Department of Pathology, University of Pittsburgh, School of Medicine, Pittsburgh, PA
Department of Medicine (Gastroenterology), University of Pittsburgh, School of Medicine, Pittsburgh, PA
Assistant Professor of Pathology (CMP) and Medicine (GI), University of Pittsburgh Cancer Institute (UPCI)-GI Oncology, McGowan Institute of Regenerative Medicine (MIRM)-Liver, University of Pittsburgh, School of Medicine, S421-BST, 200 Lothrop Street, Pittsburgh, PA 15261
Wnt/β-catenin signaling is emerging as a forerunner for its critical roles in many facets of human biology. Its roles in embryogenesis, organogenesis, and maintaining tissue and organ homeostasis demonstrate its munificent character. Its roles in pathological conditions such as cancer and other human disorders such as inflammatory disorders and fibrosis reveal its villainous disposition. In liver, it also maintains its dual personality and is clearly of essence in several physiological events such as development, regeneration, and growth. Its aberrant activation is also evident in many different tumors of the liver, and recent studies are beginning to identify its role in additional hepatic pathological conditions. It is contributing to liver physiology and pathology by regulating various basic cellular events, including differentiation, proliferation, survival, oxidative stress, morphogenesis, and others. This review discusses the contribution of the Wnt/β-catenin signaling pathway in these events and simultaneously provides an essential overview of the major developments in the field of Wnt/β-catenin and liver pathobiology. In addition, areas that are currently deficient or understudied are identified and discussed along with the avenues of translational and clinical relevance. (HEPATOLOGY 2007;45:1298–1305.)
The role of the Wnt/β-catenin signaling pathway in liver biology has come to the forefront over the last several years. This development has been fueled by a number of studies examining this signaling pathway in the essential physiologic processes in the liver, elucidating its importance in development, growth, regeneration, zonation, metabolism, and oxidative stress. Likewise, there have been advances made in understanding the role of β-catenin in the development of various liver diseases. Studies of pathological specimens and rodent models of liver diseases have demonstrated aberrations in the Wnt/β-catenin signaling pathway in conditions ranging from hepatitis to hepatocellular carcinoma (HCC). Eventually this pathway has turned out to be among the central players in maintaining liver health.
β-Catenin is the chief downstream effector of the canonical Wnt signaling pathway (Fig. 1). In the normal steady state, β-catenin is targeted for degradation by phosphorylation at serine and threonine residues through the action of casein kinase Iα and glycogen synthase kinase 3β (GSK3β).1, 2 These proteins form a larger degradation complex with axin, adenomatous polyposis coli (APC), and diversin, all of which play a role in successful post-translational modification of β-catenin, which allows for recognition and subsequent ubiquitination by the β-transducin repeat-containing protein.
An activated state is brought about when Wnt binds to its receptor frizzled, inducing the formation of a ternary complex with low-density lipoprotein receptor related protein (LRP)5/6, ultimately leading to inactivation of GSK3β via activation of disheveled (Fig. 1). This leads to hypophosphorylation of β-catenin and its release from complex with adenomas polyposis coli (APC) and axin, with ensuing nuclear translocation of β-catenin where it binds to an HMG (High Mobility Group) box containing DNA-binding protein T cell factor/lymphoid-enhancing factor family member and controls transcription of various target genes. A listing of these targets can be found at: http://www.stanford.edu/∼rnusse/pathways/targets.html
Wnt/β-Catenin Signaling Unique to Liver Cell Biology
Although β-catenin is central to liver biology, as is discussed in the subsequent sections, its upstream effectors remain obscure in the liver. Classically, there are 19 Wnt and 10 Frizzled genes, but it is unlikely that all of these are playing an essential role in liver. Our laboratory has recently identified 11 Wnts and 9 Frizzleds that are normally expressed in an adult mouse liver.3 Furthermore, a differential expression of these genes was observed in various cell types within the liver (Fig. 2). Specific differences were observed in their expression in active and resting states of various cell types. These basic findings demonstrate the diversity in the Wnt signaling between various cell types, which might be of high significance in different pathological states of the liver. Thus, future studies examining individual Wnt or Frizzled genes will be of great interest to further define their role in normal liver growth and development as well as in hepatic diseases.
Beyond the spectrum of such classical Wnt signaling, two associations involving β-catenin are important in hepatocytes (Fig. 3). First, a well-characterized association between β-catenin and E-cadherin is also noted at the hepatocyte membrane. This connection forms a link between the cytoplasmic domain of the cadherins and the actin portion of the cytoskeleton, having significant implications in cell–cell adhesion. This interaction has been previously shown to be negatively regulated by phosphorylation of β-catenin at tyrosine 654.4 This event has been shown to inhibit the β-catenin–E-cadherin association, which in turn causes the disruption of the adherens junctions. This disassembly leads to the loss of intercellular adhesion, thus imparting a motogenic characteristic to the cell. The other key association in the liver is that of β-catenin with the hepatocyte growth factor (HGF) receptor c-Met, and studies in primary hepatocyte cultures have shown HGF stimulation to lead to dissociation of this complex along with the nuclear translocation of β-catenin.5 This study identified the requirement of the intra-cytoplasmic tyrosine kinase domain of the β-subunit of Met for inducing tyrosine phosphorylation and nuclear translocation of β-catenin. More recently, HGF/Met stimulation was found to induce phosphorylation of β-catenin at tyrosine 654 and 670, to promote proliferation of the hepatocytes.6 Other reports have identified tyrosine 142 to be effected by the HGF/Met signaling in a kidney fibroblast cell line.7 However, this residue had no role in HGF-induced mitogenesis in hepatocytes and may highlight tissue or functional specificity of these interactions.
The Met/β–catenin relationship has clear implications during in vivo hepatocyte proliferation as well.8 HGF gene therapy resulted in the loss of Met-β–catenin complex, nuclear translocation of β-catenin, and contributed to hepatic growth.8 Failure of such growth-promoting effects of HGF was observed when such therapy was instituted to the β–catenin conditional null mice. Another important conclusion of the study was that prolonged HGF gene delivery over 5 weeks resulted in increased E-cadherin–β-catenin complex at the membrane. This suggested that excessive β-catenin could be sequestered at the membrane as a means of “adaptation” or protection from excessive β-catenin signaling. This could be a potential mechanism by which E-cadherin might act as a tumor suppressor. What triggers such sequestration and what will be the fate of β-catenin, if E-cadherin gets saturated, remain to be determined. Additional pertinent questions that remain unanswered include: Are β-catenin, Met, and E-cadherin together in a single complex, or are the interactions independent of one another? Are these differential interactions surface-specific in the polarized hepatocytes or are these cell-specific within the liver? How do these evolve during the processes of development and hepatocyte differentiation? Answers to these basic cell biology questions will be critical to further our understanding of the role of tyrosine phosphorylation of β-catenin in hepatic morphogenesis, mitogenesis, and motogenesis (Fig. 3).
Liver derivation from the foregut endoderm occurs around somite stages 5 to 6 as a result of signaling from mesoderm in the form of fibroblast growth factors and bone morphogenetic protein-4, both of which are incidentally downstream targets of the Wnt pathway.9, 10 However, no study has directly examined this relationship during liver development. A recent study identified active mesodermal Wnt2b signaling at these stages, which for the first time documents an important role of Wnt signaling as a positive regulator of liver specification and induction in zebrafish.11 Additional studies would be of essence to examine whether Wnt2b or additional Wnts might be the upstream regulators of the factors known to be critical for hepatic induction. We were the first to report temporal expression of β-catenin during mouse liver development with high levels at embryonic day 10 (E10)-E14 along with its nuclear/cytoplasmic location in hepatoblasts, which correlated with increased cell proliferation.6, 12 When whole livers from E10 were cultured with antisense oligonucleotides against the β-catenin gene (CTNNB1), a noteworthy decrease occurred in β-catenin protein, cell proliferation, and survival.13 Conversely, overexpression of constitutively active β-catenin in the developing liver was shown to lead to a threefold increase in liver size and an expansion of the hepatocyte precursor cell population.14 This effect on cell proliferation might be mediated by downstream targets that are playing a role in cell cycle such as cyclin D1. Thus, β-catenin is likely playing an important role during early liver development in hepatocyte expansion via its transcriptional co-activator role. Interestingly, after E15 stage, an overall decrease in total β-catenin levels was observed when it began localizing chiefly to the hepatocyte membrane.12 This membranous localization and functional complex with its partners E-cadherin and Met might be a hallmark of acquisition of hepatocytic maturation and polarity. Indeed, antisense ablation of β-catenin promoted a more immature cell type that continued to coexpress stem cell and mature hepatocyte markers.13 Additionally, studies using matrigel in primary hepatocyte cultures showed an increase in membranous complexes as a part of the differentiation process.15 This indicates that hepatocyte maturation might be a function of cell–cell adhesion properties of β-catenin with some contribution from its transcriptional coactivator function. The latter function might be associated with the function of β-catenin in regulating expression of genes involved in hepatocyte maturation, such as the cytochrome P450s, shown recently to be tentative targets of the Wnt/β-catenin pathway in liver.16, 17
Likewise, β-catenin has a role in biliary specification during liver development as shown in two studies. β-Catenin antisense ablation in E10 liver cultures led to an absence of casein kinase-positive biliary cells.13 Conversely, growth in Wnt3a-conditioned media showed survival and proliferation of predominantly 19–positive cells as compared with the control media or Wnt3a-conditioned media containing soluble Frizzled-related protein-1 (sFRP1), a Wnt inhibitor.18 Thus, β-catenin appears to play multiple important roles during hepatic morphogenesis based on in vitro and limited in vivo analysis, although whether such observations would hold true during in vivo gene targeting approaches for β-catenin or upstream effectors remains to be seen.
Role of β-Catenin in Normal Liver Growth
The first month after birth is characterized by a postnatal spurt in hepatic growth. Mice overexpressing a stable-mutant or full-length-β-catenin displayed a threefold to fourfold or 15% increase in liver size because of increased hepatocyte proliferation, respectively.19, 20 Interestingly, none of the conventional targets appear to be upregulated in any of these transgenic mice. Conversely, the β-catenin conditional null mice show decrease in their liver size within the first month after birth, where decreased basal hepatocyte proliferation is evident. Decreased cyclin D1 is seen in the β-catenin–deficient livers, and this deficit became more pronounced during states requiring de novo cyclin D1 synthesis such as during liver regeneration after partial hepatectomy.17 These observations suggest a clear role of β-catenin in normal liver growth.
β-Catenin During Liver Regeneration
In light of β-catenin's role in liver growth, multiple studies have shown its role in optimal liver regeneration. Much of this endeavor has used the two-thirds partial hepatectomy model, which is an in vivo system that has been widely used in the study of physiological and regulated liver growth. In rats, after partial hepatectomy, β-catenin protein is increased within minutes of hepatectomy as mediated by an epigenetic or post-translational mechanism that is at least initially transcription independent.21 Although this increase is transient, lasting less than 15 minutes, the accompanying β-catenin nuclear translocation is sustained and can still be found in the nucleus for up to 48 hours after hepatectomy. β-Catenin knockdown through the use of antisense in this model led to a significant decrease in liver weight to body weight ratio because of decreased cellular proliferation.22 Additionally, HGF signaling is an important player in early liver regeneration, and an increase in levels of HGF and activation occurs early after partial hepatectomy, which may have consequences on the Met-β–catenin complex as well.23 These studies suggest that β-catenin might be an early player in regeneration, potentially initiating a cascade of events that are important for successful liver regeneration. Indeed, some downstream targets of this pathway such as cyclin D1, c-myc, urokinase-type plasminogen activator receptor, matrix metalloproteinases, epidermal growth factor receptor, and others are known to be upregulated concurrently during nuclear localization of β-catenin.24 Similar findings were recently reported in mice, where conditional β-catenin knockout mice exhibit a 24-hour delay in peak regeneration after hepatectomy, at which time the hepatocyte proliferation came back with a “vengeance.”17 This study identifies factors such as IL-6, platelet-derived growth factor receptor alpha, decorin, and others that might be compensating for β-catenin loss and would need to be confirmed by additional studies.
Liver Zonation and Metabolism
Although the liver has long been known to be structurally divided into different zones that express specific proteins and partake in differing metabolic functions, not much study has been done to identify the molecular pathways involved in liver zonation. Benhamouche et al.25 recently reported a study suggesting that the interaction between β-catenin and APC may play a critical role in this zonation process. They showed that APC is highly expressed in the periportal region, where β-catenin activation and signaling is quite low, whereas in the perivenous region APC expression was absent and β-catenin activation was relatively high with subsequent β-catenin target gene expression. When APC was knocked out, periportal hepatocytes took on a similar expression profile to that of perivenous hepatocytes, further implicating APC as a key player in zonal regulation of liver physiology. These findings can be extended to the study of β-catenin's role in controlling gene expression of proteins important in ammonia metabolism. In fact, three genes involved in glutamine metabolism encoding glutamine synthetase, ornithine aminotransferase, and the glutamate transporter GLT-1 are all targets of Wnt/β-catenin signaling in transgenic and knockout mice.16, 17, 26 Likewise, overexpression of glutamine synthetase was noted selectively in mouse liver tumors that contained β-catenin mutations.27 The recent study on β-catenin regulation in zonation of the liver corroborates these findings as perivenous hepatocytes, where β-catenin is highly active, are important in glutamine synthesis while periportal hepatocytes are important in urea formation. With such relationships clearly implicated here, further studies, especially investigating inborn errors of metabolism and hepatic cancer, would be of great interest.
Oxidative Stress and β-Catenin
Funato et al.28 recently reported an important relationship between oxidative stress and β-catenin.28 Likewise, studies in C. elegans have identified β-catenin signaling in a protective response to oxidative stress leading to FOXO target gene expression, with additional implications in cell cycle regulation.29 Examining any role of β-catenin in regulating hepatic response to oxidative stress will be important. Indeed, we and others have shown that β-catenin regulates many key players that are canonically associated with regulating oxidative stress in the liver, including various cyctochrome P450s (CYP) and Glutathione S-transferases.16, 17 With oxidative stress being a key player in many pathological conditions of the liver such as alcoholic and non-alcoholic steatohepatitis, it will be crucial to come to a better understanding of the impact of β-catenin on this process. Studies addressing this issue are anticipated over the next several years.
Wnt Signaling and Hepatic Fibrosis
Lately, the role of Wnt/β-catenin signaling is also becoming evident in stellate cells. Although expression of β-catenin in cell–cell contacts of stellate cells is known, nuclear and cytoplasmic β-catenin was reported during liver regeneration more recently.22, 30 Stellate cell activation is implicated in liver fibrosis in multiple conditions including alcoholic steatohepatitis, non-alcoholic steatohepatitis, and viral hepatitis and as a consequence of metabolic liver diseases. Studies are now beginning to investigate the role of Wnt/β-catenin signaling in stellate cell activation. Recently, DNA microarray analysis of quiescent and activated rat hepatic stellate cells showed upregulation of genes involved in the non-canonical Wnt pathway, but an absence of β-catenin activation.31 Although this needs to be further characterized, another study examining absence or presence of all Wnt and frizzled genes showed no difference in their overall expression in active or resting stellate and Kupffer cells. However, this study identified the presence of Fzb (sFRP1) only in the activated cells, suggesting modulation of Wnt signaling during stellate or Kupffer cell activation.3 Additional analysis would be imperative and would have strong clinical implications because of limited existing available therapeutic opportunities.
Wnt/β-Catenin in Hepatic Tumors
Aberrant activation of the Wnt/β-catenin pathway has been implicated in tumor of multiple tissues, including the brain, breast, colon, skin, and, of relevance to this review, the liver. In fact, β-catenin appears to be an active player in hepatoblastomas, benign liver neoplasms, hepatocellular carcinoma, and cholangiocarcinomas.
Hepatoblastoma is the most common malignant liver tumor found in pediatric populations, with the incidence being highest in populations suffering from familial APC.32 Nuclear and cytoplasmic localization of β-catenin were reported in 90% to 100% of all hepatoblastomas, familial and sporadic, because of mutations in APC, CTNNB1, Axin1, and Axin2, clearly demonstrating its role in both types of hepatoblastoma.33–35 However, additional studies on hepatoblastoma are warranted, especially since the identification of activation of β-catenin normally during postnatal hepatic growth. It will be critical to identify subgroups within the hepatoblastoma population whose disease is directly affected by β-catenin activation versus those for which the finding of β-catenin activation is coincidental. Such correlation would make β-catenin activation a surrogate marker in these patients as well as having implications for defining prognosis.
Focal Nodular Hyperplasia and Hepatic Adenoma.
The Wnt/β-catenin pathway has been examined in several rare benign liver neoplasms. Although aberrant β-catenin activity was not noted in focal nodular hyperplasia, analysis demonstrated abnormal cytoplasmic or nuclear localization of β-catenin in 30% of hepatic adenomas from patients.36 A more recent analysis showed mutation of β-catenin in only 12% of adenomas, but 46% of these adenomas progressed to HCC.37 This finding suggests that development of aberrant activity in the Wnt/β-catenin pathway is an important step toward progression to HCC. This study also necessitates examination of its role in a step-by-step model of HCC as well as in oval cell activation models.
In addition to being linked to hepatoblastoma and benign liver neoplasms, the Wnt/β-catenin pathway is an important player in progression from hepatic adenoma to HCC (Fig. 4). Also, 20% to 90% of HCCs display β-catenin activation because of diverse mechanisms that include mutation in genes encoding for β-catenin or CTNNB1,38, 39AXIN-1.35, 40 and AXIN-2,35 as well as frizzled-7 upregulation41 and GSK3β inactivation.42 Interestingly no study has examined the status of any of the Wnts in HCC. In mice, liver-specific deletion of APC induces β-catenin stabilization and increased HCC.43 Similarly, zebrafish that is heterozygous for the APC function also exhibits hepatic neoplasia, secondary to Wnt/β-catenin activation.44 Likewise, transgenic mouse models overexpressing c-myc or transforming growth factor beta show mutation and/or nuclear translocation of β-catenin in liver tumors.45 Furthermore, β-catenin activation provides additional growth and invasive advantages in a model of liver cancer promotion with phenobarbital in c-myc/transforming growth factor alpha transgenic mice. Interestingly, simultaneous mutation of β-catenin and H-ras leads to 100% incidence of HCC in mice.46 These findings have clear implications in that β-catenin activation is likely an initiating or contributory factor in a significant subset of HCCs.
In HBV-positive cases of HCC, hepatitis B virus encoded X antigen is associated with decreased expression of E-cadherin and accumulation of β-catenin in the cytoplasm and/or nucleus and upregulation of the hepatitis B virus encoded X antigen effector URG11 leads to increased activation of β-catenin.47 This demonstrates multiple ways of regulation of β-catenin in just HBV pathology.
Studies have linked Wnt/β-catenin signaling to HCC that occurs in HCV patients. Interestingly, the frequency of β-catenin gene mutations was approximately double in HCC occurring in HCV patients as compared with other causes.48 The mechanism of this predilection remains obscure, calling for further investigation. An additional observation to this end has been an overall decrease in the incidence of HCC in HCV-positive patients who have been on chemotherapy for their viral illness.49 Although this might be attributable to decreased viral load in response to the antiviral agents, quite possibly these drugs might have additional modulatory effects on signal transduction such as the Wnt/β-catenin pathway, which should be investigated.
Several reports have implicated aberrant Wnt/β-catenin signaling in a subset of cholangiocarcinomas, the most common tumor of the biliary tree. Reduced expression of β-catenin and E-cadherin at the membrane along with nuclear localization is seen in a subset of tumors, which is also known to be associated with poorer histological differentiation.50–52 Although mutations in the β-catenin gene are evident only in a small subset of cholangiocarcinomas, as to their aberrant cellular localization in these tumors, additional studies examining such mechanisms as tyrosine phosphorylation-dependent β-catenin activation or Wnt/Frizzled deregulation would be necessary to clarify a bona fide role for β-catenin in these tumors. This is of further interest, especially because its role in biliary development is beginning to be understood.18, 53
β-Catenin: A Potential Therapeutic Target in HCC?
Additional studies are needed to understand the contribution of individual Wnt and frizzled genes to hepatocarcinogenesis as positive or negative regulators. However, although deregulation of β-catenin does seem to play a significant role in pathogenesis of various hepatic tumors, additional studies in conditional β-catenin knockouts will conclusively address this role. Currently mounting evidence links beta-catenin stabilization to tumor cell proliferation, survival, and perhaps “cancer stem cell” renewal, in animal models and patients, secondary to a multitude of factors that converge on β-catenin (Fig. 4). Such observations make β-catenin an attractive and novel therapeutic target. Indeed, therapeutic targeting of Wnt/β-catenin in chemoprevention or as a possible treatment of cancers is beginning to be discussed.54, 55 It is in fact quite likely that a subset of liver tumors or other cancers that are characterized by β-catenin activation will be reasonable candidates for such therapy. In addition, β-catenin suppression for chemoprophylaxis in disease states such as hepatic fibrosis, hepatitis, and cirrhosis might be achievable and highly relevant. In liver tumor cells, we have found that targeting β-catenin with an experimental drug, R-Etodolac, leads to decreased proliferation and survival of two human hepatoma cell lines (HepG2 and Hep3B).56 This drug is an enantiomer of Etodolac, an NSAID used in inflammatory diseases of the joints. R-Etodolac lacks any cox-inhibitory activity or its notorious side effects and is already in phase II trials for refractory chronic lymphocytic leukemia. Further studies are clearly warranted for examining in vivo efficacy of R-Etodolac, as well as other potential compounds, for targeting β-catenin for chemoprevention or treatment of eligible liver tumors.
Also of relevance to chemotherapy, it was recently discovered that several CYP isoenzymes are upregulated in liver tumors harboring β-catenin mutations.57 Likewise, hepatocyte-specific knockout of β-catenin in mice leads to a loss of expression of two CYP proteins.16, 17 These findings have clear implications because many CYPs either activate or inactivate common anticancer drugs; thus, further studies will be essential to address the clinical impact of these findings.
As is evident, the Wnt/β-catenin pathway plays an important role in liver biology, and at this stage we have only begun to scratch the surface in examining the impact of this signaling in liver health and disease. It will be essential to fully understand areas in which Wnt/β-catenin signaling is normally active and relevant and where its aberrant activation is in fact causing a deleterious effect. Although it is important to preserve the former, such active signaling might be rewarding in regenerative therapies, stem cell therapies, hepatocyte transplantation, and bioartificial liver devices for applications in failing livers. Conversely, it will be of high translational and clinical relevance to begin targeting Wnt/β-catenin signaling in situations of aberrant activation, to explore novel therapeutic and prophylactic opportunities. However, one theme that has emerged out of the global analysis from many laboratories is that β-catenin is good in moderation, and excess or too little of it is troublesome!