Oval cell-mediated liver regeneration: Role of cytokines and growth factors



    1. Western Australian Institute for Medical Research, *Biochemistry and Molecular Biology, School of Biomedical and Chemical Sciences, The University of Western Australia, Crawley and Department of Medicine, The University of Western Australia at Fremantle Hospital, Fremantle, Western Australia, Australia
    Search for more papers by this author
    • 1

      Authors contributed equally to this paper.


    1. Western Australian Institute for Medical Research, *Biochemistry and Molecular Biology, School of Biomedical and Chemical Sciences, The University of Western Australia, Crawley and Department of Medicine, The University of Western Australia at Fremantle Hospital, Fremantle, Western Australia, Australia
    Search for more papers by this author
    • 1

      Authors contributed equally to this paper.


    1. Western Australian Institute for Medical Research, *Biochemistry and Molecular Biology, School of Biomedical and Chemical Sciences, The University of Western Australia, Crawley and Department of Medicine, The University of Western Australia at Fremantle Hospital, Fremantle, Western Australia, Australia
    Search for more papers by this author

    1. Western Australian Institute for Medical Research, *Biochemistry and Molecular Biology, School of Biomedical and Chemical Sciences, The University of Western Australia, Crawley and Department of Medicine, The University of Western Australia at Fremantle Hospital, Fremantle, Western Australia, Australia
    Search for more papers by this author

    Corresponding author
    1. Western Australian Institute for Medical Research, *Biochemistry and Molecular Biology, School of Biomedical and Chemical Sciences, The University of Western Australia, Crawley and Department of Medicine, The University of Western Australia at Fremantle Hospital, Fremantle, Western Australia, Australia
    Search for more papers by this author

Associate Professor G Yeoh, Biochemistry and Molecular Biology, School of Biomedical and Chemical Sciences, The University of Western Australia, Entrance 2, Hackett Drive, Crawley 6009, Western Australia, Australia. Email: yeoh@cyllene.uwa.edu.au


Abstract   In experimental models, which induce liver damage and simultaneously block hepatocyte proliferation, the recruitment of a hepatic progenitor cell population comprised of oval cells is invariably observed. There is a substantial body of evidence to suggest that oval cells are involved in liver regeneration, as they differentiate into hepatocytes and biliary cells. Recently, bone marrow cells were shown to be a source of a stem cells with the capacity to repopulate the liver. Presently, the relationship between bone marrow cells and oval cells is unclear. Investigations will be greatly assisted by the availability of in vitro models based on a knowledge of cytokines that affect oval cells. While the cytokines, which regulate the different hematopoietic lineages, are well characterized, there is relatively little information regarding those that influence oval cells. This review outlines recent developments in the field of oval cell research and focuses on cytokines and growth factors that have been implicated in regulating oval cell proliferation and differentiation.

© 2003 Blackwell Publishing Asia Pty Ltd


Severe and chronic liver injury caused by drugs, viruses and toxins impairs hepatocyte proliferation. When the ability of hepatocytes to divide and replace damaged tissue is compromised, a subpopulation of liver cells, termed oval cells, are induced to proliferate. In a normal adult liver, these cells are quiescent, existing in low numbers around the periportal region; they proliferate following severe, prolonged liver trauma. Oval cells were first reported by Kinosita et al. who observed small, ovoid cells in livers of rats exposed to the carcinogenic azo dye ‘Butter Yellow’.1 These cells were later termed oval cells because of their characteristic morphology; ovoid nucleus, small size (relative to hepatocytes) and high nuclear to cytoplasmic ratio.2 Oval cells are observed to proliferate in the periportal region of the liver and, as liver damage progresses, they infiltrate into the parenchyma along the bile canaliculi between the hepatic cords. Initially considered of little importance, it was not until the last 20 years that researchers began to characterize these cells. Extensive studies in rodent models of hepatocarcinogenesis and other non-carcinogenic injury models suggest that oval cells may represent a facultative hepatic progenitor/stem cell compartment.3–5 Indeed, it is increasingly apparent that hepatic progenitor cells share common characteristics with stem cells of the hematopoietic system. In this review, we highlight recent findings in the field of hepatic progenitor cell biology, with emphasis on the biological factors that potentially influence proliferation, differentiation and migration.

Oval cells are a heterogeneous cell compartment

Many rodent models are currently used to study oval cell biology (Table 1), and it is apparent that the term ‘oval cell’ describes a heterogeneous cell population that express different combinations of phenotypic markers from both the hepatocyte and biliary lineage (Table 2). Because of the observed cellular heterogeneity, Fausto named this cell population the oval cell compartment.6 Ultrastructural analysis of ‘small cells’ in the portal region of livers from patients with chronic liver disease identified three oval cell types, distinguished by morphological characteristics and location.7 Type I cells, with a primitive phenotype, were observed in close proximity to proliferating bile ducts and in acinar arrangements around hepatocytes. Type II cells had the same general features as type I cells, however, they were exclusively located in areas of ductular proliferation, next to cells comprising small ductules. Finally, type III cells had the general characteristics of type I cells, with the phenotype of mature hepatocytes including prominent nuclei and voluminous cytoplasm.

Table 1.  Rodent models of oval cell induction
  1. AFP, alpha fetoprotein; PH, partial hepatectomy.

Choline deficient, ethionine supplemented
(CDE) diet
0.1%: Proliferation of AFP and albumin expressing ductular
oval cells;
few well-defined ductular profiles; hepatocyte
differentiation rare;
no necrosis.
Standard choline-devoid diet containing
0.05–0.1% ethionine for 1 day to 12 weeks.
0.07%: Oval cells proliferate as cords; ducts form by 2 weeks,
some form intestinal like cells by week 5.
16, 73
 0.05 or 0.1%: Parenchymal necrosis, massive oval cell
proliferation and cholangiofibrosis. No hepatocyte
differentiation within ducts.
50% choline-devoid diet and 0.15% ethionine
in drinking water for 1 day to 4 weeks (0.15%
ethionine is equivalent to 0.035% ethionine in
solid chow). Mice.
Oval cells proliferated as cords, with ducts appearing after
4 weeks.
2-actylaminofluorene and partial
Oval cell proliferation into parenchyma; intestinal metaplasia
common after 3 weeks, which develop into
cholangiofibrotic lesions. No hepatocyte differentiation.
0.02% 2-AAF for 2 weeks and PH at midway
point. Killed up to 9 weeks after PH. Rats.
Single i.p. dose (700 mg/kg) and killed 1–8 days
later. Rats.
Parenchymal necrosis. Proliferation of small clusters and
ductular oval cells that differentiate into transitional
hepatocytes and lose biliary specific markers.
Two i.p. doses of galactosamine (750 mg/kg),
6 h apart and killed 1–10 days later. Rats.
As above with oval cells expressing fetal hepatocyte enzyme
markers and normal bile ducts. Small hepatocytes and
hepatocytes lining atypical duct structures present.
Single i.p. injection of dipin (60 mg/kg),
followed 2 h later by PH. Liver examined 18
months thereafter. Mice.
Initial massive parenchymal necrosis, oval cells proliferate into
small foci of hepatocytes. Damaged parenchyma replaced
by newly formed hepatocytes.
Table 2.  Liver cell lineage markers
MarkerOval CellsHepatocytesBile Duct CellsReferences
  1. AFP, alpha fetoprotein; CK, cytokeratin; π-GST, π-glutathione; M2-PK, M2-Pyruvate kinase; OV, oval cell marker.

Albumin++23, 24, 27
AFP+Fetal23, 24, 27
CK8+++23, 24, 79
CK14+/–24, 28, 79
CK18+++23, 24, 79
CK19–/++79, 38

Similar analysis of the oval cell response in hamster liver, following Clonorchis sinensis infection and dimethylnitrosamine treatment, identified cells comparable to the human type I, II and III cells.8 Evaluation of each cell compartment identified three immunologically distinct oval cell populations: (i) a primitive oval cell that did not express α-fetoprotein (AFP), cytokeratin 19 (CK19), OV-6 or π-glutathione S-transferase (GST); (ii) an hepatocyte-like oval cell expressing AFP, but not OV-6 or π-GST; and (iii) a ductular-like oval cell, negative for AFP but expressing CK19, OV-6 and π-GST. It is our view that two distinct populations exist in rats fed a choline-deficient, ethionine-supplemented (CDE) diet, as detailed analysis of pyruvate kinase (PK) and GST isoenzyme expression reveals that most oval cells coexpress fetal and biliary specific markers, while a minor population coexpress fetal and adult hepatocyte markers.9,10

Accordingly, oval cells are a multipotent cell population with the potential to differentiate into hepatocytes and biliary epithelia, and under certain conditions, pancreatic and intestinal epithelia (Fig. 1). The oval cell–hepatocyte relationship was initially suggested in vivo by tracing 3H-thymidine-labeled oval cells into newly formed hepatocytes.11 An extension of these experiments, which included an in situ detection of AFP mRNA in oval cells and basophilic hepatocytes, supported initial findings,12 and these results have been substantiated in a recent study involving the Long–Evans Cinnamon rats that carry a defect in the Wilson's disease gene.13 The resulting copper accumulation causes acute hepatitis and significant mortality at 4 months of age, while survivors develop chronic hepatitis, during which oval cell proliferation occurs and culminates in a high incidence of hepatocellular carcinoma (HCC) after 1 year. Oval cell lines established from these rats differentiate into hepatocytes in vivo when transplanted into the liver of Nagase analbuminemic rats. Interestingly, bile ducts do not develop; however, Yasui et al. do not exclude the possibility that the established cell lines were derived from a subpopulation of oval cells committed to differentiate only along the hepatocyte lineage.13 Similarly, oval cell lines established from rats fed a CDE diet for 2 or 6 weeks (LE/2 and LE/6, respectively)14 have been observed to differentiate into hepatocytes when cultured in a 3-D collagen matrix in conjunction with a fibroblast feeder layer.15 The bipotential nature of these cells was demonstrated by the formation of ductal structures in the presence of a hepatocyte growth factor and/or keratinocyte growth factor, suggesting differentiation along the biliary lineage. This is in agreement with our observation that oval cells differentiate into bile duct cells in vivo, as shown by GST and PK immunohistochemistry.16

Figure 1.

Origin and fate of oval cells. Hepatic progenitor cells originating in the bone marrow migrate to the liver and infiltrate the portal region via the canals of Hering. These stem cells may give rise to oval cells, which can differentiate into mature hepatocytes. Oval cells are bipotential stem cells, originating from the canals of Hering, which differentiate to produce cells of either the hepatocytes or biliary epithelia lineage. Under appropriate conditions, oval cells also have the capacity to differentiate into pancreatic cells and intestinal metaplasia, and are associated with hepatocellular carcinoma (HCC).

Oval cell origin

The existence of an oval cell precursor population was initially suggested after 2-acetylaminofluorene (2-AAF) and/or allyl alcohol treatment of mice was found to induce proliferation of a periportal population of oval-like cells negative for the expression of classical oval cell markers.17 These cells began to express AFP after 2 days, possibly because of a commitment to differentiate along the hepatocyte lineage and, accordingly, it was assumed that oval cells were derived from progenitor cells located endogenously in the liver. An examination of the 3-D relationship between CK19-expressing cells in massive necrosis implies that the characteristic ductular reaction observed is caused by the proliferation of cells lining the canals of Hering, the junction between the hepatocyte canalicular system and terminal bile ducts.18 Likewise, AAF-induced oval cells are generated by the proliferation of terminal bile ducts, and they form structures representing an extension of the canals of Hering.19 Considering the first cells undergoing proliferation and differentiation into hepatocytes are found in this region, it is reasonable to assume that the canals of Hering are the source of hepatic progenitor cells and their direct progeny, the oval cell.

Hepatic progenitor cells may be resident liver cells, or they may be recruited from extrahepatic sources. Suzuki et al. have recently identified a population of hepatic stem cells that exist in developing mouse liver.20 These CD49f+, CD29+, c-Met+, c-kit, CD45, Ter119 cells, designated ‘hepatic colony-forming-unit in culture’ (H-CFU-C), possess multilineage differentiation potential and self-renewing capability.20 However, unlike oval cells, they do not express hepatocyte or cholangiocyte specific markers and c-kit.21 Cells derived from a single H-CFU-C expand in vitro and are capable of reconstituting hepatocytic, bile-ductal, pancreatic and intestinal structures in vivo.20 Whether H-CFU-C's exist in the adult liver has not been established; however, these cells may represent the resident hepatocyte progenitor.

Evidence of a cell lineage relationship between the hematopoietic system and the liver supports the extrahepatic origin of hepatic progenitor cells. Classical hematopoietic markers, including Thy-1, c-kit and CD34, are expressed on the surface of oval cells.22–25 Additionally, bone marrow transplantation of purified c-kit+, Lin and Sca-1+ hematopoietic stem cells rescued fumarylacetoacetate (FAH)-deficient mice, an animal model of fatal hereditary tyrosinemia type 1.26 Cross-gender, or cross-strain, bone marrow and whole liver transplants in mice have identified cells in the bone marrow that are capable of repopulating the liver.27,28 These results are supported by human studies using archival liver biopsy specimens from recipients of cross-gender therapeutic bone marrow transplants who later developed chronic liver damage due to recurrent disease. Analysis for the presence of the Y chromosome in cells of liver biopsy specimens showed that bone marrow-derived cells give rise to hepatocytes alone, or both hepatocytes and cholangiocytes.29,30 However, there is disagreement on the extent of engraftment. Alison et al. reported a relatively low frequency of Y-chromosome positive hepatocytes (0.5–2%); whereas Theise et al. reported that, even in mild conditions, significant engraftment (5–21% for hepatocytes and 4–18% for cholangiocytes) occurred; while in cases of severe injury, up to 64% of periportal hepatocytes and 38% of cholangiocytes were donor derived.29,30 Hence, hematopoietic cells are capable of migrating to the liver and differentiating into hepatocytes in rodents and humans; however, the role of these cells in liver regeneration remains unclear.

The origin of the hepatic oval cell remains a contentious issue. Nevertheless, it is increasingly apparent, as hypothesized by Sell,31 that three levels of proliferating cells are associated with the liver. These are: the mature adult hepatocyte; a tissue-determined stem cell, originating endogenously in adult liver in the terminal bile ducts; and a multipotent stem cell, that may be exogenously derived from circulating bone marrow stem cells. The direct lineage relationship that links bone marrow-derived hepatic stem cells and/or H-CFU-C's to oval cells has not been definitively proven.

Factors that regulate oval cell-mediated liver regeneration

Regardless of the hepatic progenitor cell origin, it is well established that oval cell proliferation occurs when the replicative capacity and function of hepatocytes is impaired. Oval cells are resistant to the effects of hepatotoxins/carcinogens, thus, they proliferate and migrate throughout damaged liver lobules to replace lost parenchyma. Given the association between experimental models of hepatotoxicity and carcinogenesis, it is not surprising that oval cells have been identified in acute and chronic human liver pathologies including HBV/HCV infection, hemachromatosis and alcoholic liver disease.32–39 An extensive inflammatory response is characteristic of many chronic liver diseases. This response is mediated through resident Kupffer, hepatic stellate cells and infiltrating inflammatory cells, which secrete chemokines, growth factors and cytokines in response to liver injury. As such, the histological changes that accompany oval cell proliferation provide useful clues for identifying factors that mediate oval cell activation.

Role of inflammatory cytokines

We have previously observed that oval cells are located in close association with inflammatory cells, particularly in hepatitis C.38 Further studies by Libbrecht et al. reported a correlation between the number of oval cells and degree of inflammatory infiltrate.40 This polymorphonuclear and mononuclear infiltrate may activate oval cells via cytokine and/or growth factor release.

Two important inflammatory cytokines associated with the regulation of liver growth are tumor necrosis factor (TNF) and interleukin-6 (IL-6). Inhibition of TNF by dexamethasone administration impairs the proliferation of hepatic cell populations following 2-AAF/partial hepatectomy (PH), completely suppressing activation of the oval cell compartment.41 In addition, in vitro studies have shown that TNF stimulates proliferation of the LE/6 murine oval cell line. These studies have outlined the importance of the TNF family of ligands and receptors in the activation of oval cells. Collectively, these findings suggest a role for TNF in oval cell proliferation. We have identified impaired oval cell proliferation in TNF receptor 1 (TNFR1) knockout mice, illustrating that TNFR1 downstream signaling events are required for maximal oval cell proliferation.42 In contrast, TNF receptor 2 (TNFR2) appears to have no involvement, as the number of oval cells induced in TNFR2 knockout mice and controls was similar. Although some TNF production can be attributed to infiltrating inflammatory cells, the majority of hepatic TNF produced during liver regeneration is Kupffer-cell derived. Depletion of Kupffer cells by treatment with gadolinium chloride prior to bile duct ligation completely ablates oval cell induction, but not ductular proliferation, suggesting that multiple cytokines produced by Kupffer cells are crucial to the process.43 This is consistent with the observation that dexamethasone treatment has no effect on bile duct proliferation following bile duct ligation, implying that TNF is not essential for the expansion of biliary epithelia.41 However, there is good evidence for its role in oval cell proliferation.

In contrast, there is conflicting evidence regarding the requirement for IL-6 in oval cell-mediated liver regeneration. Interleukin-6 production following 2-AAF/PH is inhibited by dexamethasone with an accompanying reduction in oval cell numbers. This suggests its involvement in the activation of the oval cell compartment.41 Yet, comparison of the cellular response to cocaine-induced periportal injury in normal and IL-6–/– mice demonstrated increased proliferation of periportal oval cells in IL-6–/– mice to compensate for the decrease in restorative proliferation of hepatocytes, biliary epithelia and sinusoidal cells.44 Ten days after injury, the liver was completely repaired in all mice, indicating that IL-6 is not essential for oval cell proliferation. It is feasible that other members of the IL-6 family, including leukemia inhibitory factor (LIF) and/or oncostatin M (OSM), may compensate for the absence of IL-6 in these mice.45 Indeed, LIF is increased and remains elevated during oval cell induction by 2-AAF/PH, suggesting that it may have a role in the expansion and differentiation of the oval cell compartment.46 Additionally, in situ hybridization has demonstrated LIF, LIF receptor (LIFR) and glycoprotein (gp)130 mRNA expression in oval cells, with weak expression in parenchymal cells. Oncostatin M has recently been implicated in the maturation of fetal hepatocytes in vitro and in vivo,47 and it may have a similar role in the hepatic differentiation of oval cells.

γ-Interferon (IFN-γ) is another inflammatory cytokine considered to play an integral role in controlling stem cell-mediated liver regeneration.48 Suppression subtractive hybridization following 2-AAF/PH identified genes associated with the proliferation of oval cells including IFN-γ, IFN-γ receptor α subunit (IFN-γRα), IFN-γRβ primary response genes (gp91phox), IFN-γRβ secondary response genes such as ICE, CD54/ICAM-1 and uPAR, cytokines that induce expression of IFN-γ, which include IL1-β and IL-18, and cell adhesion molecules that regulate the interactions between lymphocytes and epithelial cells; lymphocyte function-associated molecule 1-α (LFA-1α/CD11 and CD54/ICAM-1). These proteins are all part of the complex cellular response associated with the IFN-γ signaling cascade. γ-Interferon alone is not a hepatic mitogen, however, it can act synergistically with other growth factors such as EGF. Therefore, it has been suggested that interferon functions to prime certain cell populations to respond to mitogenic stimuli.

Role of hepatic stellate cells

Resident non-parenchymal liver cells can respond to injury by releasing specific paracrine factors that may activate the hepatic progenitor cell compartment. The concurrent proliferation of mesenchymal cells with oval cells was first reported by Popper et al. in 195749 and, using the 2-AAF/PH model, Evarts et al.50 showed that hepatic stellate cell proliferation is closely associated with oval cell proliferation. Hepatic stellate cells express hepatocyte growth factor (HGF), transforming growth factor-β (TGF-β), transforming growth factor-α (TGF-α) and acidic fibroblast growth factor (aFGF). Interestingly, these cytokines have all been identified in regenerating liver following PH. In vitro studies have shown that epidermal growth factor (EGF) can stimulate the growth of oval cell lines maintained in soft agar.51 Infusion of exogenous EGF during oval cell induction by 2-AAF/PH significantly enhances their proliferation and subsequent migration into the parenchyma.52 Although oval cells induced by 2-AAF/PH do not express HGF, they express mRNA for the HGF receptor, c-met.53,54 High levels of HGF mRNA are expressed by hepatic stellate cells proliferating in close proximity to oval cells, suggesting that hepatic stellate cell-derived HGF may cause oval cell proliferation via the paracrine activation of c-met.53 Furthermore, infusion of human recombinant HGF into rats following 2-AAF treatment results in the expansion of several liver cell populations, including the hepatic stellate cells and oval cells.52 Evidently, this situation also occurs in humans, as elevated levels of serum HGF are present in individuals with chronic hepatitis and cirrhosis,55–57 both conditions in which oval cell proliferation is well documented.7,18,38,40,58,59

In situ hybridization has revealed that oval cells and basophilic hepatocytes, in addition to hepatic stellate cells, express aFGF transcripts.60 A marked increase in hepatic aFGF levels has been reported at the peak of oval cell proliferation in the 2-AAF/PH model, and levels greatly exceeded those observed after PH alone, suggesting a more prominent role for aFGF in oval cell-mediated regeneration.54 High levels of TGF-α expression are observed not only in hepatic stellate cells, but also in oval cells in the 2-AAF/PH model,50,61 and TGF-α expression is detected in foci of hepatocytes that appear to be oval cell-derived. Similar results have been observed in human hepatitis B-associated HCC, where TGF-α is localized to oval cells.33 In contrast, TGF-β is implicated as a negative regulator of oval cell activation. Transforming growth factor-β1 expression on smooth muscle actin (SMA)-positive hepatic stellate cells coincides with oval cell proliferation in the 2-AAF/PH model and correlates with maximal oval cell apoptosis.62 As TGF-β1 is proposed to be a negative growth signal that controls liver size by the induction of apoptosis during compensatory hyperplasia, it is possible that TGF-β1 may assist in the remodelling of liver parenchyma during oval cell-mediated liver regeneration by terminating oval cell activation.62

Interaction with the extracellular matrix

The plasminogen activator and plasmin proteolytic cascades have an important function in stem cell-mediated regeneration, as most regenerative responses are associated with changes in the extracellular matrix. The plasminogen activator/plasmin system is complex, involving many proteins including urokinase-type plasminogen activator (uPA), tissue type plasminogen activator (tPA), the uPA receptor (uPAR), and plasminogen activator inhibitor 1 (PAI-1).63 The upregulation of uPA mRNA accompanying oval cell proliferation has been reported, and infusion of uPA enhanced the mitogenic response of cells located near bile ducts after the administration of 2-AAF. Expression of uPA, uPAR and PAI-1 is upregulated in the 2-AAF/PH model of oval cell induction, and localized to the ductal structures formed by the oval cells.64 Urokinase-type plasminogen activator expression was also detected in non-parenchymal cells along the hepatic sinusoids. In addition to the significant remodelling and oval cell migration into the parenchyma that is occurring, this system may play a role in regulating several growth factors that are involved in oval cell regeneration. The expression of uPA, uPAR and PAI-1 is regulated in vivo and in vitro by a number of cytokines and growth factors, including HGF, EGF and TGF-β. Additionally, HGF and TGF-α are both secreted as latent factors and require activation by proteases.65–69

Role of stem cell factor

Stem cell factor (SCF) and its receptor, c-kit, play a fundamental role in survival, proliferation, differentiation and migration of a variety of stem cells and may similarly affect oval cells. Stem cell factor is induced early in the activation of oval cells by 2-AAF/PH, but this is not observed following PH alone.70 Oval cell precursors express both SCF and c-kit, suggesting that an autocrine mechanism may be involved, although the precise role of this receptor–ligand system in liver regeneration is unclear. Oval cell induction is significantly suppressed in Ws/Ws (white spotting on the skin) rats, in which the c-kit receptor tyrosine kinase (KIT) activity is severely impaired.22 Once oval cells appeared in Ws/Ws rats, expression of oval cell specific marker proteins and the proliferative response of these cells was similar to controls, implying that the KIT-mediated signal transduction is not essential for the phentotype or proliferation of oval cells. It appears that the SCF/KIT system plays a crucial role in oval cell appearance, either by regulating the number of hepatic progenitor cells within the liver or by committing hepatic progenitor cells to differentiate into oval cells.

Proposed model for oval cell-mediated liver regeneration

In light of the current observations, we propose the following model of oval cell-mediated liver regeneration, as outlined in Fig. 2. In response to hepatocellular damage, the liver initiates an immune response, secreting a complex array of cytokines and chemokines. These signals promote the infiltration of Kupffer and hepatic stellate cells to the damaged area, and in situations where hepatocyte regeneration is blocked, hepatic progenitor cells may be recruited from the bone marrow by the release of chemoattractants, such as SCF. Hepatic progenitor cells migrate to the liver and infiltrate the lobules via the canals of Hering. In addition, resident primitive cells such as the H-CFU-C's may be another source of hepatic progenitor cells. These hepatic progenitor cells give rise to oval cells, which respond to the different cytokines released by the Kupffer and hepatic stellate cells. Tumor necrosis factors released by Kupffer cells stimulate oval cell proliferation, expanding the oval cell population. Hepatic stellate cells release other oval cell growth factors, including EGF, aFGF, TGF-α and HGF. Hepatocyte growth factor, EGF and TGF-β regulate the expression of uPA, uPAR and PAI-1, proteins involved in the plasmin proteolytic cascade. This cascade is responsible for the activation of TGF-α and HGF by proteolytic cleavage. Finally, TGF-β release by Kupffer and hepatic stellate cells triggers apoptotic pathways in oval cells, and ultimately, cessation of oval cell-mediated liver regeneration. Thus, the cytokines released by Kupffer, hepatic stellate and oval cells act in concert to control oval cell proliferation and remodelling of the liver parenchyma.

Figure 2.

Hepatocellular damage initiates an immune response in the liver, which leads to the secretion of a complex mixture of cytokines and chemokines. These signals promote the infiltration and/or expansion of Kupffer cell (KC) and hepatic stellate cells (HSC). When hepatocyte regeneration is blocked, hepatic progenitor cells (HPC) are recruited from the bone marrow by chemoattractants, such as stem cell factor (SCF). Hepatic progenitor cells migrate and infiltrate the liver lobules via the canals of Hering. Resident stem cells (hepatic colony-forming unit in culture (H-CFU-C)) may also be a source of hepatic progenitor cells. In the liver, HPCs give rise to oval cells (OC), which respond to cytokines released by KCs and HSCs. Tumor necrosis factor (TNF) and interleukin (IL)-6 released by KCs stimulate OC proliferation, while γ-interferon (IFN-γ) may prime OCs to respond to mitogenic stimuli. Other members of the IL-6 family, including leukemia inhibitory factor (LIF) and oncostatin M (OSM) may also enhance to OC proliferation. Oval cell growth factors released by HSCs include epidermal growth factor (EGF), acidic fibroblast growth factor (aFGF), tumor growth factor alpha and beta (TGF-α and TGF-β) and hepatocyte growth factor (HGF). Hepatocyte growth factor, EGF and TGF-β regulate the expression of urokinase-type plasminogen activator (uPA), uPA receptor (uPAR) and plasminogen activator inhibitor 1 (PAI-1), members of the plasmin proteolytic cascade. This activates TGF-α and HGF by proteolytic cleavage. Tumor growth factor-β may be a negative regulator of OC proliferation. Tumor growth factor-β triggers apoptosis in OCs and blocks OC-mediated liver regeneration. Thus, the cytokines released by KCs, HSCs and OCs act in concert to control OC proliferation and remodelling of the liver parenchyma.

The identification of factors that control this process, and clarification of their mechanism of action with respect to oval cell mediated liver regeneration will facilitate the establishment of in vitro models to investigate oval cell biology. This will help establish links, if any, between the bone marrow-derived liver stem cell and oval cells. As we expand our knowledge of factors controlling this complex process, we will be able to distinguish mechanisms that regulate oval cell-mediated liver regeneration from those that promote hepatocarcinogenesis.


The research laboratories of Associate Professors Abraham, Olynyk and Yeoh acknowledge support from the Cancer Foundation of Western Australia and the Western Australian Institute for Medical Research. Dr Croager is supported by a fellowship from the Healy Medical Research Foundation, Australia.