We retrospectively analyzed forty-one adult patients (≥18 years) who underwent OLT (May 2013–August 2015). As specified by UCLA protocols, all the recipients received routine standard of care and immunosuppressive therapy. Those who underwent re-transplantation were excluded from the study. Donor livers, procured from donation after a brain or cardiac death with standardized techniques, were perfused with and stored in UW solution (Niaspan; Bristol-Meyers Squibb Pharma, Princeton, NJ). Protocol Tru-Cut needle biopsies (Bx) from the left liver lobe were obtained pretransplant (after liver cold storage on the back table) and posttransplant (at about 2 h after portal reperfusion and before the abdominal closure). Liver Bx samples were analyzed by Western blots (WB) and qRT-PCR. Cold ischemia time was defined as the time from the perfusion of the donor liver with UW solution to its removal from the cold storage for implantation. Warm ischemia time was defined as the time from cold storage removal to the establishment of liver graft reperfusion. Recipient blood was collected before and after OLT, and sALT/sAST levels evaluated the hepatocellular function. Early allograft dysfunction (EAD) was defined by the presence of one or more of the following: bilirubin level of ≥10 mg/dl on POD (postoperative day) 7, Prothrombin time (PT)-International Normalized Ratio (INR) ≥ 1.6 on POD7, or AST/ALT level of >2,000 U/L within the first 7 days.

Wild-type (WT; Jackson Laboratory, Bar Harbor, ME) and Tim4-deficient male mice (Tim4-KO) (12), both at C57BL/6 background and 6–8 weeks of age, were used. Animals were housed in UCLA animal facility under specific pathogen-free conditions, and received humane care according to the criteria outlined in the “Guide for the Care and Use of Laboratory Animals” (NIH publication 86–23 revised 1985), while their use was reviewed and approved by UCLA Animal Research Committee.

We used a mouse model of ex-vivo hepatic cold storage followed by orthotopic liver transplantation (OLT), as described by our group (13). In brief, after the recipient liver was removed, the donor liver was placed orthotopically. The anastomosis of the suprahepatic vena cava was performed by running 10–0 nylon. The portal vein and the infrahepatic vena cava were both reconnected through a cuff technique. The bile duct reconstruction was completed with an intraluminal stent. The recipient anhepatic time was approximately 14 min. To mimic “marginal” human OLT setting, donor livers (WT or Tim4-KO) stored in UW solution (4°C/18 h) were transplanted to syngeneic recipient mice (WT or Tim4-KO). We used a syngeneic model to focus on putative homeostatic Tim4 functions while avoiding confounding effects of host allo-immune MHC responses. Liver tissue and serum samples were collected at 6 h post-reperfusion, the peak of hepatocellular damage in this model (14) and at 24 h. Separate OLT recipient groups were monitored for survival. The sham group underwent the same procedures without OLT.

Serum AST/ALT levels were measured with Infinity™ AST/ALT Liquid Stable Reagent (Thermo Scientific, Rockford, IL) and validated with Validate®GC3 (Maine Standards Company, LLC, ME).

Formalin-fixed paraffin-embedded liver sections (5 µm) were stained with hematoxylin and eosin (H&E). The severity of IRI was graded using Suzuki's criteria (15). Briefly, the degree of cytoplasmic vacuolization, sinusoidal congestion and parenchymal necrosis were scored from 0 to 4 (0: none, 1: minimal, 2: mild, 3: moderate, 4: severe).

Cell death in liver sections (5 µm) was detected by In Situ Apoptosis Detection Kit (#MK500, TAKARA BIO USA) according to the manufacturer protocol. Results were scored semi-quantitatively by blindly counting the number of positive cells in 10 HPF (high-power field)/section.

Primary mouse hepatocytes were isolated using a two-stage collagenase perfusion method (14). Briefly, mouse liver was perfused with collagenase solution. After perfusion, liver was minced and dispersed in Geys balance salt solution. The cell suspension was centrifuged at 50g for 1 min to collect hepatocytes. The hepatocytes were cultured on type 1 collagen coated plate. In some experiments, hepatocytes were stimulated with TNF-α (#575204, Biolegend, San Diego, CA) for 6 h.

Adipose tissue (AT) conditioned media was prepared, as reported (16). Briefly, epididymal fat pads collected from WT or Tim4-KO mice were cut into 2–3 mm³ pieces. For relaxation, those fragments were incubated in Dulbecco's Modified Eagle Medium (DMEM) for 24 h, then washed with PBS and re-incubated in DMEM for 24 h. After re-incubation, culture media was filter-sterilized and collected as adipose tissue conditioned media (AT-CM) for hepatocyte culture.

RNA extracted with NucleoSpin® RNA (#740955.50, TAKARA BIO USA, Inc, Mountain View, CA) was reverse-transcribed into cDNA with the PrimeScript RT Reagent Kit (#RR037, TAKARA BIO). Quantitative PCR was performed using QuantStudio 3 (Applied Biosystems, Foster City, CA). The primer sequences are listed in Supplementary Table S1. The expression of the target gene was normalized to the housekeeping Hprt (mouse) or GAPDH (human).

Proteins were extracted from tissue/cell samples, and their concentration was measured using BCA Protein Assay Kit (#23227, Thermo Fisher Scientific, Waltham, MA). An equal amount of protein was electrophoresed, blotted, incubated with primary Ab, and secondary HRP-conjugated Ab, and developed. Primary Abs used are listed in Supplementary Table S2. To compare target protein expression in multiple human OLT samples, densitometry quantification was conducted, as reported (14, 17).

Serum concentration of IL-10 was measured by ELISA kit (#431414, Biolegend, San Diego, CA) according to the manufacturer protocol.

Serum free fatty acid levels were measured by Free Fatty Acid Fluorometric Assay Kit (#700310, Cayman Chemical, Ann Arbor, MI) according to the manufacturer protocol.

Paraffin-embedded mouse liver sections were stained with rabbit anti-CD36 Ab, mouse anti-CHOP Ab or mouse anti-PPARγ antibodies with M.O.M.® (Mouse on Mouse) ImmPRESS® HRP (Peroxidase) Polymer Kit (MP-2400, Vector Laboratories, Inc. Newark, CA) according to the manufacturer's protocol. The signal was visualized by diaminobenzidine tetrahydrochloride (DAB, D5905, MilliporeSigma, Burlington, MA), and the nucleus was counterstained with hematoxylin.

Hepatocytes isolated from WT mice were seeded into the chamber slide (#177445, Thermo Fisher Scientific, Waltham, MA) and fixed with 4% paraformaldehyde for 10 min. Cells were stained with primary Abs, and Alexa-conjugated secondary Abs were used to visualize the signal. The antibodies used in the study are listed in Supplementary Table S2.

For mouse experiments, the comparison between the two groups were assessed using the Mann-Whitney U test or 1 way-ANOVA followed by Tukey's HSD test. For human data, continuous values were analyzed by the Mann-Whitney U test or 2 way-ANOVA followed by Bonferroni post hoc test and categorical variables by Fisher's exact test. Cumulative survival rates were estimated using the Kaplan-Meier method, and survival curves were analyzed using log-rank tests. All P values were 2-tailed, and P less than 0.05 was considered statistically significant.

All human studies were approved by the UCLA Institutional Research Board (IRB protocol 13-000143), and written informed consent was received from participants before inclusion in the study. All mouse experiments were approved by the UCLA Animal Research Committee (ARC #1999-094).

Livers from Tim4-deficient mice (Tim4-KO; C57BL6), stored for 18 h at 4C, were transplanted into groups of syngeneic WT recipients, followed by sampling at 6 h and 1 day after OLT. Histological analyses showed hepatic Tim4 deficiency decreased sinusoidal congestion, edema, vacuolization, and hepatocellular necrosis in IR-stressed OLT (Figure 1A). These findings translated to improved Suzuki's histological score of IR-damage (Figure 1B), attenuated sAST and sALT release (Figure 1C), suppressed frequency of TUNEL + cells (Figure 1D), depressed hepatic mRNA levels coding for Il-1β, Mcp1, Cxcl1, Cxcl2 and Cxcl10 (Figure 1E), and prolonged OLT survival (Figure 1F). These results confirm Tim4 expressed primarily by hepatic KCs functions as a proinflammatory sentinel in the mechanism of IR-triggered OLT damage.

To focus on the role of recipient-derived Tim4 signaling, IR-stressed WT donor livers were implanted into groups of Tim4-deficient or Tim4-proficient (WT) mice. Unexpectedly, recipient Tim4 deficiency markedly deteriorated OLT function and outcomes, evidenced by increased sinusoidal congestion, edema, vacuolization, and hepatocellular necrosis (Figure 2A), higher Suzuki's histological score of hepatic IRI (Figure 2B), augmented sAST/sALT release (Figure 2C), increased frequency of TUNEL + cells (Figure 2D) and enhanced hepatic mRNA levels coding for Il-1β, Mcp1, Cxcl1, Cxcl2 and Cxcl10 (Figure 2E). Strikingly, Tim4-KO recipients died within 2 days post-OLT, while all WT recipients (WT > WT) showed considerably longer survival (Figure 2F). The detailed mouse OLT data points are shown (Supplementary Table S3).

Having found opposite phenotypes in IR-stressed mouse OLT and based on previous reports (4), we hypothesized that disruption of recipient Tim4 signaling disturbs lipid metabolic pathways that normally suppress the proinflammatory response to IR stress and attenuates OLT damage (18). Indeed, OLT expression of CD36 (also known as fatty acid translocase; FAT), a key mediator of cell membrane free fatty acid (FFA) and oxidized low-density lipoprotein (ox-LDL) synthesis was significantly upregulated in Tim4-null recipients (Figure 3A). Other remarkable differences occurred for sterol regulatory element-binding protein 1 (SREBP1) and peroxisome proliferator-activated receptor gamma (PPARγ), both of which are involved in lipogenesis (Figures 3A,B). Importantly, serum FFA levels after OLT were significantly higher in Tim4-KO as compared with WT counterparts (Figure 3C), while recipient Tim4-null mutation enhanced hepatic lipid peroxidation, evidenced by increased malondialdehyde (MDA) levels (Figure 3A). These findings suggest that recipient Tim4 signaling is indispensable for maintaining metabolic homeostasis to prevent excessive lipid accumulation in IR-stressed OLT.

Since previous reports document that SREBP1 and ER stress coordinately enhance lipotoxicity (19), we confirmed that both ATF4 and CHOP ER stress markers were significantly upregulated in Tim4-KO mice (Figure 3D). Immunohistochemistry of IR-stressed OLT revealed hepatocytes as the primary source of CHOP expression (Figure 3E). In addition, Tim4-null mice exhibited augmented cell death receptors (e.g., TNFR1 or DR5) and pro-apoptotic molecules, such as active PARP [Poly (ADP-ribose) polymerase], caspase 8 and caspase 3 (Figure 3F).

To further elucidate the underlying mechanism of Tim4-dependent regulation of lipid metabolism, we co-cultured WT mouse hepatocytes with adipose tissue-conditioned media (AT-CM) from WT vs. Tim4-deficient mouse donors (Figure 4A). As shown in Figure 4B, hepatocytes co-cultured with AT-CM from Tim4-null mice showed enhanced SREBP1, caspase 8, and CHOP expression compared to AT-CM from WT counterparts. Figure 4C illustrates representative immunocytochemistry of CHOP in cultured hepatocytes after AT-CM co-culture. Next, we asked whether AT-CM might confer the susceptibility to hepatocellular death. Indeed, as shown in Figure 4D, hepatocytes primed with AT-CM from Tim4-KO mice were more sensitive to TNF-α mediated hepatocellular death than from WT counterparts. This was evidenced by increased cleaved PARP, caspase 8, and cleaved caspase 3 expression. This data illustrates AT-expressing Tim4 regulates the hepatocellular metabolic program, ER stress, and apoptotic pathway.

Having demonstrated the functional role of Tim4 expression in mouse IRI-OLT, we next aimed to validate its relevance by screening retrospectively forty-one human OLT cases. Liver Bx samples were assessed for TIM4 with GAPDH normalization by qRT-PCR (Figure 5A). Post-transplant (2 h after reperfusion) expression levels were compared to pre-transplant (after cold storage) samples (Post/Pre) to determine the perioperative TIM4 profile. Consistent with the mouse data (Supplementary Figure S1), peritransplant hepatic TIM4 levels correlated positively with early OLT function, assessed by sAST (r = 0.4028, p = 0.0090, Figure 5B) and sALT (r = 0.4223, p = 0.0060, Figure 5C) at postoperative day 1 (POD1). Then, based on the TIM4 gene expression pattern, liver Bx samples were divided into TIM4-low (n = 21) and TIM4-high (n = 20) groups, according to the median split (Figure 5D). Patients' demographic data and clinical parameters are shown in Supplementary Tables S4, S5. We found no correlation between TIM4 levels and recipient surgical parameters, including age, gender, race, ABO compatibility, pre-transplant blood tests, preoperative hospital stay, cold ischemia time, warm ischemia time, or blood transfusions during the surgery. However, BMI was significantly higher in the TIM4-high group (p = 0.020) (Supplementary Table S4). There was no correlation between TIM4 grouping and other donor data (Supplementary Table S5), including age, gender, race, BMI, pre-procurement blood tests, or donation status (after circulatory or brain death). Of note, the TIM4-high expression clinical cohort had significantly higher sAST at POD1 and sALT levels at POD1-2 (Figures 5E,F), reflecting a deteriorated OLT function. Although the difference failed to reach statistical significance, TIM4-high cases experienced an increased frequency of EAD compared to the TIM4-low group (Figure 5G, 10.5% vs. 35.0%, p = 0.0670). We analyzed graft survival rates to examine the relationship between TIM4 expression and OLT outcomes (median follow-up, 1,269 days; range, 3–1,892 days). Strikingly, the TIM4-high patient cohort experienced significantly inferior overall graft survival (Figure 5H, p = 0.0051). None of the patients underwent secondary OLT.

We also analyzed the innate, and adaptive immune gene expression profiles in human liver Bx obtained during OLT (Figure 6). Consistently, TIM4-high grafts exhibited increased mRNA levels coding for T cell activation markers, CD3 (p = 0.0246), CD8 (p = 0.0628), CD28 (p = 0.0216), and IL-17 (p = 0.0216), macrophage activation markers, CD80 (p = 0.0188), CD86 (p = 0.0072), CXCL10 (p = 0.0257), TLR2 (p = 0.0018) and TLR4 (p = 0.0024), as well as the neutrophil activation marker, Cathepsin G (p = 0.0196). Hence, the TIM4-high phenotype was accompanied by accelerated post-reperfusion innate/adaptive immune activation and enhanced hepatocellular damage in the early post-OLT phase.

Since we showed the relationship between TIM4 and ATF4, CHOP, cleaved caspase 3, caspase 8, and MDA expression in the experimental murine settings, we then analyzed our clinical liver Bx samples to validate the aforementioned findings. Consistent with the mouse data, peritransplant TIM4 levels at 2 h after reperfusion in human OLT correlated positively with hepatic ATF4 (r = 0.3958, p = 0.0104), CHOP (r = 0.4681, p = 0.0020), MDA (r = 0.2970, p = 0.0627), cleaved caspase 3 (r = 0.4906, p = 0.0011) and caspase 8 (r = 0.4352, p = 0.0045) (Figures 7A–E).