macrophages infiltration in prednisolone-treated mice was not due to the increased macrophage apoptosis because there were no differences in macrophage apoptosis between prednisolone co-treated and vehicle co-treated groups (supporting Fig. 9).

Neutrophil infiltration was determined by immunohistochemical analyses of MPO (a marker for neutrophils). As illustrated in Fig. 2B and supporting Fig. 10, CCl₄ injection resulted in neutrophil infiltration in the liver. The number of MPO⁺ cells was significantly increased 24 and 48h post CCl₄ injection and declined thereafter in the CCl₄+vehicle group. Prednisolone co-treatment delayed hepatic neutrophil recruitment, with a lower number of MPO⁺ cells at 24 and 48h but a higher number at 72h post CCl₄ administration compared with the vehicle co-treated group. Furthermore, FACS analyses also confirmed that prednisolone co-treatment attenuated hepatic neutrophil infiltration post CCl₄ injection. As illustrated in Fig. 2C, in the CCl₄+vehicle group, the total number of Gr-1^hiCD11b⁺ neutrophils was elevated, with a peak effect occurring 24h and then declining 72h post injection. Prednisolone co-treatment resulted in a lower number of neutrophils 24h but a slightly higher number 72h post CCl₄ injection compared with the vehicle co-treated mice. Such prednisolone-mediated inhibition of hepatic neutrophil number was not due to promotion of neutrophil apoptosis because FACS analyses showed that prednisolone co-treatment delayed liver neutrophil apoptosis (supporting Fig. 11).

Finally, the hepatic expression of several cytokines and chemokines, which are associated with macrophage and neutrophil recruitment,³⁵ was suppressed in the prednisolone co-treated mice compared with those in the vehicle co-treated groups (Fig. 2D).

Co-administration with prednisolone attenuates macrophage and neutrophil functions

Macrophages and neutrophils are known to be professional phagocytes that play a significant role in the clearance of apoptotic cells and in the resolution of inflammation. To analyze the effect of prednisolone on macrophage and neutrophil phagocytic activity, mice were injected with latex beads post CCl₄ plus prednisolone or vehicle co-treatment. As illustrated in Figs. 3A-B, CCl₄ injection increased the percentage of macrophage or neutrophil engulfing latex beads, which was significantly reduced in the prednisolone co-treated group. Moreover, an in vitro phagocytosis assay (Figs. 3C-D) revealed that prednisolone exposure decreased the fluorescence intensity of latex beads in peritoneal macrophages and neutrophils.

Kupffer cells, liver resident macrophages stimulate tissue damage and repair by secreting cytokines. Therefore, we examined whether prednisolone treatment altered cytokine production by Kupffer cells as well as hepatic stellate cells (HSCs) and hepatocytes. As illustrated in supporting Fig. 12, Kupffer cells produced highest levels of TNF-α and IL-6 after CCl₄ treatment, followed by HSCs and hepatocytes. Such expression in Kupffer cells from CCl₄ plus prednisolone co-treated mice was much lower than those from CCl₄ plus vehicle co-treated mice. Expression of TNF-α but not IL-6 was also attenuated in HSCs from the CCl₄ plus prednisolone group compared with the CCl₄ plus vehicle.

Finally, the effects of prednisolone on neutrophil-mediated ROS production were also examined. As illustrated in Fig. 3E, without in vitro treatment with PMA, neutrophils from CCl₄-treated mice produced a slightly higher ROS burst compared with those from mice without CCl₄ treatment. This ROS burst production was suppressed in prednisolone co-treated mice compared with vehicle co-treated mice 24h but not 72h post CCl₄ injection. In vitro incubation with PMA, which induces cells to undergo a NOX–dependent respiratory burst, markedly increased the ROS levels of neutrophils (Fig. 3E). This PMA-mediated elevation of ROS production was lower in neutrophils from prednisolone co-treated mice compared with those from vehicle co-treated mice 24h post CCl₄ injection (Fig. 3E).

Prednisolone treatment delays liver regeneration by inhibiting hepatic STAT3 and pNF-κB activation in CCl₄-induced acute hepatitis

The effect of prednisolone on liver regeneration was examined to further understand why prednisolone treatment delayed liver repair post CCl₄ injection as observed in Fig. 1. As illustrated in Fig. 4A, CCl₄ challenge markedly increased BrdU incorporation in hepatocytes, with a peak effect 48h post challenge. However, prednisolone co-treatment delayed this peak to 72h.

We next investigated the mechanisms underlying the prednisolone-mediated interruption of liver regeneration in CCl₄-induced acute hepatitis by examining the hepatic expression of pSTAT3, pNF-κB, and proliferative genes. As shown in Fig. 4B, hepatic STAT3 was activated, with a peak effect occurring at 3 to 12h after CCl₄ injection in the CCl₄+vehicle group. The hepatic expression levels of pSTAT3 were lower at 6 and 12h but higher at 24 to 72h after CCl₄ injection in the CCl₄+prednisolone group, which suggests that prednisolone treatment caused a delay in hepatic STAT3 activation. The hepatic expression of pNF-κB was also decreased at 3 and 6h in CCl₄+prednisolone group compared with the CCl₄+vehicle group (Fig. 4C). In addition, prednisolone treatment slightly reduced NF-κB acetylation at 6h post CCl₄ injection, but it did not affect STAT3 acetylation (supporting Fig. 13). The induction of PCNA and cyclin D1 expression was delayed in the prednisolone co-treated mice compared with the vehicle co-treated mice (Fig. 4B).

Neutrophil and/or macrophage depletion aggravates CCl₄–induced acute liver injury and impedes liver regeneration

The above findings suggest that prednisolone inhibits neutrophil and macrophage recruitment and functions; however, the roles of these inflammatory cells in CCl₄-induced liver injury and regeneration remain largely unclear. To understand the functions of macrophages and neutrophils, we used anti-Gr1Ab antibody to deplete these cells prior to CCl₄ challenge. As illustrated in supporting Fig. 14A, pretreatment with anti-Gr1Ab markedly reduced both macrophages (Gr1^lowCD11b⁺) and neutrophils (Gr1^hiCD11b⁺) in the liver. The depletion of both macrophages and neutrophils with anti-Gr-1 Ab induced greater elevation of serum ALT and liver injury (supporting Fig. 14B), increased the area of necrosis (supporting Fig. 14C), and delayed liver regeneration as determined by BrdU incorporation (supporting Fig. 14D).

To further dissect the roles of macrophages and neutrophils, we used anti-Ly6G and clodronate to deplete specifically neutrophils and macrophages, respectively. As illustrated in Fig. 5A, treatment with anti-Ly6G depleted neutrophils (Gr1^hiCD11b⁺) without affecting macrophages (Gr1^lowCD11b⁺) in the liver. Such treatment increased CCl₄-induced elevation of serum ALT and liver necrosis, and delayed liver regeneration (Figs. 5B-D). Additionally, treatment with clodronate depleted F4/80 positive macrophages in the liver (Fig. 6A) and also exacerbated CCl₄-induced elevation of serum ALT and liver necrosis but attenuated liver regeneration (Figs. 6B-D).

Treatment with prednisolone does not affect CCl₄-induced liver injury and regeneration in mGR KO mice

To further explore whether the detrimental effects of prednisolone on CCl₄-induced hepatitis are mediated by targeting macrophages and/or neutrophils, we used myeloid-specific GR knockout (mGR KO) mice, in which the GR gene was deleted in macrophages and neutrophils.³² Western blot analyses confirmed a loss of GR protein expression in both macrophages and neutrophils from mGR KO mice (Fig. 7A). Moreover, the effects of prednisolone on CCl₄–induced acute hepatitis were compared in WT and mGR KO mice. As illustrated in Fig. 7B, CCl₄ treatment resulted in an elevation of serum ALT levels in WT mice, and such elevation was higher in prednisolone co-treated WT mice 48h post CCl₄ injection. However, prednisolone co-treatment did not affect CCl₄-induced elevation of serum ALT levels in mGR KO mice. Liver histology analyses in Fig. 7C show that prednisolone co-treatment increased CCl₄-induced necrotic area in WT mice but did not affect this in mGR KO mice. Immunohistochemical analyses in Fig. 7D show that prednisolone co-treatment delayed neutrophil infiltration in CCl₄-treated WT mice (the peak of neutrophil infiltration occurred 48h post CCl₄ injection alone but occurred 72h post CCl₄ plus prednisolone co-treatment). In contrast, prednisolone co-treatment did not have an effect on neutrophil infiltration in CCl₄-treated mGR KO mice. Similarly, prednisolone co-treatment delayed liver regeneration in CCl₄-treated WT mice but not in mGR KO mice (Fig. 7E).