Impaired liver regeneration and increased oval cell numbers following T cell–mediated hepatitis†
Article first published online: 27 JUN 2007
Copyright © 2007 American Association for the Study of Liver Diseases
Volume 46, Issue 1, pages 229–241, July 2007
How to Cite
Hines, I. N., Kremer, M., Isayama, F., Perry, A. W., Milton, R. J., Black, A. L., Byrd, C. L. and Wheeler, M. D. (2007), Impaired liver regeneration and increased oval cell numbers following T cell–mediated hepatitis. Hepatology, 46: 229–241. doi: 10.1002/hep.21674
Potential conflict of interest: Nothing to report.
- Issue published online: 27 JUN 2007
- Article first published online: 27 JUN 2007
- Manuscript Accepted: 19 JAN 2007
- Manuscript Received: 25 JUL 2006
- NIH. Grant Numbers: AA014243, T32-AA07573, F32-AA015005
The regeneration of liver tissue following transplantation is often complicated by inflammation and tissue damage induced by a number of factors, including ischemia and reperfusion injury and immune reactions to the donor tissue. The purpose of the current study is to characterize the effects of T cell–mediated hepatitis induced by concanavalin A (ConA) on the regenerative response in vivo. Liver regeneration following a partial (70%) hepatectomy (pHx) was associated with elevations in serum enzymes and the induction of key cell cycle proteins (cyclin D, cyclin E, and Stat3) and hepatocyte proliferation. The induction of T cell–mediated hepatitis 4 days before pHx increased serum enzymes 48 hours after pHx, reduced early cyclin D expression and Stat3 activation, and suppressed hepatocyte proliferation. This inhibition of proliferation was also associated with increased expression of p21, the activation of Smad2, the induction of transforming growth factor beta and interferon gamma expression, and reduced hepatic interleukin 6 production. Moreover, the ConA pretreatment increased the numbers of separate oval cell-like CD117+ cells and hematopoietic-like Sca-1+ cell populations 48 hours following pHx. The depletion of natural killer (NK) cells, an important component of the innate immune response, did not affect liver injury or ConA-induced impairment of hepatocyte proliferation but did increase the numbers of both CD117-positive and Sca-1–positive cell populations. Finally, splenocytes isolated from ConA-pretreated mice exerted cytotoxicity toward autologous bone marrow cells in an NK cell–dependent manner. Conclusion: T cell–mediated hepatitis alters early cytokine responses, reduces hepatocellular regeneration, and induces NK cell–sensitive oval cell and hematopoietic-like cell expansion following pHx. (HEPATOLOGY 2007;46:229–241.)
Organ transplantation continues to represent the primary therapy for the treatment of severe liver disease and failure.1 The shortage of donor organs has, however, limited the availability of viable organs to the large number of eligible recipients. Alternative approaches, including partial liver and split liver transplant procedures, have been explored, although the regeneration of the total liver mass is required.1 Fortunately, the liver possesses the remarkable capacity to regenerate following major tissue loss.3 A large body of evidence suggests that hepatocytes in the area nearest the hepatic central vein undergo several rounds of division to restore functional liver mass.3 Furthermore, this process occurs very rapidly, regenerating as much as 70% of the original mass within 14 days. Alternatively, when periportal hepatocytes are damaged or their division is inhibited, a small pool of hepatic progenitor cells (HPCs), also known as oval cells, expands and differentiates into mature hepatocytes and biliary epithelial cells in an attempt to restore functional tissue mass.4–9 These regenerative processes are often complicated by the coexistence of other liver pathologies such as hepatitis C virus infection or recipient-mediated autoimmune reactions to the donor tissue.10 Moreover, ischemia and reperfusion injury occurring during the storage and transplantation procedures also complicates the regenerative response.11, 12 Together, these processes result in hepatocellular inflammation, immune cell recruitment, and tissue damage.
It is becoming increasingly apparent that components of the immune system may be capable of regulating the regenerative response. Sun and Gao13 recently reported that activated natural killer (NK) cells could impair normal cell proliferation in the regenerating murine liver. This impairment in liver cell proliferation was linked to the production of interferon gamma (IFNγ), a T helper 1 (Th1)-type cytokine, by intrahepatic NK cells. Other studies have implicated natural killer T (NKT) cells in the destruction of regenerating hepatocytes.14, 15 A reduction in the hepatocellular volume significantly increased the number of NKT cells and their cytotoxicity toward syngeneic regenerating hepatocytes.14 Therefore, it appears that direct activation of the innate immune response within the liver may be capable of regulating hepatocellular regeneration. The effect that the intrahepatic immune response has on the oval cell compartment is not well understood. Sakamoto and others16 demonstrated the ability of concanavalin A (ConA), a plant lectin capable of inducing significant NKT cell–dependent hepatocellular injury, to impair the regenerative response of the liver while stimulating the proliferation of what appeared to be hematopoietic stem cells and possibly epithelial-like stem cells, hepatic oval cells. Using this immune-based model of liver injury and regeneration, one can begin to identify the effects of specific components of the immune system on the hepatocyte and oval cell regenerative responses. Therefore, the aims of the current study were to determine the effect of existing T cell–mediated hepatitis on the hepatic regenerative response, evaluate the extent of oval cell expansion in this model, and determine the role of certain innate immune cells, specifically NK cells, during this process.
Materials and Methods
C57Bl/6 male mice (6-8 weeks of age), severe combined immunodeficient mice, and green fluorescent protein (GFP) transgenic mice were purchased from the Jackson Laboratory (Bar Harbor, ME) and maintained on a standard laboratory diet with free access to food and water. All experimental procedures complied with the Guide for the Care and Use of Laboratory Animals (revised 1996) and the Institutional Animal Care and Use Committee guidelines at the University of North Carolina at Chapel Hill.
ConA Treatment and NK Cell Depletion.
Male mice were injected intravenously with either vehicle (0.9% NaCl) or 12 mg/kg ConA Type V (Sigma Chemical Corp., St. Louis, MO) 4 days before a partial hepatectomy (pHx) or sham laparotomy. To deplete NK cells, mice were administered intraperitoneally 300 μg of anti-asialo GM1 antibody (Wako Pure Chemicals, Richmond, VA) 2 days before ConA administration, a dosage that has been shown to deplete NK cells in mice for up to 7 days.17
Animal Model of pHx.
Male mice were anesthetized with 50 mg/kg sodium pentobarbital, and a 70% hepatectomy or sham surgery was performed as described18 by the removal of the median and left lateral lobes. After 0, 6, 24, or 48 hours, the mice were euthanized, and serum and tissue were collected.
Serum ALT Measurements.
Serum levels of alanine aminotransferase (ALT) were measured for all animals with a commercially available kit at an assay temperature of 37°C (Diagnostic Chemical Laboratories, Charlottetown, PA).
Histopathology and Immunohistochemistry of Liver HPC Antigens.
Formalin-fixed, paraffin-embedded tissue was permeabilized with proteinase K (20μg/ml) at 37°C for 30 min, blocked with 3% H2O2, and stained with anti-CD117 (Neomarkers, Fremont, CA), anti- alpha fetoprotein (anti-AFP) antibodies (Santa Cruz, Santa Cruz, CA), anti-pSmad 2/3 (Santa Cruz), anti-muscle pyruvate kinase (mPK; Rockland, Gilbertsville, PA), anti-pan-cytokeratin (Dako Cytomation), and anti-p21waf (Santa Cruz) at 1:100 for 1-2 hours at room temperature, and this was followed by incubation with horseradish peroxidase–linked secondary antibodies (1:250) for 1 hour at room temperature. For stem cell antigen 1 (Sca-1) staining, slides were subjected to antigen retrieval by incubation in 10 mM sodium citrate for 20 min at 99°C in lieu of proteinase K, incubated with anti-mouse Sca-1 (eBioscience, San Diego, CA) for 2 hours at 37°C, and then incubated with a secondary antibody with a VectaStain kit (Vector Laboratories). Antibody binding was visualized with diaminobenzidine as the chromogen substrate. Hepatocytes in the S phase were analyzed by proliferating cell nuclear antigen (PCNA) staining as described.19 Hepatocyte nuclei positive for PCNA, hepatocytes in mitosis, or cytokeratin (CK) positive cells were counted from 10 random high-powered fields from each animal and expressed as positive cells per 400× field.
The total cellular protein (50μg) or nuclear protein (15 μg) was separated on an 8%-16% gradient polyacrylamide gel, transferred to nitrocellulose, and blocked with 5% bovine serum albumin for 1 hour. Primary antibodies (anti-cyclin D1, Santa Cruz Biotechnology, 1:1000; anti-cyclin E1, Santa Cruz Biotechnology, 1:1000; anti-pSmad2, Chemicon, 1:2000; Smad2, Santa Cruz Biotechnology, 1:1000; anti-pStat3, Cell Signaling Technologies, 1:1000; Stat3, Santa Cruz Biotechnology, 1:2000; p21, Santa Cruz Biotechnology, 1:500; anti-β actin, Sigma, 1:5000) were incubated for 2 hours at room temperature. Horseradish peroxidase–conjugated secondary antibodies (1:2000 dilution) were visualized by enhanced chemiluminescence. The band density was quantified with Image J software (http://rsb.info.nih.gov/ij/) and normalized to β actin expression or compared to Coomassie-stained gels. The total cellular protein was also evaluated for interleukin 6 (IL6) with an enzyme-linked immonosorbency assay with a commercially available kit from R&D Systems (Minneapolis, MN).
Real-Time Polymerase Chain Reaction.
The total RNA (5 μg) isolated with Trizol reagent (Invitrogen, Carlsbad, CA) was reverse-transcribed, and 500 ng of cDNA was subjected to real-time PCR with specific primers listed in Table 1 using a Sybr green kit from Applied Biosystems (Foster City, CA). The expression is presented as a fold change versus a vehicle-treated control with the comparative ct (cycle times) method with β-actin as the internal control.
|Gene name||Primer Sequence|
|Tumor necrosis factor α||For 5′-AGCCCACGTAGCAAACCACCAA-3′|
|Interleukin 6 (IL6)||For 5′-GAGGATACCACTCCCAACAGACC-3′|
|Interferon γ (IFNγ)||For 5′-TCAAGTGGCATAGATGTGGAAGAA-3′|
|Transforming growth factor β||For 5′-TGACGTCACTGGAGTTGTACGG-3′|
|β actin||For 5′-AGGTGTGCACCTTTTATTGGTCTCAA-3′|
Nonparenchymal Cell Isolation and Characterization.
Nonparenchymal cells were isolated from fresh liver tissue by mechanical disruption and centrifugation at 500g for 5 minutes. Approximately 1 × 106 nonparenchymal cells were stained with anti-mouse CD117 (PE-Cy5) or Sca-1 [phycoerythrin (PE)] antibodies (eBioscience) at a dilution of 1:100 for 30 minutes at room temperature and analyzed on a FACS Scan instrument (Becton Dickinson, Franklin Lakes, NJ) with Summit software (Dako Cytomation, Carpinteria, CA).
Furthermore, CD117 and Sca-1 positive cell populations were enriched by magnetic cell sorting using the AutoMACS system and bead-labeled antibodies from Miltenyi (Auburn, CA) and stained with anti-CD117 (PE-Cy5) or anti-Sca-1 (PE) antibodies (eBioscience). Sorted cells were fixed, and permeabilized cells were then stained with anti-albumin (unconjugated, ICN Biomedicals, Aurora, OH), anti-alpha-fetoprotein (unconjugated, Santa Cruz), anti-pan-cytokeratin (unconjugated, Dako Cytomation), anti-CD45 (FITC-conjugated, eBioscience), or anti-major histocompatability complex I (FITC-conjugated, eBioscience). Unconjugated antibodies were stained with FITC-conjugated secondary antibodies (anti-rabbit, Chemicon International) before analysis on a Becton Dickinson FACS Scan.
In Vitro Cytotoxicity Assay.
To determine the cytotoxicity of NK cells toward autologous bone marrow cells, total splenocytes were isolated from control mice, mice pretreated with ConA for 24 hours, or NK cell–depleted mice treated with ConA for 24 hours. Bone marrow cells were isolated from both femurs of two GFP transgenic mice via flushing with cold phosphate-buffered saline through a 25-gauge needle. For the cytotoxicity assay, 1 × 105 bone marrow cells (target cells) were incubated with 0, 1 × 106, or 2.5 × 106 total splenocytes at 37°C in 5% CO2 for 6 hours. Cells were then incubated with 7-AAD (7-amino actinomycin D), a membrane impermeant molecule. GFP-positive cells were then assessed by flow cytometry for the presence or absence of 7-AAD staining by flow cytometry. All experiments were performed in duplicate. Data are presented as percentages of the total GFP cells positive for 7-AAD.
All data are presented as the mean plus or minus the standard error of the mean for 4 or more animals per group. Differences between groups were determined by a nonparametric Student t test using InStat software (San Diego, CA), for which the significance was set at P < 0.05.
ConA Activates T Lymphocytes and Depletes NKT Cells in the Murine Liver.
ConA represents a well-established model of T cell–dependent autoimmune liver disease. Following ConA administration, hepatic T lymphocytes are rapidly activated as demonstrated by the expression of CD69 on CD4+ lymphocytes (Fig. 1A). In addition, hepatic CD3+ NK1.1+ NKT cells are rapidly depleted (Fig. 1B). Four days (96 hours) after ConA administration, however, NKT cells can be shown to repopulate the hepatic parenchyma (Fig. 1B). Moreover, ConA induces T cell accumulation 4 days after ConA administration (Fig. 1C). Together, these data demonstrate the persistence of T cell activation, the accumulation of CD3+ lymphocytes, and the repopulation of NKT cells 4 days after ConA injection, the time point chosen for pHx.
T Cell–Mediated Hepatitis Reduces Hepatocellular Proliferation Following pHx.
A number of studies have demonstrated the rapid regeneration of the murine liver following simple pHx.3 Here, we demonstrate similar robust hepatocyte proliferation following pHx (Fig. 2B). An analysis of liver cell proliferation by nuclear PCNA expression in hepatocytes revealed significant hepatocyte proliferation 48 hours following pHx in vehicle-treated controls. The pretreatment of mice with ConA 4 days before pHx significantly reduced (∼40%) the peak hepatocyte proliferative response at 48 hours after pHx but caused a small but significant increase in PCNA+ cells at 6 hours after pHx (Fig. 2B,C). The evaluation of mitotic figures confirmed the PCNA data with significantly reduced hepatocellular mitosis in ConA-pretreated livers in comparison with vehicle-treated mice 48 hours after pHx (Fig. 2D). pHx also resulted in a significant increase in liver enzyme release at 6, 24, and 48 hours after ligation (Fig. 2E). The treatment of mice with ConA 4 days before pHx increased liver injury above vehicle-treated controls only at 48 hours after pHx (Fig. 2E). To further evaluate the role of unactivated lymphocytes in the regenerative response, wild-type or severe combined immunodeficient mice were subjected to simple pHx. The absence of lymphocytes did not accelerate liver regeneration within the murine liver and, in fact, led to a small reduction in hepatocellular proliferation (data not shown). Together, these data demonstrate the importance of activated, but not quiescent, lymphocytes in the inhibition of the regenerative response.
Exploration into the potential mechanisms by which hepatocellular proliferation was impaired revealed a substantial reduction in cellular cyclin D1 expression 6 and 24 hours after pHx and cyclin E nuclear localization at 48 hours, but not 24 hours, after pHx in ConA-pretreated mice in comparison with their vehicle-treated controls (Fig. 3A,B). Moreover, phosphorylation and nuclear localization of Stat3, a key regulator of hepatocellular proliferation, were dramatically reduced in ConA-pretreated mice at 6, 24, and 48 hours after pHx in comparison with their vehicle-treated controls. Furthermore, nuclear localization of p21waf, an inhibitor of cyclin dependent kinases, was also increased in ConA-pretreated mice 6 hours after a hepatectomy in comparison with vehicle-treated controls (Fig. 3D,E). Interestingly, cyclin E was increased at 24 hours after pHx in ConA-pretreated mice in comparison with vehicle-treated controls. This increase in expression was not associated with increased cellular hepatocellular proliferation. Reduced cyclin expression, Stat activation, and inhibition of cyclin dependent kinase activity through increased p21waf expression provide potential mechanisms by which T cell–mediated hepatitis either directly or indirectly influences the regenerative response of the liver.
ConA Alters the Cytokine Response Within the Regenerating Liver.
Cytokines including TNFα and IL6 play key roles in hepatocellular regeneration. T cell–mediated liver injury induced by ConA also involves the up-regulation of several cytokines (TNFα, IFNγ, and interleukin 4) that serve to activate and recruit inflammatory cells within the liver. The current model is designed to test the effects of T cell–mediated liver damage during liver regeneration, a scenario in which overlapping cytokine signals may be present. As shown in Fig. 4A, tissue IL6 protein levels are significantly increased 6 and 24 hours after pHx in vehicle-treated mice, whereas the ConA pretreatment significantly blunts this response. Likewise, TNFα expression is increased within the regenerating liver from 0 to 48 hours after pHx (Fig. 4B). Pretreatment with ConA disrupts this response, resulting in a substantial elevation of TNFα expression at 0 hours (time of pHx) and at 24 and 48 hours after pHx, although the levels do not reach statistical significance. In contrast, the expression of transforming growth factor beta 1 (TGFβ1), a cytokine associated with the inhibition of hepatocyte proliferation, dropped from 24 to 48 hours after pHx in vehicle-treated mice (Fig. 4C). In ConA-pretreated mice, TGFβ expression was elevated at 0 hours (time of pHx) and decreased over the period of regeneration (from 24 to 48 hours after pHx), although the levels remained significantly higher than those in vehicle-pretreated mice. Finally, the expression of IFNγ, a known mediator of ConA-dependent liver injury20 and a recently reported positive regulator of the oval cell response,21 was similar in vehicle- and ConA-pretreated mice at 0 and 6 hours after the hepatectomy but did significantly increase by 48 hours after the hepatectomy (Fig. 4D). Together, these data demonstrate the ability of pre-existing T cell–mediated hepatitis to significantly alter important mitogenic and proliferation-inhibitory cytokines.
TGFβ acting through Smad signaling plays an important role in liver growth and regeneration.22, 23 To determine the significance of the early enhancements in TGFβ production, an analysis of downstream Smad activation was assessed. As shown in Figure 4E, T cell–mediated hepatitis induces a significant increase in Smad-2/3 phosphorylation at 6 hours after a hepatectomy in hepatocytes in comparison with vehicle-treated controls. An analysis of nuclear Smad2 activation confirmed these results, demonstrating large increases in Smad2 activation at 24 hours after pHx in comparison with vehicle-treated controls subjected to pHx (Fig. 4F). Interestingly, at 48 hours after pHx, phosphorylation of Smad2 was reduced to nearly baseline levels in ConA-pretreated mice, whereas it continued to increase in vehicle-pretreated mice. Smad activation is also associated with the activation of hepatic stellate cells. With the impairment in liver regeneration, the increased expression of TGFβ, and the activation of key Smad proteins, the development of hepatic fibrosis would be favored. Despite the environment provided by ConA pretreatment in the regenerating liver, stellate cell activation and collagen deposition were not observed (Fig. 4G). Given recent studies demonstrating the ability of TGFβ to dampen the proliferative response within the liver, these data provide further mechanistic insight into the impaired hepatocyte proliferation observed following T cell–mediated hepatitis. Together, these findings support the hypothesis that T cell–mediated liver injury induced by ConA is capable of altering the cytokine milieu, specifically down-regulating the expression of key mitogenic cytokines while simultaneously increasing the expression of proliferation inhibitor cytokines within the regenerating liver.
T Cell–Mediated Hepatitis Increases the Numbers of Oval-Like Cells Within the Regenerating Liver.
A simple reduction in the liver mass does not result in the proliferation of oval cells.24 However, when a reduction is combined with hepatocellular damage, oval cells are known to increase in number.24 A histopathological inspection of tissue sections from ConA-pretreated mice subjected to pHx revealed the presence of large numbers of small, oval-shaped cells located periportally (Fig. 5A), which were only present following the ConA pretreatment and pHx. To evaluate the oval cell response in this model of T cell–mediated hepatitis and liver regeneration, sections were stained for several established oval cell markers, including mPK, pan CK, and AFP. As shown in Figure 5B, limited numbers of mPK+ cells are present in the regenerating livers of the mice administered vehicle. Moreover, pan CK staining in vehicle-pretreated regenerating livers is confined to typical periportal ductular structures. The presence of T cell–mediated liver damage before pHx results in large increases in the numbers of pan CK and mPK positive cells within the liver. Indeed, significant numbers of pan CK+ cells (Fig. 5C) can be seen accumulating periportally, distinct from the typical pan CK+ ductular structures. A subsequent analysis of AFP expression reveals divergent patterns of expression in vehicle- and ConA-pretreated regenerating livers (Fig. 5B). Specifically, although AFP is expressed in what appear to be mature hepatocytes in vehicle-pretreated mice subjected to pHx, its expression is localized to small, oval-shaped cells in ConA-pretreated mice subjected to pHx. Together, these data demonstrate the ability of ConA-mediated hepatitis within the regenerating liver, possibly through the inhibition of mature hepatocyte proliferation, to increase the numbers of cells expressing oval cell markers.
Expansion of CD117+ and Sca-1+ Cell Populations and Their Co-Expression with Oval Cell Markers in the Damaged and Regenerating Liver.
Hepatic oval cells share similarities with hematopoietic and bone marrow–derived stem cells.4, 6, 25–27 For example, CD117 and Sca-1 are found on cells that also express certain hepatocyte markers such as albumin and are capable of transplantation and differentiation into mature hepatocytes.25–27 Given the oval cell response in this model of injury and regeneration and its previously reported ability to activate hematopoietic-like progenitor cells, we evaluated the ability of these markers to be used in the detection of these cell populations. In the first series of studies, a simple immunohistochemical assessment of CD117 expression revealed significant specific staining in vehicle-treated mice subjected to pHx on small cells lining the sinusoid, possibly endothelial cells (Fig. 6A). When mice were treated with ConA before pHx, small, round CD117+ cells could be seen clustering. Immunohistochemistry for Sca-1 revealed limited staining in vehicle-pretreated mice subjected to pHx, primarily on elongated cells lining the larger central veins. ConA-pretreated mice subjected to pHx showed increased numbers of Sca-1 positive cells both periportally and near areas of accumulating small, ovoid-shaped cells (Fig. 6A). To further quantify the numbers of CD117 and Sca-1 positive cells in this model, isolated hepatic nonparenchymal cells were stained for these markers and assessed by flow cytometric analysis.ConA pretreatment before pHx increased the numbers of both CD117 and Sca-1 positive cells in comparison with vehicle-pretreated mice 48 hours after pHx (Fig. 6B,C). Interestingly, the number of CD117+ cells, but not Sca-1+ cells, increased in both vehicle-pretreated and ConA-pretreated mice at 48 hours after ligation in comparison with their respective sham-operated controls. Moreover, there appeared a transient spike in Sca-1+ cells at 24 hours after ligation in vehicle-pretreated and ConA-pretreated mice. In addition, there appears to be a small subpopulation of cells expressing both CD117 and Sca-1 antigens. Together, these data demonstrate increased numbers of 2 morphologically distinct populations of cells expressing the naïve surface markers CD117 and Sca-1 within the damaged and regenerating liver.
CD117 and Sca-1 antigens are present on a number of hematopoietic cells and their progenitors, including lymphocytes and bone marrow–derived stem cells.28, 29 Further characterization of these cell populations was conducted with markers of hepatocytes, oval cells, and hematopoietic cells. As shown in Figure 7A, substantial numbers of CD117 positive cells isolated 48 hours following pHx in ConA-pretreated mice express markers of oval cells, including albumin (28%), AFP (18.64%), and CKs (37.02%). In addition, smaller numbers of these cells express the hematopoietic cell marker CD45 (16.02%), whereas the majority express the major histocompatibility complex (MHC) class I (82.66%). In contrast, limited numbers of Sca-1 positive cells express oval cell markers, including albumin (9.82%), AFP (0.72%), and CKs (9.37%). Interestingly, only limited numbers of Sca-1 positive cells express the hematopoietic marker CD45 (22.46%), and only 50% express the MHC class I complex. A further examination of CD117 and Sca-1 cell populations by serial section staining revealed significant colocalization of CD117 and AFP but not Sca-1 and AFP (Fig. 7B,C). Together, these data demonstrate that substantial numbers of CD117 positive cells share oval cell markers, whereas far fewer Sca-1 positive cells express previously defined oval cell markers in the damaged and regenerating liver.
NK Cells Inhibit the Oval Cell Response in the Damaged and Regenerating Liver.
NK cells represent an important component of the innate immune response within the liver.30 NK cells are responsible for the recognition of nonself antigens, including those displayed on cells infected by virus, and have been linked to the impairment of transplanted cell engraftment.31 More recently, NK cells have been associated with the impairment of hepatocellular regeneration following routine pHx13 and have been shown to be recruited to the liver during ConA-induced T cell–mediated hepatitis.32 Together, NK cells play an important immunomodulatory role within the normal and regenerating liver. Given this ability, we explored the impact of NK cells on the regulation of liver regeneration and oval cell expansion within the T cell–damaged and regenerating liver. As shown in Table 2, the depletion of NK cells with anti-asialo GM1 was unable to restore the regenerative response of ConA-injured hepatocytes or improve the hepatocellular injury associated with this combination of procedures. In contrast, both CD117 and Sca-1 cell populations were significantly enhanced by NK cell depletion, suggesting a potential inhibitory effect of NK cells on these naive or progenitor cell populations (Fig. 8A,B).
|Treatment||ALT (U/L)||LW/BW (%)||PCNA+ Cells/HPF||Spleen NK Cell Count (% total cells)|
|ConA||444 ± 70||3.01 ± 0.08||38.42 ± 6.01||6.20 ± 1.15|
|ConA + Ab||354 ± 31||2.95 ± 0.07||29.94 ± 4.47||3.80 ± 0.73|
Finally, to determine if ConA was capable of increasing the cytotoxicity of NK cell populations, total splenocytes from ConA-treated control or NK cell–depleted mice were incubated with total bone marrow cells. As shown in Figure 8C, splenocytes from ConA-pretreated mice showed enhanced cytotoxicity toward autologous bone marrow cells in comparison with total splenocytes from untreated wild-type mice. The treatment of mice with anti-asialo GM1 significantly reduced the ConA-induced increases in splenocyte cytotoxicity toward autologous bone marrow cells. In summary, these data implicate NK cells as inhibitors of the progenitor cell response but not the mature hepatocyte proliferative response to reduced hepatocellular volume likely through increased cytotoxicity toward primitive cell populations.
The regeneration of donor liver tissue following transplantation is crucial to graft and subsequently recipient survival.1, 2 Complicating factors including the activation of the innate immune response are often present in the regenerating organs, although their effects on the overall outcome of tissue regeneration are not well understood. The current series of studies has focused on the effects of T cell–mediated liver injury during the regenerative response. From this investigation, it is apparent that liver injury induced by the NKT cell activator, ConA, significantly impairs the early regenerative response (impaired cyclin D1 and E expression and Stat3 activation) within hepatocytes in conjunction with alterations in the expression of key mitogenic cytokines (IL6) and activation of known inhibitors of the regenerative response (i.e. TGFβ, IFNγ, Smad phosphorylation, and p21waf expression). Conversely, ConA pretreatment stimulates the NK cell–sensitive expansion of oval cells (CD117, AFP, albumin, and CK positive cells) as well as a less well differentiated Sca-1 positive cell population. Together, these studies demonstrate the ability of T cell–mediated liver injury to impair the mature hepatocyte proliferative response and increase the number of NK cell–sensitive oval cells within the liver.
T Cell–Mediated Inflammation and Liver Regeneration.
Central to the regenerative response is the production of certain mitogenic yet pro-inflammatory cytokines such as TNFα and IL6.3 Early studies reported a supportive role for TNFα signaling through TNFα receptor 1 in the mitogenic response of hepatocytes to pHx.33, 34 Similar to the effects of TNFα, the IL6-associated activation of gp130 was also implicated in Stat3 activation and hepatocyte proliferation in the reduced size liver.35–37 In contrast, other cytokines, including TGFβ1 and IFNγ, serve to inhibit the regenerative response of hepatocytes.23, 38 TGFβ signaling through TGFβ receptor 2 dampens the early regenerative response of the liver through the inhibition of cyclin expression and a subsequent G1-to-S transition.23 Furthermore, the deletion of Smad2, a downstream signaling molecule of TGFβ, significantly enhances the proliferative response of hepatocytes in vitro.22 Our studies demonstrate the ability of T cell–mediated hepatitis to both decrease the expression of the pro-mitogenic cytokine IL6 and inhibit Stat3 activation while simultaneously increasing the expression of IFNγ and TGFβ as well as the phosphorylation of Smad2. These alterations in cytokine responses occur at a time point (6 hours after pHx) when cyclin D1 expression is reduced and the expression of p21waf, an inhibitor of cyclin dependent kinases, is enhanced. The source of elevated TGFβ in this model is not clear, although several possibilities exist. For example, CD4+/CD25+ regulatory T cells produce large amounts of TGFβ when activated; this action is associated with the inhibition of cytolytic T cell activation.39 Interestingly, apoptotic T cells have also been shown to produce large quantities of TGFβ.40 The activation of T lymphocytes, specifically CD4+ T cells, is associated with apoptotic cell death.41 As ConA is known to activate and rapidly deplete NKT cells from the murine liver, it may also serve to activate other intrahepatic lymphocytes, leading to their death. Together, it appears that T cell–mediated hepatitis induced by ConA generates a hepatic microenvironment that is not conducive to hepatocyte replication, a clinically relevant scenario in which the paths of the innate and adaptive immune response and hepatocellular regeneration converge.
Not only does ConA alter the cytokine response, but it may also change the immune cell populations present or prime their level of activation.32, 42 Recent studies by Sun and Gao13 demonstrated the ability of poly dI:C activated NK cells through the production of IFNγ to significantly impair mature hepatocyte proliferation and damage regenerating hepatocytes. Similarly, NKT cells activated by interleukin 12 caused significant liver injury during the regenerative response in a TNFα-dependent manner.15 Within the current model, ConA induces the rapid activation and depletion of NKT cells, which serve to initiate the injury response (Fig. 1). However, NKT cells have repopulated the liver at the time of pHx and are therefore capable of contributing to the regenerative response. As NKT cells are required to initiate the injury response in this model, the determination of their direct role in this response is unclear. Further investigation using the selective depletion of NKT cells following the initiation of liver injury will be required to fully elucidate their role in this response.
NK cells do not appear to play a role in the impairment of normal hepatocyte replication induced by ConA as the depletion of NK cells fails to restore the regenerative response at 48 hours after a hepatectomy. Moreover, NK cells do not contribute significantly to the increased hepatocellular injury observed in the regenerating ConA-pretreated livers. In summary, NK cells do not appear to directly affect the regenerative response of the liver following ConA administration and pHx, whereas the direct role of NKT cells remains unclear.
Oval Cells, Liver Regeneration, and the Innate Immune Response.
Hepatic oval cells represent an important component of the liver's regenerative response when mature hepatocyte replication is impaired.24 Sakamoto and others16 previously reported the ability of ConA-induced liver injury coupled with regeneration to increase the number of hematopoietic-like stem cells. We extended these findings, demonstrating the presence of an oval cell-like population in the damaged and regenerating liver. Using 2 established markers of hepatic oval cells, CD117 and Sca-1, we demonstrated the expansion of 2 populations of naive cells within the damaged and regenerating liver. Consistently with previous reports, CD117 colocalized with several oval cell markers, including albumin, AFP, and CKs.43 Interestingly, 48 hours following pHx in vehicle-treated mice, increased numbers of CD117+ cells could be observed. An analysis of CD117+ cells in these mice by immunohistochemistry revealed a vascular staining pattern consistent with endothelial cell populations. Consistent with this staining pattern are studies by Takamiya and others44 that report the expression of CD117 on endothelial progenitor cells. Further investigation is required to fully characterize the CD117+ cell population present in the normally regenerating liver.
In contrast to CD117+ cells, only small numbers of Sca-1 positive cells expressed oval cell markers. In a 3,5-diethoxycarbonyl-1,4-dihydro-collidine diet–induced model of liver injury and oval cell expansion, Sca-1+ served as a specific marker of the oval cell population.4 In the current model, Sca-1+ cells could represent a more primitive cellular subset that precedes CD117+ oval cells. This hypothesis is supported by the level of expression of MHC class I in this cell population. Only 50% of Sca-1+ cells from the ConA-pretreated and regenerating liver express MHC class I, with greater than 80% of CD117+ cells expressing this antigen by flow cytometric analysis. Recent studies have reported the presence of an MHC class I negative subset of HPCs that appear to precede traditional oval cell populations.45 Alternatively, Sca-1+ cells from the T cell–damaged and regenerating liver could also represent important hematopoietic progenitors. Indeed, the fetal liver is an important source of hematopoietic precursors and site of immune cell development.46 Further exploration into the potential of this cell population is required to concretely determine its differentiation potential.
Little data exist identifying how oval cells may interact with or be affected by components of the hepatic immune system. Given their similarities to other naive cell populations, including bone marrow stem cells and HCC cell lines, it is plausible that this cell population may also be sensitive to the components of the innate immune response, including NKT and NK cells. It is well appreciated that both NKT cells and NK cells are capable of and responsible for the clearance of intrahepatic tumor cells.17, 32 Crowe and others47 demonstrated the capacity of liver-derived NKT cells to limit the expansion of both transferred sarcoma cells and B16 melanoma cells. Moreover, the adoptive transfer of NKT lymphocytes exposed ex vivo to HCC-derived antigens presented by dendritic cells led to the suppression of HCC in mice, an effect that was associated with increased NKT and CD8+ T cell numbers in the liver.48 Additionally, NK cells specifically inhibited the engraftment of human CD34+ stem cells in the mouse, suggesting that primitive progenitor cells may be directly regulated by components of the innate immune response.31 As hepatic progenitor or oval cells share characteristics with HCC and stem cell populations, NKT and NK cells may also limit their expansion within the regenerating liver. Within the current study, the direct depletion of NK cells resulted in increased numbers of both CD117 and Sca-1 cell populations, and this is consistent with a negative effect of NK cells on the oval and perhaps hematopoietic cell response. Moreover, ConA increased the NK cell–dependent cytotoxicity of autologous bone marrow cells, and this finding is similar to that of Miyagi et al.32 In their studies, ConA activated NK cells to exert cytotoxicity toward Colon-26 colon cancer cells in vitro and in vivo. Again, the role that NKT cells play in the oval cell response is not clear from the current investigation. Upon activation, NKT cells can exert significant Fas(CD95)-mediated and TNFα-mediated cytotoxicity toward regenerating hepatocytes, a response that could also limit the expansion of primitive cell populations such as oval cells.15, 49 Alternatively, activated NKT cells are capable of producing large amounts of IFNγ, which may promote the oval cell response.50 Further studies evaluating the activity of NKT cells within the damaged and regenerating liver are required to more fully elucidate their role. Together, however, data from the current study strongly implicate components of the innate immune response in the regulation of oval cells and possibly other progenitor cell populations.
In conclusion, the presence of T cell–mediated liver damage inhibits the normal regenerative response within the liver. Both hepatocellular damage and altered cytokine responses (i.e., IL6 production and Stat3 activation and TGFβ-induced Smad activation) may be responsible for this impairment and may also play a major role in the expansion of the oval cell compartment. Moreover, components of the intrahepatic innate immune response (NK cells and NKT cells) significantly reduce the expansion of the oval cell compartment. Therefore, therapies directed at modulating the innate immune response within the liver may be useful in preserving the mature hepatocyte regenerative response while also potentially protecting progenitor cell populations.