Extracellular gp96 is a crucial mediator for driving immune hyperactivation and liver damage

Abstract

Liver failure leads to the massive necrosis of hepatocytes, releasing large amounts of intracellular components including damage-associated molecular patterns (DAMPs). We found that extracellular gp96 levels in serum were elevated in patients with chronic hepatitis B infection (CHB) and acute-on-chronic liver failure (ACLF). Meanwhile, the gp96 level positively correlated with hepatic necroinflammation. We employed two mouse liver damage and liver failure models induced by lipopolysaccharide (LPS) plus D-galactosamine (D-Galn), and concanavalin A (ConA) to identify the function of extracellular gp96. As a result, the inhibition of extracellular gp96 by a specific peptide efficiently mitigated both LPS/D-Galn- and ConA-induced liver injury and immune hyperactivation, whereas exogenous gp96 aggravated the symptoms of hepatic injury in mice but not in Kupffer cells-ablated mice. The exposure of Kupffer cells to gp96 induced the secretion of pro-inflammatory cytokines. Collectively, our data demonstrate that gp96 released from necrotic hepatocytes aggravates immune hyperactivation and promotes liver damage and possibly the development of liver failure mainly by activating Kupffer cells.

Figure 1

Serum gp96 levels are elevated in patients with CHB and ACLF. (a) Serum ALT or (b) gp96 levels in patients with CHB and healthy controls were measured by ELISA. (c) Correlation analysis between serum ALT and gp96 levels in patients with CHB. Pearson’s correlation coefficient (R) and P-value were analysed. (d) Serum ALT or (e) gp96 levels in patients with ACLF induced by HBV infection, drug, or unknown factors were measured by ELISA. (f) Correlation analysis between serum ALT and serum gp96 levels in patients with ACLF. (g) Serum cytokines levels of TNF-α, IL-6, and IFN-γ in patients with ACLF were measured by ELISA. Correlation analysis between serum TNF-α (h) or IL-6 (i) and serum gp96 levels in patients with ACLF. Data are presented as mean ± SEM. *p < 0.05, **p < 0.01, and ***p < 0.001 compared to the control.

Figure 2

Serum gp96 levels are elevated in two mouse liver failure models. (a–e) Female C57 mice were challenged intraperitoneally (i.p.) with LPS (30 ng/g)/d-Galn (500 μg/g) or PBS as a control (n = 5/group). After 6 h, mice were sacrificed. Serum ALT levels (a) and liver histology (arrows indicate the necrotic areas) (b) were assessed. Serum gp96 (c) and pro-inflammatory cytokines including TNF-α, IL-6, and IL-1β (d) were quantified by ELISA. Intrahepatic immune cells were isolated, and then the total amounts of different cells were measured by flow cytometer (e). (f,g) Male BALB/c mice were challenged intravenously (i.v.) with ConA (15 μg/g) or PBS as a control (n = 5/group). After 8 h, mice were sacrificed. Serum gp96 (f) and pro-inflammatory cytokines including TNF-α, IL-6, and IFN-γ (g) levels were quantified by ELISA. Data are presented as mean ± SEM from two independent experiments.

Figure 3

Inhibition of gp96 by the specific peptide mitigates immune hyperactivation and liver injury in mice induced by LPS/d-Galn or ConA. (a–d) Female C57 mice were challenged i.p. with LPS (30 ng/g)/d-Galn (500 μg/g). After 1 h, 100 μg of the gp96 inhibitor or control peptide was injected i.p. (n = 5/group). At 6 h after LPS/d-Galn treatment, mice were sacrificed. Serum ALT levels (a) and liver histology (arrows indicate the necrotic areas) (b) were assessed. Serum inflammatory cytokine (TNF-α, IL-6, and IL-1β) levels were measured by ELISA (c). Intrahepatic immune cells were isolated and analysed by flow cytometer (d). (e–g) Male BALB/c mice were challenged i.v. with ConA (15 μg/g). After 1 h, 100 μg of the gp96 inhibitor or control peptides was injected i.p. At 8 h after ConA treatment, mice were sacrificed. Serum ALT levels (e), liver histology (f), and serum inflammatory cytokine levels (g) were determined. Data are presented as mean ± SEM from three independent experiments.

Figure 4

Exogenous gp96 promotes liver injury in mice after challenge with a low dose of LPS/d-Galn or ConA. (a) Female C57 mice were injected i.p. with 10 μg Cy5-labelled gp96 for different times as indicated before they were sacrificed. Photographs of livers were taken by IVIS spectrum. (b) Mice were injected i.p. with the indicated doses of gp96 respectively (n = 5/group) at 3 h after the challenge with a low dose of LPS (5 ng/g)/d-Galn (500 μg/g). The serum ALT levels was measured at 6 h after LPS/d-Galn treatment. (c–e) Female C57 mice were challenged i.p. with a low dose of LPS (5 ng/g)/d-Galn (500 μg/g). After 3 h, 2 μg gp96 or mouse serum albumin (MSA) were injected i.p. (n = 5/group). At 6 h after LPS/d-Galn treatment, mice were sacrificed. Serum ALT levels (c), liver histology (arrows indicate the necrotic areas) (d) was assessed. Serum inflammatory cytokine level were measured by ELISA (e). (f–h) Male BALB/c mice were challenged i.v. with a low dose of ConA (10 μg/g). After 5 h, 2 μg gp96 or MSA were injected i.p. (n = 5/group). At 8 h after ConA treatment, mice were sacrificed. Serum ALT levels (f), liver histology (arrows indicate the necrotic areas) (g) was assessed. Serum inflammatory cytokine level were measured by ELISA (h). Data are presented as mean ± SEM from two independent experiments.

Figure 5

Gp96 induces expression of inflammatory cytokines of intrahepatic immune cells in vitro. (a,c) Total intrahepatic immune cells (a) or Kupffer cells (c) were isolated from the C57 mice and incubated with different doses of gp96 with or without the gp96 inhibitor. At 24 h after treatment, secreted TNF-α in the culture supernatant was detected by ELISA. (b,d) Total intrahepatic immune cells or Kupffer cells were treated with 10 μg/ml gp96 with or without the gp96 inhibitor(2 μg/ml) for 12 h. Relative mRNA expression of multiple inflammatory cytokines as indicated of total intrahepatic immune cells (b) or Kupffer cells (d) was detected by quantitative real-time PCR. Data are presented as mean ± SEM from three independent experiments. Ns, not significant, *p < 0.05, **p < 0.01, and ***p < 0.001 compared to the control.

Figure 6

Exogenous gp96 loses its promotion effect on LPS/d-Galn- or ConA-induced liver injury in Kupffer cells-ablated mice. Mice were injected i.v. with 200 μl clodronate liposome for 48 h. (a–c) Female C57 mice were challenged i.p. with LPS (10 ng/g)/d-Galn (500 μg/g). After 3 h, 2 μg gp96 or MSA were injected i.p. (n = 5/group). At 6 h after LPS/d-Galn treatment, mice were sacrificed. Serum ALT levels (a), liver histology (arrows indicate the necrotic areas) (b), and serum inflammatory cytokine levels (c) were determined. (d–f) Male BALB/c mice were challenged i.v. with ConA (25 μg/g). After 5 h, 2 μg gp96 or MSA were injected i.p. (n = 5/group). At 8 h after ConA treatment, mice were sacrificed. Serum ALT levels (d), liver histology (e), and serum inflammatory cytokine levels (f) were assessed. Data are presented as mean ± SEM from two independent experiments.

Figure 7

Schematic illustration of how extracellular gp96 promotes immune hyperactivation and liver damage during the development of liver failure. Multiple stimuli, including different pathogens, chemicals, or drugs, induces liver damage and necrosis of hepatocytes, which results in the release of extracellular gp96. Extracellular gp96 may recognize and bind to its receptors (e.g. CD91 or TLR2/4, etc.) to stimulate the activation of Kupffer cells, which secrete large amounts of pro-inflammatory cytokines, resulting in the infiltration of immune cells into the liver and hepatic hyperactivation. This may in turn lead to further damage.