Growth arrest–specific protein 6 is hepatoprotective against murine ischemia/reperfusion injury

Authors

  • Laura Llacuna,

    1. Liver Unit, Hospital Clinic, Institut d'Investigacions Biomèdiques August Pi i Sunyer/Centre d'Investigació Biomèdica Esther Koplowitz, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain
    2. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Cristina Bárcena,

    1. Liver Unit, Hospital Clinic, Institut d'Investigacions Biomèdiques August Pi i Sunyer/Centre d'Investigació Biomèdica Esther Koplowitz, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain
    2. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Lola Bellido-Martín,

    1. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    Search for more papers by this author
  • Laura Fernández,

    1. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    Search for more papers by this author
  • Milica Stefanovic,

    1. Liver Unit, Hospital Clinic, Institut d'Investigacions Biomèdiques August Pi i Sunyer/Centre d'Investigació Biomèdica Esther Koplowitz, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain
    2. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    Search for more papers by this author
  • Montserrat Marí,

    1. Liver Unit, Hospital Clinic, Institut d'Investigacions Biomèdiques August Pi i Sunyer/Centre d'Investigació Biomèdica Esther Koplowitz, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain
    2. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    Search for more papers by this author
  • Carmen García-Ruiz,

    1. Liver Unit, Hospital Clinic, Institut d'Investigacions Biomèdiques August Pi i Sunyer/Centre d'Investigació Biomèdica Esther Koplowitz, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain
    2. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    Search for more papers by this author
  • José C. Fernández-Checa,

    Corresponding author
    1. Liver Unit, Hospital Clinic, Institut d'Investigacions Biomèdiques August Pi i Sunyer/Centre d'Investigació Biomèdica Esther Koplowitz, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain
    2. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    3. Research Center for Alcoholic Liver and Pancreatic Diseases, Keck School of Medicine, University of Southern California, Los Angeles, CA
    • Liver Unit, Hospital Clinic i Provincial, c/Villarroel 170, 08036 Barcelona, Spain
    Search for more papers by this author
  • Pablo García de Frutos,

    Corresponding author
    1. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    • Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Barcelona, Spain
    Search for more papers by this author
    • These authors share senior authorship of this work.

  • Albert Morales

    Corresponding author
    1. Liver Unit, Hospital Clinic, Institut d'Investigacions Biomèdiques August Pi i Sunyer/Centre d'Investigació Biomèdica Esther Koplowitz, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain
    2. Department of Cell Death and Proliferation, Instituto de Investigaciones Biomédicas de Barcelona, Consejo Superior de Investigaciones Científicas, Rosselló 160, pta 6, 08036 Barcelona, Spain
    • Liver Unit, Hospital Clinic i Provincial, c/Villarroel 170, 08036 Barcelona, Spain
    Search for more papers by this author
    • These authors share senior authorship of this work.


  • Potential conflict of interest: Nothing to report.

Abstract

Growth arrest–specific gene 6 (GAS6) promotes growth and cell survival during tissue repair and development in different organs, including the liver. However, the specific role of GAS6 in liver ischemia/reperfusion (I/R) injury has not been previously addressed. Here we report an early increase in serum GAS6 levels after I/R exposure. Moreover, unlike wild-type (WT) mice, Gas6−/− mice were highly sensitive to partial hepatic I/R, with 90% of the mice dying within 12 hours of reperfusion because of massive hepatocellular injury. I/R induced early hepatic protein kinase B (AKT) phosphorylation in WT mice but not in Gas6−/− mice without significant changes in c-Jun N-terminal kinase phosphorylation or nuclear factor kappa B translocation, whereas hepatic interleukin-1β (IL-1β) and tumor necrosis factor (TNF) messenger RNA levels were higher in Gas6−/− mice versus WT mice. In line with the in vivo data, in vitro studies indicated that GAS6 induced AKT phosphorylation in primary mouse hepatocytes and thus protected them from hypoxia-induced cell death, whereas GAS6 diminished lipopolysaccharide-induced cytokine expression (IL-1β and TNF) in murine macrophages. Finally, recombinant GAS6 treatment in vivo not only rescued GAS6 knockout mice from severe I/R-induced liver damage but also attenuated hepatic damage in WT mice after I/R. Conclusion: Our data have revealed GAS6 to be a new player in liver I/R injury that is emerging as a potential therapeutic target for reducing postischemic hepatic damage. (HEPATOLOGY 2010;)

The growth arrest–specific gene 6 (GAS6) product and its tyrosine kinase TAM receptors (Tyro3, Axl, and Mer) are involved in growth and survival processes during tissue repair and development.1, 2 GAS6 is a vitamin K–dependent protein that has high structural homology with the natural anticoagulant protein S; they share the same modular composition and 40% of their sequence identity. Despite these common features, the biological roles of GAS6 and protein S are clearly differentiated, with GAS6 being mainly involved in cell protection and tissue formation and less involved in the coagulation cascade.3, 4

The low concentration of GAS6 in plasma and its specific pattern of tissue expression suggest a unique function of GAS6 among vitamin K–dependent proteins. In the liver, GAS6 is mainly expressed in Kupffer cells at levels below those observed in other tissues such as lung, kidney, and heart tissues.3 However, after a specific liver injury, other hepatic cell types may participate in its production. For instance, GAS6 produced by hepatic stellate cells and its receptor Axl participate in the signaling involved in the wound healing response to liver injury by carbon tetrachloride, and oval cells induce GAS6 production after hepatectomy.5-7 In addition, increased plasma levels of GAS6 have been reported in different pathologies, such as cancer,8-10 acute coronary syndrome,11 pulmonary embolism,12 acute pancreatitis,13 and severe sepsis,14, 15 both in patients and in experimental models. In particular, GAS6 expression correlates with disease severity in patients with septic shock, especially with respect to renal and hepatic dysfunction.15 However, the role of GAS6 in hepatocyte signaling and liver injury after ischemia/reperfusion (I/R) has not been previously reported to the best of our knowledge.

GAS6 and its signaling through TAM receptors (Mer, Axl, and Tyro3) have been proposed not only as a protective pathway in several cell types, including endothelial and epithelial cells, neurons, and fibroblasts,16-19 but also as a molecular device modulating cytokine secretion. For instance, mice deficient in TAM receptors or with mutated Mer displayed high susceptibility to endotoxic shock, with monocytes showing increased tumor necrosis factor (TNF) secretion after lipopolysaccharide (LPS) challenge.20 Moreover, recent data on monocytes/macrophages have shown that exogenous GAS6 reduces LPS-induced TNF and interleukin-1 (IL-1) stimulation via Mer signaling but not via Axl or Tyro3 signaling.21 Therefore, our aim was to address the role of GAS6 during hepatic I/R injury and the potential mechanisms involved. Our results showed that plasma GAS6 levels increased early during hepatic I/R as the hepatic GAS6 messenger RNA (mRNA) content decreased before major liver injury. Exogenous GAS6 induced protein kinase B (AKT) phosphorylation and thus protected primary hepatocytes from hypoxia. In addition, partial I/R was lethal in GAS6 knockout (KO) mice because of massive hepatocellular injury, an event that was abrogated by a recombinant GAS6 intravenous injection. Overall, these findings indicate that GAS6 protects against liver I/R injury, and it is emerging as a potential novel target in diverse clinical settings in which hepatic I/R damage occurs, such as liver transplantation, hemorrhagic shock, and liver surgery.

Abbreviations

AKT, protein kinase B; ALT, alanine aminotransferase; CM, control medium; GAS6, growth arrest–specific gene 6; H&E, hematoxylin and eosin; I/R, ischemia/reperfusion; IL, interleukin; JNK, c-Jun N-terminal kinase; KO, knockout; LPS, lipopolysaccharide; mRNA, messenger RNA; NF-κB, nuclear factor kappa B; PBS, phosphate-buffered saline; TNF, tumor necrosis factor; TUNEL, terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling; WT, wild type.

Materials and Methods

Partial Hepatic Ischemia and Treatments.

The experimental animal protocol was approved by the animal care and use committee of the Institut d'Investigacions Biomèdiques August Pi i Sunyer. Wild-type (WT) C57BL/6 and Gas6−/− male mice, backcrossed into the C57BL/6 background (8-12 weeks), were generated and propagated as previously characterized.22 Hepatic partial warm ischemia was performed for 90 minutes,23 as detailed in the supporting information. Mice were pretreated for 15 to 20 minutes before surgery with the recombinant GAS6 protein (0.5-10.0 μg/mouse; R&D, St. Louis, MO) or with an equal volume of the vehicle [a phosphate-buffered saline (PBS) solution]. To follow animal survival, we monitored the mice every 12 hours for 1 week.

Cell Culture and Hypoxia Exposure.

Primary hepatocytes, obtained from mouse livers by collagenase digestion and cultured on collagen-coated plates,23 were routinely grown at 37°C with 5% CO2 in Dulbecco's modified Eagle's medium/F12 medium with 10% fetal bovine serum under a normoxic atmosphere or were exposed to hypoxia (1% O2 and 5% CO2) as previously described24, 25 (see the supporting information). In some experiments, conditioned media from GAS6-expressing HEK293 cells (100 ng GAS6/mL) or from control HEK293 pcDNA3-transfected cells were added to cultured mouse hepatocytes.26 Cell survival was measured with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay and trypan blue exclusion.

Western Blotting and Nuclear Extract Preparation.

Cell and nuclear extracts were prepared as previously described,27 and protein levels were analyzed with specific antibodies (see the supporting information).

Statistical Analyses.

The results are expressed as means and standard deviations; the number of individual experiments is detailed in the figure legends. Statistical significance was established by one-way analysis of variance followed by Dunnett and Tukey-Kramer post hoc tests. Animal survival was evaluated with the Kaplan-Meier method and compared with the log-rank test.

Results

I/R Decreases Hepatic GAS6 mRNA Levels While Increasing the Presence of GAS6 in Serum.

We evaluated whether I/R modulated hepatic GAS6 homeostasis in WT C57BL/6 mice subjected to partial ischemia for 90 minutes; we assessed the GAS6 mRNA content and GAS6 levels in serum after different reperfusion times. The GAS6 mRNA levels, determined from liver biopsy samples, fell early after reperfusion and remained below control levels up to 16 hours after reperfusion (Fig. 1A). In contrast, enzyme-linked immunosorbent assay analyses of serum indicated a time-dependent increase in the levels of GAS6 detected as soon as 3 hours after reperfusion, and they remained above control levels for 24 hours after reperfusion (Fig. 1B). Although this model of partial I/R typically results in maximal liver damage between 4 and 8 hours after reperfusion, increased serum alanine aminotransferase (ALT) levels were already detected as soon as 1 hour after reperfusion (659 ± 284 U/mL), and this coincided with the decrease in hepatic GAS6 mRNA levels and the initiation of the progressive increase observed in GAS6 serum levels. Thus, GAS6 homeostasis is regulated during hepatic I/R.

Figure 1.

Hepatic GAS6 regulation and susceptibility of Gas6−/− mice to partial hepatic I/R. (A) GAS6 mRNA levels in liver biopsy samples from mice subjected to 90 minutes of liver ischemia and different times of reperfusion were measured by real-time polymerase chain reaction with β-actin as a control (n = 3). (B) Serum samples from ischemic mice were collected after different times of reperfusion, and the GAS6 protein content was measured with an enzyme-linked immunosorbent assay kit. (C) WT mice and Gas6−/− mice were subjected to liver ischemia for 90 minutes, and their survival for 7 days was monitored. No changes in the animal living status were observed after day 3. *P = 0.008 versus I/R-exposed WT mice by a log-rank test (n = 10 in each group).

Partial Hepatic I/R Is Lethal in GAS6-KO Mice Because of Massive Hepatocellular Damage and a Proinflammatory State.

The model of partial hepatic I/R follows a typical time-dependent pattern characterized by initial tissue damage that is resolved within 24 to 48 hours because of liver regeneration. In order to critically assess the role of GAS6 during I/R-mediated liver injury, we generated Gas6−/− mice as previously described20; they exhibited abnormalities in platelet function, erythropoiesis, and endothelial activation but normal liver morphology and TAM receptor expression. Interestingly, 9 of 10 Gas6−/− mice died within the first 12 hours of reperfusion, with 50% of the animals dying within 8 hours of reperfusion. In contrast, 90% of WT mice survived partial I/R (Fig. 1C), and this was in line with our previous studies.23 The deaths of the Gas6−/− mice during hepatic I/R were most likely related to liver failure due to massive hepatocellular damage because serum aminotransferase levels 6 hours after reperfusion were dramatically elevated in GAS6-deficient mice versus WT mice (Fig. 2A). Moreover, this outcome mirrored histological findings revealing severe deterioration of the liver parenchyma after I/R exposure in GAS6-deficient mice with respect to WT mice (Fig. 2B). In addition, parallel liver sections from Gas6−/− mice undergoing I/R displayed extensive cell death detected by terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling (TUNEL) staining, which affected areas all over the hepatic parenchyma; this contrasted with the confined TUNEL-positive areas observed in WT mice (Fig. 2B). Finally, we examined whether GAS6 regulates inflammatory mediators during I/R as previously proposed.20, 21 Interestingly, TNF and IL-1β mRNA levels were markedly elevated in the null mice versus the WT animals exposed to I/R (Fig. 2C,D) after 6 hours of reperfusion, and this suggests that GAS6 counterbalances the inflammatory response after hepatic ischemia. Thus, these findings clearly demonstrate overall that the lack of GAS6 sensitizes the liver to I/R and results in massive hepatic destruction incompatible with life.

Figure 2.

GAS6-deficient mice displayed severe liver damage after I/R exposure. (A) WT mice and Gas6−/− mice were subjected to liver ischemia for 90 minutes, and aminotransferase levels in serum samples were measured after 6 hours of reperfusion (n = 3-4). (B) Representative images of H&E and TUNEL staining of liver biopsy samples from I/R-treated WT and Gas6−/− mice after 6 hours of reperfusion. (C,D) mRNA levels of proinflammatory cytokines (TNF and IL-1β) were detected by quantitative reverse-transcription polymerase chain reaction in liver samples after 6 hours of reperfusion (n = 6).

GAS6-Deficient Mice Exhibit Defective Hepatic I/R-Induced Mer Signaling and AKT Phosphorylation.

Although massive liver injury in GAS6-null mice was evident 3 to 6 hours after reperfusion, a significant increase in ALT levels was already detected as soon as 1 hour after ischemia in comparison with WT mice (1835 ± 893 U/mL for Gas6−/− mice versus 695 ± 248 U/mL for WT mice). Therefore, we next evaluated early changes in hepatic signaling responsible for the hepatic sensitization to I/R observed in Gas6−/− mice and focused particularly on protective/apoptotic pathways. Because c-Jun N-terminal kinase (JNK) activation has been shown to contribute to hepatic I/R injury, we analyzed the phosphorylation state of JNK 60 minutes after reperfusion. The robust phosphorylation of JNK p46 and p54 isoforms observed after reperfusion was comparable in the livers of both WT and GAS6-null mice (Fig. 3A); this particular pathway in the sensitization of GAS6-KO mice to hepatic I/R was discarded. Because AKT regulates cell survival, we examined the phosphorylation state of AKT during hepatic I/R. In contrast to JNK activation, hepatic AKT phosphorylation was clearly reduced in Gas6−/− mice versus I/R-exposed WT mice (Fig. 3B). In addition to AKT, nuclear factor kappa B (NF-κB) regulates hepatocellular susceptibility to I/R,28 and hence we compared the extent of activation of NF-κB in WT and GAS6-deficient mice. In contrast to AKT, the translocation of p65 to hepatic nuclear extracts after 60 minutes of reperfusion was similar in WT and GAS6-null mice (Fig. 3C).

Figure 3.

Impaired hepatic AKT activation and Mer phosphorylation in GAS6-deficient mice during I/R. WT mice and Gas6−/− mice were subjected to hepatic ischemia for 90 minutes, and liver protein extracts were obtained for western analysis after 60 minutes of reperfusion. (A) JNK activation, (B) AKT phosphorylation, (C) NF-κB nuclear p65 translocation, and (D) GAS6 and phosphorylation of Axl and Mer were analyzed. Representative blots of two independent experiments are shown.

Because the GAS6 serum concentration increases after I/R, we evaluated whether ischemia stimulates GAS6 signaling through activation of TAM receptors. First, GAS6 protein levels increased in liver extracts from I/R-exposed WT animals (Fig. 3D), and as expected, these changes were undetectable in GAS6-deficient mice. Axl and Mer are TAM receptors located in liver cells that are phosphorylated after GAS6 binding. Therefore, we decided to verify their participation in I/R-induced TAM signaling. Although no changes in Axl activation were evident after I/R, an increase in Mer phosphorylation was detected in WT mice exposed to I/R, but this response was blunted in GAS6-KO mice (Fig. 3D). Hence, our data indicate that GAS6 levels increase in the liver after I/R and induce Mer-dependent signaling and AKT phosphorylation independently of NF-κB activation. The lack of these events in GAS6-KO mice may contribute to their susceptibility to hepatic I/R injury.

GAS6 Supplementation Protects Primary Mouse Hepatocytes from Hypoxia-Induced Cell Death.

In light of the previous findings, we extended the in vivo observations to cultured hepatocytes and examined whether the exogenous administration of GAS6 directly regulates AKT phosphorylation and hypoxia susceptibility. First, we analyzed NF-κB activation after the addition of preconditioned media from GAS6-overexpressing HEK293 cells to primary mouse hepatocytes. GAS6 supplementation did not change the p65 nuclear levels in cultured mouse hepatocytes (Fig. 4A). However, a marked increase in AKT phosphorylation was detected after the addition of a GAS6-containing medium. As soon as 15 minutes after the administration of the GAS6 conditioned medium, primary hepatocytes displayed robust AKT phosphorylation (Fig. 4B). Moreover, in accordance with the in vivo findings, no changes in JNK activation were observed after hepatocyte incubation with the conditioned medium containing GAS6 (Fig. 4C). These finding confirm that parenchymal cells are targets of GAS6, which results in AKT phosphorylation regardless of p65 nuclear translocation, suggesting that a similar mechanism is occurring in vivo after I/R.

Figure 4.

GAS6 induced AKT activation in primary mouse hepatocytes and protected against oxygen deprivation. (A) Cultured mouse hepatocytes were treated with a control medium (CM) or a medium enriched in GAS6 (CM+GAS6), and cell extracts were prepared for western analysis at different times. (A) NF-κB nuclear translocation, (B) AKT phosphorylation, and (C) JNK activation were analyzed. Representative blots of three independent experiments are shown. (D) Cell viability of primary hepatocytes exposed for 24 hours to hypoxia (1% O2) in the presence or absence of the GAS6-enriched medium (n = 3). (E,F) The effect of GAS6 before incubation (20 minutes) on the cytokine production (TNF and IL-1β) of RAW264.7 cells treated with LPS (50 ng/mL) was analyzed after 2 hours (n = 4).

To verify that the signaling effects induced by GAS6 administration could have a protective effect against oxygen deprivation, primary mouse hepatocytes exposed to hypoxia (1% O2) were preincubated with a conditioned medium with or without GAS6. First, we verified that hypoxia activated hypoxia inducible factor 1 alpha, a known target of oxygen deprivation. In agreement with previous findings,24 the nuclear levels of hypoxia inducible factor 1 alpha increased in hepatocytes cultured with 1% O2 (not shown). Interestingly, GAS6 supplementation protected cultured hepatocytes against hypoxia-induced cell death (survival of 25% ± 4% in control cells versus 40% ± 5% in GAS6-supplemented cells; Fig. 4D). Thus, these in vitro findings further confirm the link between GAS6 and AKT signaling observed in vivo during I/R and establish a potential protective mechanism against hypoxia. Therefore, our work suggests that defective AKT activation in Gas6−/− mice may contribute to the sensitivity of the liver to I/R. The identification of other intracellular mechanisms that may play a relevant role in the signaling triggered by GAS6 downstream of AKT in hepatic I/R deserves further investigation. In this respect, GAS6 has been shown to activate forkhead box O1a in cultured endothelial cells.29

Moreover, because GAS6 has been shown to reduce LPS-induced inflammatory cytokine release in human monocytes21 and in murine Sertoli cells30 and because the LPS/toll-like receptor pathway is increasingly recognized as an important contributing mechanism in I/R-induced liver injury,31 we next decided to determine if this mechanism could also modulate the response of murine macrophages after LPS challenge. RAW264.7 macrophages greatly increased TNF and IL-1β mRNA levels after LPS treatment, and this response was significantly reduced by GAS6 (Fig. 4E). Hence, these findings indicate that the intrahepatic increase in GAS6 after I/R restrains the overgeneration of inflammatory cytokines and that the lack of this pathway in the absence of GAS6 further contributes to the sensitization to I/R-induced liver damage.

Recombinant GAS6 Protects GAS6-Deficient Mice Against Liver I/R Damage.

We next evaluated whether the severe liver injury of Gas6−/− mice after I/R could be prevented by the administration of recombinant GAS6. GAS6-deficient mice were intravenously injected with a commercial mouse recombinant protein (5 μg/mouse) before they were subjected to partial ischemia. Remarkably, Gas6−/− mice that received recombinant GAS6 protein 15 to 20 minutes before ischemia displayed reduced liver damage that was comparable to the injury seen in WT mice; this was reflected by the lower ALT and aspartate aminotransferase concentrations detected in serum (Fig. 5A and Supporting Fig. 1). Moreover, doses of recombinant GAS6 greater than 5 μg/mouse (up to 10 μg/mouse) exerted a similar protective effect against I/R (not shown), and GAS6 even at doses 10 times lower (0.5 μg/mouse) was able to induce liver protection but to a lesser extent (Supporting Fig. 2). In parallel with the aminotransferase levels, liver biopsy samples from GAS6-injected KO mice displayed preserved parenchymal architecture and organization with lesser areas of hepatocellular damage, as shown by hematoxylin and eosin (H&E) staining (Fig. 5B). Moreover, TNF and IL-1β expression after I/R was repressed at mRNA levels by GAS6 administration to both WT and null mice (Supporting Fig. 3). Thus, these results confirm that the sensitivity of Gas6−/− mice to hepatic I/R injury was due to the lack of expression of GAS6 and not due to other previously unnoticed phenotypic changes. Importantly, when recombinant GAS6 was administered to WT mice, preservation of the liver structure and significant prevention of hepatic injury after I/R were observed in GAS6-injected WT mice; this was measured by aminotransferase levels and H&E staining (Fig. 5A,B). These results indicate not only that a GAS6 deficiency sensitizes mice to I/R-induced liver damage but also that the elevation of serum GAS6 levels is a protective strategy against liver I/R injury.

Figure 5.

GAS6 administration attenuated hepatic I/R injury in Gas6−/− mice and WT mice. GAS6 or a vehicle (PBS) was intravenously injected into WT mice and Gas6−/− mice subjected to 90 minutes of liver ischemia. (A) After 6 hours of reperfusion, serum ALT levels were measured in blood samples (n = 3-6). (B) Representative H&E staining of liver biopsy samples obtained after a sham operation or 6 hours of reperfusion from WT mice and Gas6−/− mice that received an intravenous injection of the GAS6 solution (5 μg/mouse) or vehicle (PBS).

Discussion

Our findings have disclosed a novel role for GAS6 in hepatocellular defense against hypoxia and I/R and have expanded our knowledge of the biological facets of this atypical member of the vitamin K–dependent family, which has negligible anticoagulant function but is recognized to be involved in cancer, inflammation, and the regulation of liver regeneration and fibrosis.1, 2, 5, 6 Specifically, we have shown that GAS6 is dispensable for the activation of JNK and NF-κB during hepatic I/R but is required for efficient AKT phosphorylation in hepatocytes; this agrees with previous findings reported for other cell types.3 In addition to increased susceptibility to hepatic I/R injury, GAS6-null mice exhibit enhanced expression of IL-1β and TNF, and this suggests that GAS6 may mediate the suppression of liver inflammation during I/R. These findings are in line with previously reported data on human and murine monocytes/macrophages20, 21, 32 and Sertoli cells30 via the regulation of TAM receptor signaling.1 Together, our findings indicate that the susceptibility to hepatic I/R injury in the absence of GAS6 reflects the combination of the inability to turn on survival pathways dependent on AKT activation and the overstimulation of inflammatory cytokines. Interestingly, this hepatoprotective role displayed by GAS6 against I/R contrasts with findings in mouse hearts subjected to warm ischemia, in which GAS6 promoted graft destruction by enhancing interactions between endothelial cells, platelets, and leukocytes33; this exemplifies the importance of organ-dependent mechanisms in GAS6 signaling.

The outcome of GAS6-null mice after hepatic I/R is reversible upon the addition of GAS6, and this suggests the maintenance of TAM receptors in the null mice and the functional interaction of GAS6 with TAM receptors. Among TAM receptors, most of the protective effects of GAS6 described in different cell types are due to the binding of GAS6 to Axl.34-37 In contrast to these findings, Axl seems to play a minor role in hepatic I/R because increased Axl phosphorylation was not detected after I/R exposure. Instead, Mer becomes phosphorylated in WT mice after I/R, and this response is blunted in GAS6-KO mice. Moreover, recent data on differentiated monocytes and macrophages have shown that GAS6 inhibits TNF and IL-1β stimulation in response to LPS via Mer activation.21 Moreover, the binding of GAS6 to Mer but not to Axl or Tyro3 results in the activation of AKT via glycogen synthase kinase 3β phosphorylation, which in turn prevents NF-κB activation. Together, these data and the present study expand our knowledge of the biological impact of the GAS6-Mer interaction, which elicits a combined anti-inflammatory and survival pathway to protect against hepatic I/R. Although we did not examine the involvement of glycogen synthase kinase 3β in the protective action of GAS6 in hepatic I/R, we did observe that GAS6 is dispensable for NF-κB activation in normothermic I/R.

In agreement with a protective role of GAS6 during hepatic I/R, the serum levels of GAS6 increase early after I/R, and this parallels the up-regulation of GAS6 in hepatic extracts. Recent findings in liver regeneration and chemical-induced liver damage have indicated predominant expression of GAS6 in Kupffer cells and hepatic stellate cells.5, 6 Although we did not estimate the relative contribution of these putative sources of GAS6 during I/R, GAS6 reproduced the anti-inflammatory effect of down-regulating TNF and IL-1β in RAW264.7 macrophages, a surrogate cell line for Kupffer cells. In this scenario, it would be tempting to speculate that GAS6 derived from hepatic macrophages initiates a paracrine signaling event via Mer in hepatocytes to activate protective and anti-inflammatory pathways of relevance in hepatic I/R.

In summary, our work has identified GAS6 as a survival factor released during hepatic I/R damage that protects hepatocytes from oxygen deprivation and reduces inflammatory cytokine production. It is quite interesting that GAS6 not only rescued null mice from I/R-mediated liver injury but also proved useful in protecting WT mice against hepatic I/R damage. Because of the broad implications of hepatic I/R injury, GAS6 is emerging as a novel pharmacological therapy of potential relevance in different clinical settings.

Acknowledgements

The technical assistance of Susana Nuñez and Anghara Menendez is highly appreciated.

Ancillary