Potential conflict of interest: Nothing to report.
We have shown that activation of toll-like receptor 4 (TLR4) and its interferon regulatory factor 3 (IRF3)-dependent downstream signaling pathway are required for the development of liver ischemia/reperfusion injury (IRI). This study focused on the role of TLR4-IRF3 activation pathway products, in particular, chemokine (C-X-C motif) ligand 10 (CXCL10). The induction of CXCL10 by liver IR was rapid (1 hour postreperfusion), restricted (ischemic lobes), and specific (no CXCL9 and CXCL11 induction). Functionally, CXCL10 was critical for IR-induced liver inflammation and hepatocellular injury. CXCL10 knockout (KO) mice were protected from IRI, as evidenced by reduced serum alanine aminotransferase (sALT) levels and preserved liver histological detail. The induction of pro-inflammatory genes, such as tumor necrosis factor alpha (TNF-α), interleukin 1β (IL-1β), IL-6, and IL-12β was diminished, whereas the induction of the IL-10 gene remained intact in CXCL10 KO mice, indicating an altered liver response against IR. This was accompanied by selective down-regulation of extracellular signal-regulated kinase (ERK), but intact Jun N-terminal kinase (JNK), activation in the KO IR livers. This altered liver inflammation response was (1) specific to IR, because lipopolysaccharide (LPS) induced a comparable pro-inflammatory response in CXCL10 KO and wild-type (WT) mice; and (2) responsible for liver cytoprotection from IR, because neutralization of IL-10 restored local inflammation and hepatocellular damage. Conclusion: CXCL10 regulates liver inflammation response against IRI, and its deficiency protected livers from IRI by local IL-10–mediated cytoprotection. Targeting CXCL10 may provide a novel therapeutic means to ameliorate liver IRI in clinics. (HEPATOLOGY 2008.)
Ischemia/reperfusion injury (IRI) develops in the absence of exogenous Ag, and innate immunity has been thought to play a dominant pathogenic role.1–3 Liver ischemia activates Kupffer cells, and to a lesser degree endothelial cells as well as hepatocytes, leading to the formation of reactive oxygen species and secretion of pro-inflammatory cytokines/chemokines. The oxidant stress directly damages endothelial cells/hepatocytes, whereas the soluble factors are largely responsible for leukocyte recruitment and activation, leading to the full development of intrahepatic inflammation causing further organ damage. Although excessive pro-inflammatory response has been recognized as the key element leading to IRI, the mechanisms that initiate and regulate liver inflammation cascade remain to be elucidated.
We and others have reported that mammalian sentinel receptor toll-like receptor 4 (TLR4) is involved in the initiation of IRI.4–7 We found that livers in TLR4 knockout (KO) mice were protected from IRI and associated inflammation.4–7 Furthermore, MyD88-independent signaling mediated by IRF3, downstream of TLR4 activation, was critical, because mice deficient in interferon regulatory factor 3 (IRF3) but not MyD88 were protected from IRI.5 MyD88-dependent signaling in TLR4 activation pathway leads to direct nuclear factor kappa B activation and induction of pro-inflammatory cytokines, whereas the IRF3-mediated signaling induces type 1 IFN and IFN-inducible genes, such as chemokine (C-X-C motif) ligand 10 (CXCL10).8, 9 Our findings prompted us to analyze gene products downstream of IRF3, in particular, CXCL10, which was induced in a TLR4-dependent and IRF3- dependent manner in livers undergoing IRI.5
CXCL10, a chemokine that targeted activated T cells and natural killer cells expressing CXCR3,10, 11 has been implicated in inflammatory diseases, mostly associated with antigen-specific T cell responses.12–14 Its up-regulation has also been detected in IR-exposed livers, kidneys, and hearts.15–17 However, the mechanism of CXCL10 induction and function in the pathogenesis of innate dominated IRI, particularly in the early acute phase, is not clear. Although T cell recruitment into inflamed tissues by local CXCL10 has been established, it generally associates with the activation of antigen-specific T cells and occurs in days. In the acute IRI phase (hours after reperfusion) without T cell activation by exogenous Ag, de novo recruitment of peripheral T cells into IRI livers was not significant.18, 19 Thus, in this study, we first determined the function of CXCL10 in the acute phase of liver IRI using gene KO mice, and then analyzed IR-induced liver inflammation in the presence or absence of CXCL10. We show that CXCL10 is required for the induction of pro-inflammatory response and the development of hepatocellular injury.
Male wild-type (WT; C57BL/6) mice (8-12 weeks old) were used (Jackson Laboratory, Bar Harbor, ME). CXCL10-deficient (KO; C57BL/6) mice were provided by Dr. Andrew D. Luster (MGH; Boston, MA). Animals were housed in the University of California Los Angeles 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” prepared by the National Academy of Sciences and published by the National Institute of Health (NIH publication 86-23 revised 1985).
Liver IRI Model.
We have developed a warm hepatic IRI model in mice, as described.5, 6 Briefly, mice were anesthetized and injected with heparin, and an atraumatic clip was used to interrupt the arterial and portal venous blood supply to the liver cephalad lobes. After 90 minutes of partial hepatic warm ischemia, the clip was removed. Mice were sacrificed after various times of reperfusion; liver, spleen, and peripheral blood were collected. Serum alanine aminotransferase (sALT) levels, an indicator of hepatocellular injury, were measured by auto analyzer (ANTECH Diagnostics, Los Angeles, CA). Specimens were fixed in 10% buffered formalin; liver sections (4 μm) were stained with hematoxylin-eosin and analyzed blindly. Sham WT controls underwent the same procedure, but without vascular occlusion.
For the lipopolysaccharide (LPS)-induced liver inflammation study, 1 μg LPS (Sigma, St. Louis, MO) in 500 μL phosphate-buffered saline was administered intraperitoneally into WT/CXCL10 KO mice; livers and spleens were harvested at 3 hours. For interleukin-10 (IL-10) neutralization experiments, rat anti-mouse IL-10 monoclonal antibody (mAb) (JES5-2A5, Bio Express, West Lebanon, NH) was administered (0.5 mg/mouse intraperitoneally) at the end of lobe clamping. Normal rat immunoglobulin G was used as control.
Liver Neutrophil Infiltration.
Livers were snap frozen, and cryostat sections (5 μm) were fixed in acetone. Endogenous peroxidase activity was inhibited with 0.3% hydrogen peroxidase. Sections were then blocked with 10% normal goat serum. Primary rat mAb against mouse polymorphonuclear leukocytes (Ly-6G, 1A8; BD Pharmingen) was diluted (1/250 in 3% normal goat serum), and 100 μL was added to each section. The primary antibody was incubated overnight at 4°C. The secondary antibody, a biotinylated goat anti-rat immunoglobulin G (Vector; diluted 1:200), was incubated for 40 minutes at room temperature. Sections were incubated with immunoperoxidase (ABC Kit, Vector), washed, and developed with a 3,3′-diaminobenzidine kit (Vector). Slides were counterstained with hematoxylin, rinsed with ammonia, and then rinsed with tap water. Negative control was prepared by omission of primary antibody. The sections were evaluated blindly by counting labeled cells in triplicates in 30 high-power fields (HPF); results are expressed as average number of positive cells/HPF. In parallel, the accumulation of activated neutrophils in the livers was assessed by staining for chloroacetate esterase, a specific marked for neutrophils, using the Naphthol ASD-chloroacetate esterase kit (Sigma). Polymorphonuclear cells were identified by positive staining and morphology.
Two and a half micrograms RNA was reverse-transcribed into complementary DNA using SuperScriptTM III First-Strand Synthesis System (Invitrogen, Carlsbad, CA). Quantitative polymerase chain reaction was performed using the DNA Engine with Chromo 4 Detector (MJ Research, Waltham, MA). In a final volume of 25 μL, the following were added: 1× SuperMix (Platinum SYBR Green qPCR Kit, Invitrogen, Carlsbad, CA), complementary DNA, and 0.5 mM of each primer. Amplification conditions were: 50°C (2 minutes), 95° (5 minutes) followed by 50 cycles of 95°C (15 seconds), 60°C (30 seconds).
Western Blot of Liver Mitogen-Activated Protein Kinases.
Protein was extracted from liver tissues with ice-cold PBSTDS (1% Triton X-100, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate in PBS, pH 7.4) buffer. Proteins (40 μg) in sodium dodecyl sulfate–loading buffer were subjected to 10% to 20% sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane. Antibodies against phosphorylated and total cellular extracellular signal-regulated kinase (ERK), Jun N-terminal kinase (JNK), and p38 mitogen-activated protein (MAP) kinase proteins (Cell Signaling Technology, Santa Cruz, CA), and β-actin (Abcam Inc., Cambridge, MA) were used. Relative quantities of protein were determined using a densitometer (Kodak Digital Science 1D Analysis Soft-ware, Rochester, NY).
All values are expressed as mean ± standard deviation (SD). Data were analyzed with an unpaired 2-tailed Student t test. P < 0.05 was considered statistically significant.
IR Induces Local CXCL10 Expression in the Ischemic Liver Lobes.
We have shown that TLR4-dependent liver IRI is accompanied by intrahepatic CXCL10 induction that is mediated by IRF3 signaling.5 In this study, a significant increase in liver CXCL10 expression was noted as early as at 1 hour after reperfusion (after 90 minutes of ischemia), with a peak at 4 hours (Fig. 1A). IR-induced CXCL10 was restricted to ischemic lobes (26-fold increase versus 5-fold increase in nonischemic lobes), with no changes in the spleen (1.7-fold) (Fig. 1B). CXCL10 belongs to the CXCR3 ligand family, which includes 2 other structurally similar chemokines, chemokine (C-X-C motif) ligand 9 (monokine induced by interferon-gamma) (CXCL9 [Mig]) and chemokine (C-X-C motif) ligand 11 (interferon-inducible T cell alpha chemoattractant) (CXCL11 [I TAC], that target activated T helper cell type 1 (Th1) natural killer cells. However, induction of neither CXCL9 nor CXCL11 could be detected in IR livers (Fig. 1C). Thus, IR triggers a rapid and specific up-regulation of CXCL10 in the liver.
CXCL10 Is Critical for the Development of Liver IRI.
To test the functional significance of CXCL10, mice deficient in CXCR3 ligand were subjected to liver IRI. Livers were harvested at 6, 24, and 48 hours after reperfusion. Unlike WT counterparts, CXCL10 KO mice were protected from IRI. Indeed, hematoxylin-eosin staining of IR-liver lobes showed near normal histology, with preservation of lobular architecture and absence of large necrotic areas/sinusoidal congestion (Fig. 2). Unlike in WT, the early signs of hepatocellular injury in CXCL10 KO mice never progressed into zonal and panlobular necrosis. Moreover, sALT levels remained lower in CXCL10 KO as compared with WT (Fig. 3; WT = 2772 ± 874.9 versus KO = 352.7 ± 71.61, P < 0.03). Thus, CXCL10 signaling is required for the development of liver IRI.
IR Alters Inflammation Response in CXCL10–Deficient Livers.
To analyze the mechanism of cytoprotection in the absence of CXCL10 signaling, we contrasted IR-induced intrahepatic inflammation in WT and CXCL10 KO mice. Because the induction of chemokine/cytokine gene programs varies, the ischemic livers were harvested serially and subjected to quantitative reverse transcription polymerase chain reaction. Tumor necrosis factor alpha (TNF-α), interleukin-1β (IL-1β), interleukin-6 (IL-6), interleukin-12β (IL-12β), inducible nitric oxide synthase, and interleukin-10 (IL-10) have all been implicated in the pathogenesis of liver IRI with distinctive roles. Indeed, these genes were readily induced by IR (Fig. 4) and peaked at either 2 hours (TNF-α IL-12β, IL-10) or 6 hours (IL-1β, inducible nitric oxide synthase) or remained up-regulated throughout (IL-6). Liver IR in CXCL10 KO mice failed to induce TNF-α, IL-1β, or IL-6 at all time points studied but kept IL-10 induction relatively intact (Fig. 4). The early IL-12β induction (1 hour) in KO livers diminished thereafter. In contrast, inducible nitric oxide synthase induction was reduced early at reperfusion but recovered thereafter. Thus, CXCL10 KO mice elicited an altered inflammation response, with reduced pro-inflammatory gene, and concomitant preserved anti-inflammatory IL-10 formation.
In addition to reduced liver inflammation, we measured IR-induced liver neutrophil infiltration at 6 hours by immunohistology (anti-Ly-6G) and esterase staining (Naphthol). Significantly decreased numbers of activated neutrophils were detected by both methods in CXCL10 KO livers, as compared with WT counterparts (Fig. 5, positive cells/HPF by immunohistology, 2.6 ± 1.3 versus 0.3 ± 0.6, P < 0.05). The induction of chemokines associated with neutrophil function, including macrophage inflammatory protein 2 (MIP-2) and Kupffer cells, was also diminished in CXCL10 KO livers (data not shown), indicating that neutrophil infiltration/activation was mediated, at least in part, by CXCL10.
CXCL10 KO Mice Respond to LPS.
To determine whether reduced liver pro-inflammatory IR response in CXCL10 KO mice resulted from intrinsic defective innate immune function, LPS (stimulates TLR4 on Kupffer cells) was administered into WT and CXCL10 KO mice. Similar degree of LPS-induced pro-inflammatory as well as anti-inflammatory gene induction was observed in both animal groups, as evidenced by up-regulated liver TNF-α, IL-1β, IL-6, IL-12β, and IL-10 expression (Fig. 6). Furthermore, parallel cultures of bone marrow–derived macrophages from WT and CXCL10 KO mice have revealed comparable in vitro responses to LPS stimulation (not shown). Thus, CXCL10 KO mice were capable of mounting pro-inflammatory responses in vivo and in vitro, and their failure to do so in response to liver IR insult was not attributable to putative intrinsic defects in their innate immune system.
Selective Reduction of IR-Induced Liver ERK Activation in CXCL10 KO Mice.
To further dissect the mechanism of altered liver immune response to IR in the absence of CXCL10, we analyzed the activation of inflammation-associated MAP kinases, including ERK, JNK, and p38, which have been implicated in various signaling pathways leading to pro-inflammatory gene up-regulation.20 Liver IR induces activation of ERK and JNK, but not p38, as shown by increased phosphorylated ERK and JNK protein levels (Fig. 7A). Interestingly, activation profiles of MAP kinases in livers were different, because the phosphorylated ERK level initially reduced after ischemia (0 hours) did increase at 30 minutes to 2 hours post-reperfusion, whereas the phosphorylated JNK increased right after ischemia, peaked at 30 minutes after reperfusion and became undetectable by 1 hour. Thus, the absence of CXCL10 had a selective impact on the ERK, but not JNK activation, providing evidence that IR altered immune response in CXCL10 KO livers.
IL-10 Neutralization Restores Liver Inflammation and IRI in CXCL10KO Mice.
Because the liver inflammatory response to IR in CXCL10 KO mice was characterized by intact IL-10 induction, the question of whether IL-10 played a cytoprotective role was addressed by neutralizing IL-10 at the onset of reperfusion. Significant increase in liver injury was observed after treatment of CXCL10 KO mice with anti-IL-10 mAb, as shown by both sALT levels (Fig. 8A, 1101 ± 222 in KO, versus 5618 ± 1570 in KO/mAb; P < 0.05), and liver histology. Livers in CXCL10 KO mice in which IL-10 has been neutralized showed both zonal and panlobular necrosis in parenchyma, comparable with WT (Fig. 2). In parallel, IR inflammation was restored, as evidenced by increased TNF-α, IL-1β, and IL-6 levels (Fig. 8b). Furthermore, IR-triggered liver ERK activation was also restored in CXCL10 mice treated with anti-IL-10 mAb (Fig. 7B). Thus, by re-creating the pro-inflammatory response, IL-10 neutralization rendered CXCL10 KO mice susceptible to IR-induced hepatocellular injury.
Although up-regulation of CXCL10 during IR has been reported in animal models and humans,12, 16, 17, 21 it has never been singled out in a strict functional IRI setting. Our data show that the absence of CXCL10 protected livers from injury by reducing liver inflammation, skewing the local response against IR from pro-inflammatory to anti-inflammatory, coupled with decreased neutrophil infiltration and activation. This altered inflammation response was specific to IR insult, because LPS stimulation was equally effective in eliciting pro-inflammatory responses in vivo in livers and in vitro in primary macrophage cultures from WT or CXCL10 KO mice. Additionally, this alteration was reversible, because IL-10 neutralization restored liver inflammation/RI in CXCL10 KO mice. These data indicate that CXCL10 may regulate response against liver IR by promoting pro-inflammatory gene induction, and the balance between CXCL10 and IL-10 determines the nature of response to IR insult and the development of hepatocellular injury. The functional role of CXCL10 was in the acute (within 6 hours) rather than the chronic disease phase during which de novo lymphocyte chemotaxis was not significant.
CXCL10 plays a key role in initiating acute allograft rejection.12, 22 By using gene KO donors, its receptor CXCR3 KO recipients, or CXCL10 neutralizing/CXCR3 blocking antibodies, cardiac allograft survival was prolonged, as compared with WT/untreated counterparts.12, 22 Reduced intragraft pro-inflammatory cytokine production/lymphocyte infiltration accompanied the protective effect of the CXCL10 signaling blockade. Other IFN-inducible chemokines, including Mig or iTAC, were also induced early posttransplantation.23–25 In particular, Mig was the dominant factor directing T cells into rejecting grafts.26 The role of CXCR3 and its ligands has also been reported in skin and small bowel transplant models, diabetic insulitis, adjuvant-induced peritonitis, and delayed type hypersensitivity (DTH) response.14, 27–29 Most of these studies focused on Ag-specific T lymphocyte chemotaxis occurring days after the disease process started. In our study, conversely, the syngeneic setting without exogenous antigen precluded the full activation of adaptive immune T/B cells. Moreover, most of our end-points were within 6 hours of the disease process, which makes it unlikely that the reduced de novo CXCR3+ T cell infiltration was responsible for diminished liver inflammation. Indeed, we found constant hepatic CD3 levels throughout the reperfusion period and between WT and CXCL10 KO mice (data not shown). Thus, the novel finding in our study is that CXCL10 deficiency does impair liver inflammation responses in the absence of exogenous antigen stimulation.
In the absence of CXCL10 signaling, although liver inflammation during IR was diminished, the IL-10 formation remained intact, a clear indication of an altered response. Moreover, reduced ERK, but intact JNK activation were observed in CXCL10 KO livers, providing further evidence for an altered rather than generally suppressed response. These findings are also indicative of a possible differential role of MAP kinase activation associated with liver inflammation/IRI. Although JNK activation has been associated with liver IRI, its role in liver cell death or tissue inflammation remains controversial.30, 31 The ability of IR to induce JNK activation in CXCL10 KO mice indicates that it associates rather with the triggering event, that is, the initial ischemia-induced hepatocellular injury than the pro-inflammatory response per se. Conversely, the suppression of ERK activation suggests its role in liver inflammation further downstream IR cascade, that is, the inflammation-induced hepatocellular damage. Furthermore, our data suggest that CXCL10 may regulate the threshold of inflammatory response in livers counteracting with IL-10, which induction was independent of CXCL10. Although IL-10, a known Th2-type immunosuppressive and anti-inflammatory cytokine,32 plays protective role in IRI,33, 34 its relationship with pro-inflammatory cytokine programs has not been fully elucidated. Our analysis of liver cytokine profiles in WT and CXCL10-deficient mice indicates distinctive mechanisms regulating the 2 sets of genes. Indeed, the induction of pro-inflammatory genes, but not IL-10, required CXCL10. Our data of restored liver inflammation by IL-10 neutralization indicate no endogenous defects in the ability of CXCL10 KO mice to generate inflammation response, consistent with our hypothesis that immune regulation is the key cytoprotective mechanism in livers deficient of CXCL10 signaling. Moreover, we have shown that treatment with anti–IL-10 mAb further deteriorated the hepatocellular damage in WT recipients.35
CXCL10 KO mice are born without phenotypical defects in the immune system, as evidenced by normal lymphocyte numbers, phenotypes, and subset distribution.36 However, they do exhibit impaired T cell responses, with reduced proliferation and interferon gamma (IFN-γ) secretion against antigen stimulation.36 Although it has been established that T cells, particularly CD4+ subset, play a key role in IRI,18, 37 how T cells function in this innate immune-dominated response remains unclear. We observed in rat orthotopic syngeneic liver transplants that CXCR3+ T cells are critical in executing innate immune function during IRI,38 data supported by a mouse renal study.39 Our results further indicate that CXCL10 and its interaction with CXCR3+ cells may be the key link for T cell recruitment and function in IRI activation. We found that intrahepatic T cells are enriched with CXCR3+ T cells, particularly the CD4 subset, even before IR onset without any injury or stimulation. Because activated T cells, particularly type I CD4+ helpers, express CXCR3, it is possible that they may be able to respond to CXCL10 without further antigen stimulation by releasing cytokines/cytotoxic molecules, and promoting liver innate activation, followed by hepatocyte damage. Indeed, it has been found that CXCL10 enhances T lymphocyte proliferation and effector cytokine production in vitro.40
In summary, our study documents the critical role of CXCL10 signaling in liver inflammation and pathophysiology of IRI. Our data unravel a novel cytoprotective mechanism based on the balance between CXCL10 and IL-10. The role of CXCL10 in mediating adaptive T cell function in TLR4-mediated activation provides a clue for further exploration of the elusive question as to how T cell functions in IRI in the absence of exogenous antigen. Targeting this chemokine may provide a novel means to prevent or treat IRI and thus maximize organ donor pool through the safer use of liver transplants despite prolonged periods of ischemia.