Suppression of Hepatocyte Ferroptosis via USP19-Mediated Deubiquitination of SLC7A11 in Ischemia-Free Liver Transplantation

Abstract

Ischemia-free liver transplantation (IFLT) is developed as a novel clinical approach to avoid ischemia-reperfusion injury (IRI). This study aims to identify the most distinguished programmed cell death pathway in grafts undergoing IFLT versus conventional liver transplantation (CLT) and to explore the underlying mechanism. Ferroptosis is the most distinct programmed cell death form between IFLT and CLT grafts. Among various cell death inhibitors, the ferroptosis inhibitor (Ferrostain-1) is the most effective one to prevent hepatocytes from damage induced by oxygen deprivation/reoxygenation (OGD/R). Hepatocyte ferroptosis is significantly alleviated in IFLT versus CLT grafts in both human beings and pigs. Ubiquitination enzyme screening identifies augmented amounts of ubiquitin-specific protease 19 (USP19) in IFLT versus CLT grafts. The upregulation of USP19 in the grafts is correlated with reduced pathological Suzuki's score, lower post-transplant peak liver enzyme level, and less early allograft dysfunction in liver transplant recipients. USP19 overexpression mitigates post-transplant liver injury in mice. Mechanistically, USP19 inhibits the degradation of solute carrier family 7 member 11 (SLC7A11) by removing its K63-linked ubiquitin chains. Notably, USP19 overexpression reduces ferroptosis and IRI in a SLC7A11-dependent manner in mice. Collectively, USP19-mediated suppression of hepatocyte ferroptosis via deubiquitinating SLC7A11 is a key mechanism by which IFLT abrogates graft IRI.

Keywords: ferroptosis; ischemia reperfusion injury; liver transplantation; ubiquitin specific protease 19.

Figure 1

IFLT reduces IRI and suppresses ferroptosis in clinical practice. A–C) IFLT reduces IRI. The peak serum AST (A) and ALT (B) within 7 days and TBIL (C) on day 7 post‐liver transplantation were compared between IFLT and CLT groups (IFLT, n = 16; CLT, n = 48). D) The differences in GSVA scores between the four programmed cell death pathways are shown. E) Viability of mouse hepatocytes following OGD/R and treatment with inhibitors of apoptosis (Z‐VAD), ferroptosis (Fer‐1), and necrosis (Nec‐1). Z‐VAD (10 µM), Fer‐1 (5 µM), or Nec‐1 (10 µm) were administered to the hepatocyte culture medium prior to subjecting the hepatocytes to 4 h of oxygen deprivation followed by 4 h of reoxygenation. Finally, cell viability was assessed. F) Ultrastructural analysis of liver biopsy from CLT and IFLT. m, mitochondrion. scale bar, 1 µm; × 8000 magnification. G,H), MDA (G), and SOD (H) were compared between CLT and IFLT liver biopsies (IFLT, n = 16; CLT, n = 48). I–K) The ferroptosis negative regulation axis of SLC7A11 (I)‐GSH (J)‐ GPX4 (K) were measured in both IFLT and CLT liver tissues (IFLT, n = 16; CLT, n = 48). L,M) The correlation analysis between ferroptosis suppression score and the peak serum AST (L) and ALT (M) within 7 days after liver transplantation in IFLT and CLT groups (IFLT, n = 16; CLT, n = 48). N–P) The peak serum AST and ALT were compared between the SLC7A11 (N), GSH (O), and GPX4 (P) high‐ and low‐expression groups (n = 48). The error bars represent the standard deviation of the indicated dataset. P values in (A–E), (G–K), and (N–P) were analyzed by unpaired two‐tailed Student's t‐test (Prism; GraphPad), and (L & M) were analyzed by linear regression algorithm.

Figure 2

IFLT reduces IRI and suppresses ferroptosis in pigs. A–C), IFLT attenuates IRI. Liver injury was compared between IFLT and CLT groups (n = 6 for each group) by AST, ALT levels (A) and pathological manifestations (B & C) at 6 h post‐implantation. Red arrows, infiltrating inflammatory cells; green arrows, swelling of hepatocytes with loose cytoplasm and light staining; blue arrows, balloon degeneration, cytoplasmic vacuolization; yellow arrows, visible hemorrhage within the field of view. scale bars, 50 µm, × 20 magnification. D) The mitochondrial ultrastructure was captured by TEM. m, mitochondrion. left panel: scale bars, 2 µm, × 4000 magnification; right panel: scale bars, 500 nm, × 15 000 magnification. E,F) Levels of lipid peroxidation in livers were measured by MDA and SOD assay. G,H) Representative images of the fluorescence assay of ROS in livers using the fluorescent probe C11‐BODIPY (green) and DHE (red), with nuclei stained with DAPI (blue). Scale bar, 50 µm.

Figure 3

Deubiquitinase USP19 suppresses ferroptosis of hepatocytes. A) The GSVA scores for each deubiquitinase family were analyzed in Erastin‐induced hepatocyte ferroptosis. B) Proteomic analysis showed the numbers of peptide segments of 5 deubiquitinases (USP7, USP19, USP24, USP39, and USP47) were significantly higher in IFLT than in CLT grafts. C) The comparison of cell viability among MIHA cells exposed to OGD/R with knockdown of USP7, USP19, USP24, USP39, and USP47. D,E) Cell survival assay of hepatocytes treated with Sorafenib and Fer‐1 was measured using CCK‐8. Primary hepatocytes of WT or Usp19‐KO mice were stimulated with Sorafenib at the indicated concentrations for 12 h (D) or 2.5 µMSorafenib for different time points (E). F,G) Cell survival assay in control or Usp19‐overexpressed MIHA cells under Sorafenib treatment in indicated doses for 18 h (F) or 10 µM Sorafenib for a series of time points with or without Fer‐1 (5 µM) co‐treatment (G). H) Analysis of WT or KO mice survival upon treatment with a half‐lethal dose of concanavalin A (ConA) (n = 10 mice in each group). I) Serum levels of AST and ALT were assessed in WT or KO mice treated with ConA (n = 6 mice in each group). J) H&E staining of liver sections from ConA‐treated mice. Red arrows, infiltrating inflammatory cells; black arrows, cell coagulation necrosis, nuclear pyknosis, and deep staining; green arrows, swelling of liver cells with loose cytoplasm and light staining. left panel: scale bars, 500 µm, × 40 magnification; right panel: scale bars, 100 nm, × 200 magnification. (K), Total GSH levels in livers of indicated groups. The error bars represent the standard deviation of the indicated dataset. P values in (B‐C), (J), and (K) were analyzed by unpaired two‐tailed Student's t‐test (Prism; GraphPad). Two or more groups were compared by ANOVA (D‐G). The survival analysis was described by Kaplan‐Meier Curve (Prism; GraphPad) (H).

Figure 4

USP19 expression is upregulated in IFLT and associated with ferroptosis and clinical IRI severity. A) The changes in USP19 mRNA expression between pre‐procurement and post‐reperfusion in IFLT (n = 16) and CLT (n = 48) cohorts. USP19 mRNA ratio was calculated by the USP19 mRNA expression post‐revascularization/at preprocurement. B) The protein levels of USP19 post‐revascularization were measured in IFLT and CLT recipients (n = 6 in each group). C–E) Correlation analyses of USP19 protein levels against the levels of SLC7A11 (C), GSH (D), and GPX4 (E) in CLT grafts (n = 48). These proteins were assessed by ELISA and normalized by the total amount of proteins. F) The correlation analysis between USP19 expression and the peak serum AST and ALT within 7 days after CLT (n = 48). G,H) Based on the median expression of USP19 in the grafts, recipients were divided into high‐expression and low‐expression groups. Suzuki's score (G) and EAD incidence of the two groups were compared. I) The distribution of L‐GrAFT risk groups was compared between the USP19 high‐ and low‐expression groups. P values in (A) were analyzed by unpaired two‐tailed Student's t‐test (Prism; GraphPad), The comparison in the indicated groups (H) was calculated by the Wilcoxon Rank Sum Test. P values in (H) and (I) were analyzed by chi‐square test. The correlation of USP19 and SLC7A11‐GSH‐GPX4 in (C‐E), and liver enzymes in (F) were analyzed by linear regression algorithm.

Figure 5

USP19 restrains IRI and inhibits ferroptosis. A) Serum AST and ALT levels in ADV‐Usp19‐treated and ADV‐NC‐treated mice at 6 h after OLT (n = 6). B) The levels of cytokines (IL‐1β, IL‐6, and TNF‐α) in ADV‐Usp19‐treated and ADV‐NC‐treated mice serum post‐transplantation. C,D) The representative H&E staining of liver sections (C) and the statistics of Suzuki's score for the pathological analysis (D) from ADV‐Usp19‐treated and ADV‐NC‐treated mice (n = 6). Red arrows, infiltrating inflammatory cells; black arrows, cell coagulation necrosis, nuclear pyknosis, and deep staining; blue arrows, balloon degeneration, cytoplasmic vacuolization. Scale bar, 50 µm, × 200 magnification. E) Representative hepatic TEM images of ADV‐Usp19‐treated group and ADV‐NC‐treated group. m, mitochondria. Top panel: scale bar, 2 µm, × 8000 magnification; bottom panel, scale bar, 500 nm, × 15 000 magnification. F,G) Levels of lipid peroxidation in livers were measured by MDA (F) and SOD (G) assay. H,I) The intracellular ROS content demonstrated by the fluorescence intensity of DHE (H) and BODIPY (I). (H), Scale bar, 50 µm, × 30 magnification. (I), Scale bar, 50 µm, × 40 magnification.

Figure 6

USP19 deficiency aggravates ferroptosis and cell OGD/R injury in vitro. A,B) Cell viability was evaluated by CCK8. Primary hepatocytes were isolated from WT and Usp19‐KO (A) or ADV‐Usp19‐treated and ADV‐NC‐treated (B) mice, cultured overnight and then subjected to OGD/R. Ferroptosis inhibitor (5 µm Fer‐1) was added to the culture for pretreatment prior to OGD/R. C–F) The contents of LDH (C), ATP (D), membrane potential (E) and GSH (F) were compared between the indicated groups. G–K) Cytoplasmic (G–I) and mitochondrial ROS (J,K) were detected and compared between WT and KO hepatocytes exposed to OGD/R. Cytoplasmic ROS of primary hepatocytes was detected with DCFH‐DA fluorescent probes (G,H) and DHE fluorescent probes (I). The DCF fluorescence intensity was detected by spectrophotometer (G) and flow cytometry (H) after being mounted on DCFH‐DA fluorescent probes. The intensity of red fluorescence after DHE staining also indicates ROS production (I). The bright red fluorescence was indicative of high intracellular oxidation in KO group. The primary mouse hepatocytes exposed to OGD/R were labeled with MitoSOX ™ red reagent which specifically targets mitochondria in living cells, and then the red fluorescence emitted by the reagent was detected by flow cytometry (J). The statistical analysis of the mitoSOX Red⁺ cells between groups was presented in the bar chart (K). Scale bar, 100 µm. P values were analyzed by unpaired two‐tailed Student's t‐test.

Figure 7

USP19 inhibits ferroptosis by deubiquitinating SLC7A11. A) The cell viability of MIHA cells treated with 10 µM Sorafenib for 0–18 h was compared between the control and USP19‐overexpression group, with or without SLC7A11 knockdown. B,C) Immunoblot analysis of SLC7A11 expression in MIHA cells with USP19 overexpression (B) or knockdown (C). D) Co‐Immunoprecipitation (Co‐IP) and immunoblot analysis of the interaction of USP19 and SLC7A11. MIHA cells were transfected with HA‐USP19 and Flag‐USP19, and then the cell extracts were harvested for analysis. E) Co‐Immunoprecipitation (Co‐IP) and immunoblot analysis of the interaction of USP19 and SLC7A11. The specific antibody against SLC7A11 (anti‐SLC7A11) was used to immunoprecipitate SLC7A11‐associated protein complexes revealing the endogenous association between SLC7A11 and USP19 in MIHA cells. F) MS analysis of proteins that interacted with USP19, including the SLC7A11 peptide. G) MIHA with or without transfected with Flag‐USP19 were treated with cycloheximide (CHX; 100 µg mL⁻¹) at the indicated time points. Cell lysates were then collected for immunoblotting (left). Quantification of the protein levels of SLC7A11 by ImageJ software (NIH) (right). H) Immunoblot analysis of SLC7A11 stabilization in MIHA cells transfected with HA‐USP19, followed by DMSO, MG132 (5 µm), or CQ (50 µm) treatment for 6 h. I) Immunoblot analysis of SLC7A11 expression in MIHA cells transfected with Myc‐tagged USP19 (WT, CS, CS/HA, or ΔTMD) plasmids. J) Co‐IP and immunoblot analysis of HEK293T cells transfected with vectors for the expression of Flag‐SLC7A11 and HA‐tagged ubiquitin (Ub) in the presence of Myc‐USP19 after SDS denaturation (input: 10%). K) Lysates of HEK293T cells transfected with plasmid expressing Flag‐SLC7A11 and HA‐tagged Ub (wild type, K6‐linked‐Ub, K11‐linked‐Ub, K27‐linked‐Ub, K29‐linked‐Ub, K33‐linked‐Ub, K48‐linked‐Ub, or K63‐linked‐Ub), together with the empty vector or expression vector of Myc‐USP19 and treated with BafA1 (0.3 µM) for 6 h, were immunoprecipitated with anti‐Flag and immunoblotted with anti‐HA. L) Immunoprecipitation and immunoblot analysis of HEK293T cells transfected with vectors expressing HA‐USP19 and Flag‐SLC7A11 WT or SLC7A11 mutants. M) Co‐IP and immunoblot analysis of HEK293T cells transfected with vectors for the expression of Flag‐SLC7A11 WT, or the indicated mutants and HA‐tagged Ub in the presence of Myc‐USP19 after SDS denaturation (input: 10%). For all immunoblot data, similar results were obtained from three independent biological experiments.

Figure 8

USP19 restrains IRI in a SLC7A11‐dependent manner. A–D) AAV9‐shSlc7a11 and ADV‐Usp19 were injected in the lateral tail vein of mice to knockdown Slc7a11 and overexpress Usp19, then mice were subjected to IRI. A,B) Serum AST (A), and ALT levels (B) were evaluated 6 h after surgery in the indicated groups. C,D) The histopathological changes, as depicted by the H&E staining images (C) and Suzuki's score (D), suggest a milder injury in the Usp19 overexpression group. Red arrows, infiltrating inflammatory cells; black arrows, cell coagulation necrosis, nuclear pyknosis, and deep staining; green arrows, swelling of hepatocytes with loose cytoplasm and light staining; blue arrows, balloon degeneration, cytoplasmic vacuolization. Scale bar, 100 µm, × 200 magnification. E–H) The ferroptosis were evaluated by LPO (E), MDA (F), SOD (G), and GSH (H). P values in (A & B) and (E‐H) were analyzed by unpaired two‐tailed Student's t‐test (Prism; GraphPad).