ZBP1-mediated apoptosis and inflammation exacerbate steatotic liver ischemia/reperfusion injury

Abstract

Steatotic donor livers are becoming more and more common in liver transplantation. However, the current use of steatotic grafts is less acceptable than normal grafts due to their higher susceptibility to ischemia/reperfusion (I/R) injury. To investigate the mechanism underlying the susceptibility of steatotic liver to I/R injury, we detected cell death markers and inflammation in clinical donor livers and animal models. We found that caspase-8-mediated hepatic apoptosis is activated in steatotic liver I/R injury. However, ablation of caspase-8 only slightly mitigated steatotic liver I/R injury without affecting inflammation. We further demonstrated that RIPK1 kinase induces both caspase-8-mediated apoptosis and cell death-independent inflammation. Inhibition of RIPK1 kinase significantly protects against steatotic liver I/R injury by alleviating both hepatic apoptosis and inflammation. Additionally, we found that RIPK1 activation is induced by Z-DNA binding protein 1 (ZBP1) but not the canonical TNF-α pathway during steatotic liver I/R injury. Deletion of ZBP1 substantially decreases the steatotic liver I/R injury. Mechanistically, ZBP1 is amplified by palmitic acid-activated JNK pathway in steatotic livers. Upon I/R injury, excessive reactive oxygen species trigger ZBP1 activation by inducing its aggregation independent of the Z-nucleic acids sensing action in steatotic livers, leading to the kinase activation of RIPK1 and the subsequent aggravation of liver injury. Thus, ZBP1-mediated RIPK1-driven apoptosis and inflammation exacerbate steatotic liver I/R injury, which could be targeted to protect steatotic donor livers during transplantation.

Keywords: Apoptosis; Transplantation.

Figure 1. Apoptosis and inflammation are exacerbated in steatotic liver I/R injury.

(A) Serum ALT and AST levels of normal donor liver or steatotic donor liver recipients within 7 days after transplantation. (B and C) Prereperfusion and postreperfusion donor liver specimens were analyzed with H&E staining (B) and TUNEL staining (C). Arrows indicate congestion (B). n = 25 for normal donor livers and n = 24 for steatotic donor livers. (D) Recipients’ serum levels of the proinflammatory cytokines (n = 10) on POD1. (E) Left: Representative immunoblot result of cell death markers in hepatocytes isolated from pre- and postreperfusion donor livers. Protein expression was normalized to tubulin levels and shown as relative values. n = 25 for normal donor livers and n = 24 for steatotic donor livers. Right: PHH cell were isolated and cultured in vitro. After H/R challenge, the death markers were analyzed (n = 6). (F–H) ND-, HFD-, and CDHFD-fed mice (n = 8) underwent 1 hour ischemia/6 hours reperfusion operation and donor livers from ND- or CDHFD-fed rats (n = 6) underwent 18 hours cold storage/6 hours reperfusion. Serum ALT/AST levels (F and H) and cell death markers in isolated hepatocytes (G) were detected. All data are presented as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. 2-way ANOVA, post hoc Bonferroni’s test (A, C, F, and H). Fisher’s exact test (B). Unpaired 2-tailed Student’s t test (D and E). Scale bars: 100 μm

Figure 2. RIPK1 drives inflammation and caspase-8–mediated apoptosis, contributing to steatotic liver I/R injury.

(A and B) Hepatocytes were isolated from CDHFD-fed WT mice (n = 3) and exposed to 10 hours hypoxia after 1 hour pretreatment of z-VAD-fmk (zVAD, 10 μM), Nec-1s (10 μM), GSK′872 (1 μM), and disulfiram (50 μM). Cell death was detected after 24 hours reoxygenation (A) and cell death markers were detected after 3 hours reoxygenation (B). (C) PMH isolated from ND-, HFD-, and CDHFD-fed Ripk1^WT/WT or Ripk1^{D138N/ D138N} mice (n = 3) were exposed to 10 hours hypoxia. Cells were analyzed with immunoblot for p-RIPK1 (S166) after 3 hours reoxygenation. (D) Expression of activated RIPK1 of whole cell lysates in hepatocytes of prereperfusion or postreperfusion donor livers were detected with immunoblot (n = 6). (E–I) ND-, HFD-, CDHFD-fed Ripk1^WT/WT or Ripk1^D138N/D138N mice (n = 8) underwent 1 hour ischemia/6 hours reperfusion operation. Serum ALT/AST detection (E), H&E staining (F), TUNEL staining (G), cell death analysis (H), and detection of serum cytokine concentrations (I) were performed. (J and K) Livers of ND- or CDHFD-fed rats (n = 6) were perfused with and stored in 4°C University of Wisconsin solution supplemented with Nec-1s to a final concentration of 10 μM. After 18 hours cold storage/6 hours reperfusion, cell death analysis (J) and detection of serum cytokine concentrations (K) were performed. All data are presented as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. 1-way ANOVA, post hoc Dunnett’s test (A). 2-way ANOVA, post hoc Bonferroni’s test (E–G, I, and K). Scale bars: 200 μm.

Figure 3. RIPK1 is activated by ZBP1 in steatotic liver I/R injury.

(A–E) ND-, HFD-, CDHFD-fed WT, Ripk3^–/–, Trif^–/–, Zbp1^–/– mice (n = 8) underwent 1 h ischemia/6 h reperfusion operation. Cell death analysis (A), serum ALT/AST detection (B), H&E staining (C), and TUNEL staining (D) were performed. (E) Transcriptome profiling of the WT and Zbp1^–/–livers (n = 3) was conducted. Differentially expressed genes were filtered with the criteria P_adj < 0.05 and |log₂ FC| > 1. The genes downregulated in Zbp1^–/– livers were enriched for GO enrichment analysis. (F–H) After AAV8-mediated Zbp1 knockdown, ND- or CDHFD-fed rat livers (n = 6) underwent 18 h cold storage/6 h reperfusion. Cell death analysis (F), serum ALT/AST detection (G), H&E staining (H) were performed. (I and J) Interaction of ZBP1 and RIPK1 was analyzed in hepatocytes of I/R-challenged mouse livers (I) and postreperfusion donor livers (J). All data are presented as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. 2-way ANOVA, post hoc Bonferroni’s test. Scale bars: 200 μm.

Figure 4. ZBP1 is increased in steatotic livers, underlying its specificity in promoting steatotic liver I/R injury.

(A) ZBP1 protein levels were detected in normal and steatotic donor livers after reperfusion. ZBP1 protein levels relative to tubulin were compared between normal donor livers and steatotic donor livers. n = 25 for normal donor livers and n = 24 for steatotic donor livers. (B) The correlation between ZBP1 protein level and recipients’ serum ALT and AST levels in the POD1 was analyzed (n = 49). (C–H) After AAV8-mediated Zbp1 overexpression and Nec-1s pretreatment, ND-fed mice (n = 8) were subjected to I/R challenge. Serum ALT/AST detection (C), H&E staining (D), TUNEL staining (E), cell death analysis (F), and detection of serum cytokine concentrations (G) were performed. (H) Transcriptome profiling of the mouse livers was conducted. The expression of genes that were upregulated by ZBP1 overexpression and rescued by Nec-1s pretreatment were analyzed for GO enrichment. (I–K) PMH isolated from AAV-mediated Zbp1-overexpression normal mouse livers (n = 3) were pretreated with Nec-1s and exposed to H/R challenge. Cell death (I), apoptosis (J), mRNA levels of cytokines (K) and were analyzed. All data are presented as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Unpaired 2-tailed Student’s t test (A). Pearson’s correlation test (B). 2-way ANOVA, post hoc Bonferroni’s test (C–E, G, I, and K). Scale bars: 200 μm.

Figure 5. Palmitic acid upregulates hepatocyte ZBP1 expression level through JNK pathway to exacerbate I/R injury of steatotic livers.

(A and B) ZBP1 mRNA levels in donor livers (n = 6) or mouse I/R model (n = 4 for sham and n = 8 for I/R) were quantified. (C and D) mRNA and protein levels of ZBP1 after different concentration of PA in normal PMH for 48 hours (n = 5). (E and F) mRNA and protein levels of ZBP1 after stimulation of 0.4 mM PA and 10μM SP600125 in normal PHH for 24 hours (n = 5). (G and H) mRNA and protein levels of ZBP1 after stimulation of 0.4 mM PA for 24 hours in c-Jun knockdown normal PHH (n = 5). (I) ZBP1 promoter mutation schema and luciferase activity after c-Jun overexpression (n = 6). (J) Binding of c-Jun to ZBP1 promoter after stimulation of PA and SP600125 treatment or c-Jun knockdown in normal PHH (n = 6). (K–M) Zbp1^+/+ or Zbp1^–/– PMH (n = 3) were pretreated with 0.2 mM PA for 24 hours and subjected to 10 h hypoxia. Cell death (K) was analyzed after 24 hours reoxygenation and apoptosis (L) and mRNA levels of cytokines (M) was analyzed after 3 hours reoxygenation. All data are presented as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. 2-way ANOVA, post hoc Bonferroni’s test (A, B, I, J, and M). 1-way ANOVA, post hoc Dunnett’s test (C, E, and G). Unpaired, 2-tailed Student’s t test (K).

Figure 6. Z-NA sensing is not required for ZBP1-induced liver I/R injury.

(A–F) Different truncation mutants of Zbp1 (A) were packaged into AAV8 and injected to ND-fed mice and I/R operation was conducted 1 month later (n = 8). Cell death and ZBP1-RIPK1 interaction analysis (B), serum ALT/AST detection (C), H&E staining (D), TUNEL staining (E), and detection of serum levels of the proinflammatory cytokines (F) were performed. All data are presented as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. 1-way ANOVA, post hoc Dunnett’s test. Scale bars: 200 μm.

Figure 7. ROS triggers ZBP1 aggregation and activation in liver I/R injury.

(A) Flag immunostaining was performed in I/R-challenged ND-fed mouse livers overexpressing Zbp1-Flag (n = 8). (B) ZBP1 aggregation was detected with immunoblot under nonreducing or reducing conditions in I/R-challenged mouse livers or transplantation-challenged donor livers (n = 3). (C) hZBP1-GFP plasmids were transfected into HEK293T cells. After 1 mM H₂O₂ treatment for 4 hours or H/R challenge after pretreatment with 10 mM NAC, ZBP1 oligomers were detected from green fluorescence (n = 8). (D–H) After AAV8-mediated Zbp1-Flag overexpression in ND-fed mice and pretreatment with NAC, I/R operation was performed (n = 8). Flag immunostaining (D), ZBP1 aggregation and ZBP1-RIPK1 interaction (E), serum ALT/AST levels (F), H&E and TUNEL staining (G), and detection of serum levels of the proinflammatory cytokines (H) were detected. All data are presented as the mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Unpaired, 2-tailed Student’s t test (A and D). 2-way ANOVA, post hoc Bonferroni’s test (C, G, and H). Scale bars (A, C, and D): 10 μm, (G) 200 μm.