Potential conflict of interest: Nothing to report.
This work was supported by the National Institutes of Health (grant nos.: DK56029 and AG025881; to A.B.L.).
Previous studies have demonstrated the significance of signaling through the CXC chemokine receptor-2 (CXCR2) receptor in the process of recovery and regeneration of functional liver mass after hepatic ischemia/reperfusion (I/R). CXCR2 is constitutively expressed on both neutrophils and hepatocytes; however, the cell-specific roles of this receptor are unknown. In the present study, chimeric mice were created through bone marrow transplantation (BMT) using wild-type and CXCR2-knockout mice, yielding selective expression of CXCR2 on hepatocytes (Hep) and/or myeloid cells (My) in the following combinations: Hep+/My+; Hep−/My+; Hep+/My−; and Hep−/My−. These tools allowed us to assess the contributions of myeloid and hepatocyte CXCR2 in the recovery of the liver after I/R injury. Flow cytometry confirmed the adoption of the donor phenotype in neutrophils. Interestingly, Kupffer cells from all chimeras lacked CXCR2 expression. Recovery/regeneration of hepatic parenchyma was assessed by histologic assessment and measurement of hepatocyte proliferation. CXCR2Hep+/My+ mice showed the least amount of liver recovery and hepatocyte proliferation, whereas CXCR2Hep−/My− mice had the greatest liver recovery and hepatocyte proliferation. CXCR2Hep+/My− mice had enhanced liver recovery, with hepatocyte proliferation similar to CXCR2Hep−/My− mice. Myeloid expression of CXCR2 directly regulated CXC chemokine expression levels after hepatic I/R, such that mice lacking myeloid CXCR2 had markedly increased chemokine expression, compared with mice expressing CXCR2 on myeloid cells. Conclusion: The data suggest that CXCR2 on myeloid cells is the predominant regulator of liver recovery and regeneration after I/R injury, whereas hepatocyte CXCR2 plays a minor, secondary role. These findings suggest that myeloid cell-directed therapy may significantly affect liver regeneration after liver resection or transplantation. (HEPATOLOGY 2013)
Hepatic ischemia/reperfusion (I/R) injury is a cause of significant morbidity and mortality after liver transplantation, hemorrhagic shock, and extended liver resection for cancer.1-4 I/R injury is a biphasic response comprised of an acute injury phase resulting from oxidative stress and a later phase resulting from an intense inflammatory response that culminates in the hepatic recruitment of neutrophils and subsequent neutrophil-dependent parenchymal injury.5-10 An emerging interest has been focused on the processes of liver recovery, repair, and regeneration after I/R injury, which are highly relevant to transplantation and shock.
Our laboratory has previously characterized the recovery, repair, and regenerative responses after liver I/R injury in a murine model. The reparative process spans several days and involves the clearance of dead tissue and the replacement of functional liver mass by hepatocyte proliferation.11-13 Interestingly, we recently discovered that CXC chemokines that contain the amino-terminal ELR motif play an important role in the recovery, repair, and regeneration of the liver after I/R.12-14 These same chemokines are potent neutrophil chemoattractants and have been shown, by us and others, to contribute significantly to the inflammatory response induced by liver I/R.15, 16 CXC chemokines bind to the receptors, CXC chemokine receptor 1 and 2 (CXCR1 and CXCR2), which are expressed on a variety of cell types, including myeloid cells, endothelial cells, and hepatocytes.17, 18 In neutrophils, these receptors are responsible for cell priming, chemotaxis, and induction of respiratory burst.5, 19, 20 In hepatocytes, we have shown that these receptors can regulate cell proliferation and cell death.12, 13 Our studies with knockout (KO) mice for CXCR1 or CXCR2 demonstrate that signaling through CXCR2 is a major regulatory component of the reparative and regenerative response, whereas CXCR1 appears to play a minor, counterbalancing role.12, 13
Despite our current understanding of the importance of signaling through CXCR2 during the recovery response to hepatic I/R, the cell-specific roles for CXCR2 in this response are unknown. Therefore, in the current study, we employed bone marrow transplantation (BMT) to create chimeric mice to allow us to examine the functional roles of hepatocyte versus myeloid expression of CXCR2 in the reparative and regenerative response after I/R injury.
This project was approved by the University of Cincinnati Animal Care and Use Committee (Cincinnati, OH) and was in compliance with the National Institutes of Health guidelines. Male BALB/c and CXCR2−/− (BALB/c background) were obtained from The Jackson Laboratory (Bar Harbor, ME). Chimeric mice were generated using BMT. For all experiments, mice were exposed to whole-body irradiation at 6-8 weeks of age. Mice received 9.5 Gy of gamma-radiation, given as a split dose of 4.75 and 4.75 Gy, separated by 3 hours. Immediately thereafter, 10 × 106 cells of wild-type (WT) or CXCR2−/− bone marrow was injected into the tail vein. These experiments generated chimeric mice with hepatocytes (Hep) and myeloid cells (My) either positive or negative for CXCR2 expression in the following combinations: Hep+/My+; Hep−/My+; Hep+/My−; and Hep−/My−. Mice were then kept in a pathogen-free environment with ad libitum access to food and water and a 12-hour light/dark cycle for a total of 4 weeks before undergoing procedures.
Hepatic I/R Model.
Partial hepatic ischemia was induced as previously described.15 All mice underwent either sham or I/R operation. Briefly, mice were anesthetized using sodium pentobarbital (65 mg/kg). A midline laparotomy was performed and an atraumatic vascular occlusion clip was used to interrupt blood supply to the left lateral and median lobes of the liver. The caudal lobes retained intact portal and arterial inflow and venous outflow, preventing intestinal venous congestion. After 90 minutes of partial hepatic ischemia, the clip was removed to initiate hepatic reperfusion. Sham control mice underwent the same protocol without vascular occlusion. Mice were sacrificed after the indicated periods of reperfusion, and blood and liver samples were taken for analysis.
Blood and Tissue Analysis.
Blood was obtained by intracardiac puncture at time of sacrifice. Serum was then analyzed for alanine aminotransferase (ALT) as a measure of hepatocellular injury. Measurements of serum ALT were made using a diagnosis kit by bioassay (Wiener Laboratories, Rosario, Argentina). Serum levels of tumor necrosis factor alpha (TNF-α), keratinocyte chemokine (KC), and macrophage inflammatory protein-2 (MIP-2) were measured by enzyme-linked immunosorbent assay (ELISA), in accord with the manufacturer's instructions (R&D Systems, Minneapolis, MN). Hepatic tissue was harvested at time of sacrifice for measurement of neutrophil accumulation and histologic analysis. Liver neutrophil accumulation was determined by myeloperoxidase (MPO) content. MPO content was assessed using methods described elsewhere.21 For histologic analysis, tissue samples were fixed in 10% neutral-buffered formalin (Richard Allen Scientific, Kalamazoo, MI), processed, and embedded in paraffin before staining with hematoxylin and eosin (H&E). Quantitative morphometric analysis of hepatocellular necrosis was performed in a blinded fashion with histologic sections at low power (×10) using Photoshop image analysis software (Adobe Systems, Inc., San Jose, CA). Necrotic area was expressed as percentage of total area examined.
Proliferating Cell Nuclear Antigen Staining.
Immunohistochemical (IHC) staining for proliferating cell nuclear antigen (PCNA) was performed on paraffin-embedded liver tissue with anti-PCNA antibody (Ab) using a Dako ARK Peroxidase kit (Dako, Copenhagen, Denmark). Briefly, a three-step peroxidase method was performed according to the manufacturer's instructions. PC-10 monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was used at a dilution of 1:50 for 15 minutes at room temperature. Sections were counterstained with hematoxylin. Evaluation of PC-10 immunostaining was performed based on the percentage of positive nuclei of 400-600 hepatocytes from four to six highest positive fields at high power (×400) and was expressed as the PCNA labeling index.
Isolation and Flow Cytometric Analysis of Neutrophils and Kupffer Cells.
Analysis of blood leukocytes was conducted similarly as previously described.22 First, 50-100 uL of whole blood were added to glass tubes with 4 mL of ACK lysing buffer and were mixed by inversion. Samples were incubated at room temperature for 5 minutes and then spun down for 5 minutes at 400×g. Supernatant was decanted and 0.5 mL of fluorescence-activated cell sorting (FACS) buffer was used as a wash, followed by centrifugation at 400×g for 5 minutes. Cells (0.5-1.5 × 106/tube) were then placed in polystyrene tubes containing 1 μg of anti-CD16/32 and 5% rat serum to block nonspecific Ab labeling. Alexa Fluor-700 rat antimouse CD11b (Clone: M1/70; BD Biosciences, San Diego, CA), fluorescein isothiocyanate (FITC) rat antimouse Ly6-G (Clone: 1A8; BD Biosciences), and allophycocyanin (APC) rat immunoglobulin G (IgG)2A antimouse CXCR2 (Lot: LMC0711011; R&D Systems) or APC-conjugated rat antimouse IgG2A isotype control (Lot: KEU1111011; R&D Systems) were used as Abs. Labeling antibodies (0.5-1 μg of Ab/tube) were added and incubated at 4OC for 20 minutes. Cells were washed two times with cold FACS buffer (phosphate-bufferd saline [PBS] with 1% bovine serum albumin [BSA] and 0.1% azide) and fixed with 250 μL of 1% paraformaldehyde (PFA). Samples were analyzed with an LSR II flow cytometer and FACS Diva software (BD Biosciences, Mountain View, CA).
For the isolation of Kupffer cells, chimeric mice were anesthetized with sodium pentobarbital (65 mg/kg), and a midline laparotomy was performed. The portal vein was cannulated using a 22-G angiogath (Abbocath-T; Hospira Venisystems, Lake Forest, IL), and the liver was flushed with 10 mL of PBS. Liver leukocytes were isolated using a gentleMACS dissociator (Miltenyi Biotec, Auburn, CA), and the instructions were provided by the manufacturer. Isolated cells were resuspended in 7 mL of Hank's balanced salt solution (HBSS) and carefully layered on 7 mL of Lympholyte M (Cedarlane Laboratories, Hornby, Ontario, Canada). Tubes were spun at 1,500×g for 20 minutes at 25°C. The cell layer was collected and placed into FACS tubes with HBSS and spun down at 450×g for 5 minutes at room temperature. Analysis of cell-surface antigen expression in situ was performed as described elsewhere.23 Phycoerythrin-conjugated anti-CD68 (Clone: FA-11; BioLegend, San Diego, CA), Pacific Blue antimouse F4/80 (Clone: BM8; BioLegend, San Diego, CA), Alexa Fluor-700 rat antimouse CD11b (Clone: M1/70; BD Biosciences, FITC rat antimouse Ly-6G (Clone: 1A8; BD Biosciences), and APC rat IgG2A antimouse CXCR2 (Lot: LMC0711011; R&D Systems) or APC-conjugated rat antimouse IgG2A isotype control (Lot: KEU1111011; R&D Systems) were purchased for cell staining. Cells (0.5-1.5 × 106/tube) were then placed in polystyrene tubes containing 1 μg of anti-CD16/32 and 5% rat serum to block nonspecific Ab labeling. Labeling Abs (0.5-1.0 μg of Ab/tube) were then added and incubated at 4OC for 20 minutes. Cells were washed two times with cold FACS buffer (PBS with 1% BSA and 0.1% azide) and fixed with 250 μL of 1% PFA. Samples were analyzed with an LSR II flow cytometer and FACS Diva software (BD Biosciences). For some experiments, isolated Kupffer cells were plated and treated with TNF-α (10 ng/mL) or lipopolysaccharide (LPS; 100 ng/mL) for 30 minutes before staining and FACS analysis for CXCR2 expression.
All data are expressed as the mean ± standard error of the mean (SEM). Data were analyzed with a one-way analysis of variance, with a subsequent Student t test. Differences were considered significant when P < 0.05.
BMT Effectively Modulates Neutrophil Expression of CXCR2, but Is Associated With Significant Pathology in CXCR2−/− Mice.
The role of neutrophils in the pathogenesis of hepatic I/R injury is well known.5, 19 We have shown that CXCR2 is important for acute neutrophil recruitment, and that mice nullizygous for CXCR2 have a lesser degree of hepatic injury acutely after I/R than WT mice.12 CXCR2 chimeric mice (CXCR2Hep+/My+, CXCR2Hep−/My+, CXCR2Hep−/My−, and CXCR2Hep+/My−) were generated as described above. Mice receiving bone marrow from WT mice (CXCR2Hep+/My+ and CXCR2Hep−/My+) retained expression of CXCR2 on neutrophils, whereas mice receiving bone marrow from CXCR2−/− mice (CXCR2Hep−/My− and CXCR2Hep+/My−) lacked neutrophil CXCR2 expression (Fig. 1A). Expression of CXCR2 on Kupffer cells has not been previously studied. Examination of liver Kupffer cells from all chimeric phenotypes as well as healthy WT (BALB/c) and CXCR2−/− mice demonstrated no expression of CXCR2 (Fig. 1B). To determine whether CXCR2 expression might be inducible, Kupffer cells were isolated and stimulated with either LPS (100 ng/mL) or TNF-α (10ng/mL) for 30 minutes. Surface expression of CXCR2 was determined by flow cytometry. Stimulation of Kupffer cells did not induce expression of CXCR2 (data not shown).
Although BMT yielded the appropriate myeloid and hepatocyte expression phenotypes, some chimeric mice displayed pathologies that precluded their use in I/R experiments. Survival of the different chimeric mice after BMT was not significantly different among the different groups (Fig. 2A). At the time of laparotomy for hepatic I/R, a number of pathologies were noted in the different chimeras that prevented their use in experiments (Fig. 2B). These pathologies and their frequencies are shown in Table 1. A low frequency of hepatic lesion and infection was observed in CXCR2Hep+/My+, CXCR2Hep−/My+, and CXCR2Hep+/My− (Table 1). In contrast, CXCR2Hep−/My− mice had a high frequency of splenomegaly and hepatic lesions that limited their use in I/R experiments (Table 1). Splenomegaly was observed in 24% of CXCR2Hep−/My− mice (Fig. 3A). In those animals, closer examination of splenic histology revealed extensive hematopoiesis (Fig. 3B). Hepatic lesions were observed in 30% of CXCR2Hep−/My− mice. These lesions were found to be focal areas of organizing necrosis (Fig. 3C). The incidence and frequency of splenomegaly and hepatic lesions we observed in CXCR2Hep−/My− mice are consistent with the CXCR2-KO phenotype.24
Table 1. Pathologic Findings at Time of Laparotomy in Chimeric Mice
Hep+/My+ (n = 41)
Hep−/My+ (n = 46)
Ventral neck lesion, n = 1 (0.02%)
Hepatic lesion, n = 1 (0.02%)
Hep+/My− (n = 36)
Abdominal wall infection, n = 1 (0.03%)
Hep−/My− (n = 33)
Massive splenomegaly, n = 8 (24%)
Hepatic lesion, n = 10 (30%)
Liver Injury and Recovery After I/R Is Impacted by Both Myeloid and Hepatocyte CXCR2.
Our previous studies using global KO mice for CXCR2 suggested that CXCR2 expression on hepatocytes is detrimental to liver recovery after I/R.12 However, these studies were unable to directly assess the function of CXCR2 on different cell types in vivo. In vitro studies demonstrated that hepatocytes exposed to high concentrations of CXC chemokines had increased cytotoxicity.12 Using chimeric mice, we assessed the effect of cell-specific CXCR2 expression on the recovery response after I/R. Seven days after I/R, we examined liver histology, necrotic area, and serum ALT as indices of remaining liver injury and extent of repair and recovery. In CXCR2Hep+/My+ mice, necrotic liver was prominent, with little recovery observed (Fig. 4). These mice also had the highest level of circulating ALT (Fig. 4). CXCR2Hep−/My+ mice had a similar degree of liver necrosis and similar ALT levels to CXCR2Hep+/My+ mice (Fig. 4). In contrast, CXCR2Hep+/My− mice had reduced hepatocellular necrosis, compared with CXCR2Hep+/My+ mice, and had significantly lower ALT levels than both CXCR2Hep+/My+ and CXCR2Hep−/My+ mice (Fig. 4). CXCR2Hep−/My− mice had markedly less hepatocellular necrosis than all other chimeric groups and had ALT levels that were significantly lower than both CXCR2Hep+/My+ and CXCR2Hep−/My+ mice (Fig. 4).
Myeloid, but Not Hepatocyte, CXCR2 Negatively Regulates Expression of CXC Chemokines and Is Critical for Hepatocyte Proliferation After I/R.
Our previous studies have demonstrated that CXCR2-KO mice have greatly increased expression of CXC chemokines.12 These studies also showed that CXCR2 mediated proliferative effects of CXC chemokines on hepatocytes in vitro. However, subsequent experiments demonstrated that CXC chemokines are not direct mitogens for hepatocytes and that the proliferative effects of CXC chemokines required factors in the media serum supplement.25 To determine the roles of myeloid and/or hepatocyte CXCR2 in CXC chemokine expression and hepatocyte proliferation in vivo, we examined chemokine expression and hepatocyte proliferation in CXCR2 chimeric mice. Because peak chemokine production after I/R injury occurs 8 hours after reperfusion, we used this time point for analysis of MIP-2 and KC. CXCR2Hep+/My+ and CXCRHep−/My+ mice had similar levels of MIP-2 and KC (Fig. 5). Furthermore, levels of MIP-2 and KC were consistent with expression levels previously observed in WT mice.12 Interestingly, expression of MIP-2 and KC was increased by up to 15-fold in both CXCR2Hep+/My− and CXCR2Hep−/My− mice (Fig. 5), which was similar to the increase previously reported in CXCR2-KO.12
We next evaluated the hepatic regenerative response after I/R using PCNA staining as an index of hepatocyte proliferation. CXCR2Hep+/My+ and CXCRHep−/My+ mice had low numbers of PCNA-positive hepatocytes, whereas CXCR2Hep+/My− and CXCR2Hep−/My− mice had far higher numbers of PCNA-positive hepatocytes (Fig. 6). When numbers of PCNA-positive hepatocytes were quantified, CXCR2Hep+/My− and CXCR2Hep−/My− mice demonstrated up to 4-fold increases in proliferation, compared with CXCR2Hep+/My+ and CXCRHep−/My+ mice (Fig. 6).
Our laboratory has previously demonstrated that signaling by CXCR2 is detrimental to liver recovery after hepatic I/R, and that high concentrations of CXC chemokines were cytotoxic to hepatocytes in vitro.12 However, because both myeloid cells as well as hepatocytes are known to constitutively express the CXCR2 receptor, it was previously undefined as to their individual contributions within the schema of hepatic I/R injury. By generating chimeric mice with various combinations of CXCR2 expression on hepatocyte and myeloid cells, we have advanced our understanding of the role of this receptor in recovery, repair, and regeneration processes occurring after liver I/R.
ELR-positive CXC chemokines are key to the pathophysiology of I/R injury, serving as primary chemoattractants that mediate neutrophil recruitment, resulting in subsequent neutrophil-dependent liver injury.15, 16, 26 CXC chemokines are elevated after I/R,15 reaching maximal expression acutely. Previous studies have shown that CXCR2-KO mice express far higher levels of CXC chemokines than their WT counterparts.12 These results suggested that the lack of CXCR2 receptor/ligand interaction interfered with a possible negative feedback process. Whether this negative feedback was occurring through myeloid CXCR2 signaling or hepatocyte CXCR2 signaling was unknown. Our analysis of CXC chemokine levels in CXCR2Hep−/My+ mice after I/R demonstrated levels that were comparable to those of CXCR2Hep+/My+ mice, which are the equivalent of a WT control. Conversely, in CXCR2Hep+/My− mice, there was a dramatic increase in CXC chemokine levels after I/R, similar to that observed in CXCR2Hep−/My− mice (the equivalent of a CXCR2 ABL-KO). These results indicate that the negative feedback mechanism of CXC chemokine signaling appears to be primarily driven by CXCR2 expressed on myeloid cells, and that hepatocyte CXCR2 is not involved in this regulation.
Kupffer cells have a significant role in the hepatic response to I/R. Their production of proinflammatory mediators, such as TNF-α, propagates the inflammatory response throughout the liver and induces the hepatic expression of CXC chemokines and endothelial adhesion molecules.7 In mice subjected to total-body irradiation and BMT, it was important to assess whether or not resident Kupffer cells of the liver expressed CXCR2, because these cells lie at the interface of the vasculature and parenchyma. Previous research analyzing Kupffer- cell turnover after BMT has demonstrated almost complete replacement of Kupffer cells with those expressing the bone marrow donor cell phenotype at 4 weeks.27 Interestingly, our data demonstrate that Kupffer cells from all phenotypes of chimeric mice (CXCR2Hep+/My+, CXCR2Hep−/My+, CXCR2Hep+/My−, and CXCR2Hep−/My−) lacked expression of CXCR2. In addition, Kupffer cells from healthy WT BALB/c and CXCR2−/− mice that did not undergo BMT also lacked expression of CXCR2. Furthermore, stimulation of Kupffer cells with LPS or TNF-α failed to induce expression of CXCR2. These findings demonstrate, for the first time, a lack of CXCR2 expression by Kupffer cells, and strongly suggest that the observed effects of myeloid CXCR2 deficiency in the current study is related to a lack of expression of CXCR2 on neutrophils.
Neutrophil, and not hepatocyte, CXCR2 also appears to be a primary regulator of liver recovery and regeneration after I/R injury. Seven days after injury, mice lacking myeloid CXCR2 expression had less hepatocellular injury and increased hepatocyte proliferation, compared with those retaining myeloid CXCR2. Myeloid CXCR2 appears to function primarily by contributing to hepatocellular toxicity in the presence of high levels of CXCR2 ligands found after reperfusion. This is supported by the findings that CXCR2Hep+/My− mice had less liver necrosis than CXCR2Hep+/My+ mice and far lower ALT levels than both CXCR2Hep+/My+ and CXCR2Hep−/My+ mice. The greatest repair and recovery, as determined by histological evidence of necrosis and circulating ALT, was found in CXCR2Hep−/My− mice. Hepatocyte proliferation was 3-4-fold greater in mice lacking myeloid CXCR2, suggesting that chemokine signaling in myeloid cells, most likely neutrophils, directly stimulates hepatocyte proliferation. The precise mechanism of this interaction needs to be elucidated.
In summary, the current study demonstrates that myeloid CXCR2 is the primary regulator of liver recovery and regeneration after hepatic I/R injury. Our data also show that Kupffer cells lack CXCR2 expression, suggesting that it is primarily neutrophil CXCR2 that regulates the hepatic recovery response. Hepatocyte CXCR2 appears to play a minor role in this process. CXCR2 continues to be an important therapeutic target and has relevance in a number of clinically relevant liver injuries.