Potential conflict of interest: Nothing to report.
Recent studies have suggested that β-catenin is involved in the regulation of hepatocyte proliferation in multiple contexts, including organ development and tumorigenesis. We explored the role of β-catenin during liver regeneration using T cell factor/lymphoid enhancer factor (TCF/LEF)-reporter mice (TOPGal mice) and liver-specific β-catenin knockout mice. Liver-specific β-catenin knockout mice showed a delayed onset of DNA synthesis after hepatectomy, whereas recovery of liver mass was not affected. Among putative β-catenin target genes examined, the induction of Ccnd1 expression was reduced, whereas the expression of Myc and Egfr was unaffected. Furthermore, cyclin D1 protein levels were not induced, and the expression of cyclins A, E, and proliferating cell nuclear antigen was delayed. Intriguingly, the analysis of TOPGal mice showed that hepatocytes with active TCF/LEF transcription are confined to the pericentral zone and are not increased in number during regeneration, indicating an uncoupling between β-catenin/TCF signaling activity and hepatocyte proliferation. Conclusion: Our results indicate that β-catenin is critical for the proper regulation of hepatocyte proliferation during liver regeneration; however, the activity of β-catenin/TCF signaling does not correlate with hepatocyte proliferation, suggesting that this regulation might be indirect/secondary. (HEPATOLOGY 2007;45:361–368.)
The liver has a remarkable capacity to regenerate. Under normal conditions, hepatocytes rarely divide; however, hepatocytes rapidly proliferate in response to loss of liver mass due to surgical removal, chemical injury, or infection.1–3 Partial hepatectomy, involving the removal of two thirds of the liver, has been widely used as a rodent model to study the molecular mechanisms of liver regeneration.4, 5 In hepatectomized rodents, most hepatocytes enter the cell cycle and repopulate the original liver mass after only one or two cell divisions. Recent studies have shown that multiple pathways, including cytokines, growth factors, and metabolic networks, are involved in this regeneration process.1–3
β-Catenin is an integral part of the canonical Wnt signaling pathway known to regulate cell proliferation, differentiation, and stem cell maintenance in a wide variety of tissues.6–9 Wnt ligands are secreted proteins that bind to cognate Frizzled receptors present on the cell membrane of Wnt-responsive cells. In the absence of Wnt signals, cytoplasmic β-catenin is phosphorylated by casein kinase I and glycogen synthase kinase-3β, modifications that trigger rapid proteosomal degradation of β-catenin. On stimulation by Wnt ligands, cytoplasmic β-catenin is stabilized and translocates into the nucleus, where it forms active transcription complexes with members of the T-cell factor/lymphoid enhancer factor (TCF/LEF) transcription factor family.6–9
There is accumulating evidence that suggests a crucial role for β-catenin in the regulation of hepatocyte proliferation in settings outside of liver regeneration. During liver development, the level of nuclear/cytoplasmic β-catenin accumulation correlates with the proliferative activity of hepatocytes,10 and modulation of β-catenin expression alters hepatocyte proliferation. Expression of a constitutively stabilized form of β-catenin results in increased hepatocyte proliferation and hepatomegaly.11, 12 Deletion of APC, a component of the β-catenin degradation complex, in the adult liver also stabilizes β-catenin and induces hepatomegaly.13, 14 Conversely, suppression of β-catenin expression by antisense oligonucleotides in ex vivo liver organ culture results in decreased cell proliferation and increased apoptosis of hepatocytes,15 whereas loss of β-catenin in hepatocytes in vivo substantially reduces liver mass.16
During liver regeneration, a potential role for β-catenin in hepatocyte proliferation has also been reported. Monga et al.17 showed that Wnt/β-catenin signaling is regulated after hepatectomy in rats. The same group also reported that knocking down β-catenin by a morpholino oligonucleotide caused impaired rat liver regeneration.18 To further clarify the role of β-catenin in liver regeneration, we studied 2 transgenic mouse models: liver-specific β-catenin knockout and TOPGal reporter mice. These two transgenic models allow for efficient blocking of β-catenin/TCF signaling in hepatocytes and monitoring of signaling activity in vivo, respectively. Our data support the requirement of β-catenin for normal liver regeneration. However, our findings indicate that the β-catenin/TCF pathway is not an immediate regulator of hepatocyte proliferation during liver regeneration.
Mice used in the this study were maintained in the barrier facility according to protocols approved by the Committee on Animal Research at the University of California, San Francisco. Mice were kept on a 12-hour light/dark cycle.
Generation of liver-specific β-catenin knockout mice (Albumin-Cre;Ctnnb1loxP/loxP; referred to as β-catenin knockout mice hereafter) and TCF/LEF reporter mice (TOPGal mice) has been described.16, 19 The β-Catenin knockout mouse colony was maintained by mating Albumin-Cre;Ctnnb1loxP/loxP with Ctnnb1loxP/loxP mice to obtain β-catenin knockout mice and control mice from the same litters. Genotyping was performed by PCR analysis using genomic DNA isolated from the tail tip as described.20, 21
Male and female TOPGal 7- to 9-week-old mice were examined in the current study. For liver-specific β-catenin knockout mice, 7- to 9-week-old, male Albumin-Cre;Ctnnb1loxP/loxP mice were used whereas Ctnnb1loxP/loxP littermates served as controls. Two-thirds hepatectomy was performed as previously described.4 Briefly, mice were anesthetized with isoflurane. After opening the upper abdomen, the three most anterior lobes were individually ligated and resected. All procedures were done in the morning to avoid the effect of diurnal variations. Animals were injected with BrdU 50 μg/g 2 hours before killing at the indicated postoperative time points. Four mice were studied per time point for each genotype.
Mice were perfused with ice-cold fixative [2% paraformaldehyde and 0.25% glutaraldehyde in phosphate-buffered saline (PBS)]. Liver samples were fixed for 90 minutes at 4°C. Tissues were then washed with PBS for 1 hour and stained with LacZ solution (1 mg/ml X-gal, 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, 2 mM MgCl2, 0.02%(v/v) NP40, and 0.01%(w/v) sodium-deoxycholate in PBS) at room temperature for 24 hours. After staining, tissues were post-fixed with buffered formalin, embedded in paraffin, sectioned, and counterstained with nuclear fast red.
Liver tissue samples were fixed overnight in zinc-containing neutral-buffered formalin, embedded in paraffin, and cut into 5-μm-thick sections. Deparaffinized and rehydrated slides were incubated in 1N HCl for 8 minutes at 65°C. Slides were blocked with 0.3% H2O2 for 20 minutes, washed in PBS, and blocked with 1% (w/v) bovine serum albumin in PBS. Slides were incubated with rat anti-BrdU antibody (Serotec, Raleigh, NC; BU1/75, 1:200 dilution) for 1 hour at room temperature. Biotinylated anti-rat antibody (Jackson Immunoresearch, West Grove, PA) was used as a secondary antibody at a 1:200 dilution and visualized with peroxidase reaction performed with the ABC Vectastain kit (Vector Laboratories, Burlingame, CA). 3–3′-Diaminobenzidine tetrahydrochloride was used as a chromogen.
Total RNA was prepared from liver tissue samples using Trizol (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. Reverse-transcription reaction was performed using Superscript First-Strand Synthesis System (Invitrogen). PCR reactions were performed in a 25 μL reaction mixture containing 1× SYBR Green PCR master mix (Applied Biosystems, Foster City, CA) and 300 nM of each primer. Amplification was performed by initial polymerase activation for 10 minutes at 95°C, and 40 cycles of 94°C for 30 seconds followed by elongation for 1 minute at 60°C. The following primers were used: GAPDH: TGTGTCCGTCGTGGATCTGA and ACCACCTTCTTGATGTCATCATACTT; Axin2: GCCAATGGCCAAGTGTCTCT and GCGTCATCTCCTTGGGCA. Primers for Ccnd1, Myc, and Egfr were described previously.16 To exclude contamination with nonspecific PCR products, melting curve analysis was applied to all final PCR products after the cycling protocol. RNA samples without reverse-transcription were also subjected to PCR reaction to exclude contamination of genomic DNA. Expression of liver genes was compared with the expression level of GAPDH.
Tissue samples were homogenized in RIPA buffer [50 mM Tris-HCl, pH 7.4, 1% (v/v) NP40, 0.1% (w/v) SDS, 0.25% (w/v) Na-deoxycholate, 1 mM EDTA) with phosphatase inhibitors (1 mM Na-orthovanadate, 40 mM NaF, 10 mM glycerophosphate, and 5 mM pyrophosphate) and protease inhibitors (Complete; Roche, Basel, Switzerland). Equal amounts of proteins were electrophoresed on 10% or 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis gels and transferred to PVDF membrane (Bio-Rad, Hercules, CA), and processed for immunoblotting with antibodies for β-catenin (clone 14; BD Bioscience, San Diego, CA), GAPDH (sc-25778; Santa Cruz, Santa Cruz, CA), cyclin A (sc-751; Santa Cruz), cyclin D1 (sc-718; Santa Cruz), cyclin E (07-687; Millipore, Billerica, MA), and proliferating cell nuclear antigen (PC-10; Santa Cruz). Horseradish peroxidase–conjugated anti-mouse or rabbit secondary antibody was used at 1:5,000 dilution and detected using ECL (Amersham Biosciences).
The results are presented as the mean ± standard deviation (SD). Statistical significance was determined by Student two-tailed t test, with a P value of <0.05 considered significant.
Efficient Disruption of the β-Catenin/TCF Pathway in β-Catenin Knockout Mice During Liver Regeneration.
To determine whether β-catenin is required during liver regeneration, we performed two-thirds hepatectomy on β-catenin knockout mice (Albumin-Cre;Ctnnb1loxP/loxP) and control littermates (Ctnnb1loxP/loxP). Liver samples were harvested after indicated time points and protein extracts were analyzed by Western blotting to examine the changes in β-catenin expression levels and to verify the efficient suppression of β-catenin during liver regeneration in the knockout mice. Levels of β-catenin expression remained virtually constant during regeneration in control mice (Fig. 1A). Expression of β-catenin was consistently low in β-catenin knockout mice compared with control mice at all time points examined. The residual β-catenin expression is likely to be derived from nonparenchymal cells, including endothelial cells and fibroblasts retaining β-catenin expression.16
Next, we performed quantitative PCR against Axin2, a direct and ubiquitous target of TCF/LEF-dependent transcription to determine whether β-catenin/TCF pathway activity changes during liver regeneration.22, 23 Control mice showed a modest induction of Axin2 expression 12 to 96 hours after hepatectomy, indicating a moderate increase in β-catenin/TCF pathway activity during liver regeneration (Fig. 1B). As expected, Axin2 expression remained at a low level in β-catenin knockout mice throughout liver regeneration, confirming efficient disruption of β-catenin/TCF pathway in β-catenin knockout mice.
Distribution of Hepatocytes With Active β-Catenin/TCF Signaling During Liver Regeneration.
To identify hepatocytes with active TCF/LEF-dependent transcription during liver regeneration in wild-type mice, we examined TOPGal mice for LacZ expression after hepatectomy. In TOPGal mice, the bacterial LacZ gene is under control of minimal promoter and optimal TCF/LEF binding sites.19 Because stabilized β-catenin activates TCF/LEF-dependent transcription, LacZ expression marks cells with active β-catenin signaling in this mouse model. TOPGal mice were subjected to partial hepatectomy and examined at 3, 6, 12, 24, 36, 48, 72, and 120 hours and 1 and 2 weeks after the procedure.
In mice that did not undergo hepatectomy, β-galactosidase-positive hepatocytes were confined to a few layers of hepatocytes surrounding central veins (Fig. 2A). This distribution of LacZ-positive cells was unchanged throughout hepatectomy-induced liver regeneration, and no detectable staining was observed in nonparenchymal cells (Fig. 2B-D). These results indicate that β-catenin/TCF signaling is consistently activated in pericentral hepatocytes, and that hepatectomy does not alter either the number or spatial distribution of cells with active TCF/LEF transcription within the liver lobules.
Normal Recovery of Liver Mass in β-Catenin Knockout Mice.
Next, we followed liver mass recovery after hepatectomy in β-catenin mutant mice. As previously reported, the baseline liver mass of β-catenin knockout mice is significantly smaller than that of control mice16 (Fig. 3A). Resection of the three most anterior lobes resulted in liver mass reduction of 62% and 66% in β-catenin knockout and control mice, respectively. To facilitate comparison between wild-type and mutant mice, the liver/body weight ratio at each time point post-hepatectomy was normalized to the pre-hepatectomy body/liver weight ratio (Fig. 3B). The results showed that loss of β-catenin does not affect the extent or the timing of liver mass recovery after partial hepatectomy.
Delayed DNA Synthesis After Partial Hepatectomy in β-Catenin Knockout Mice.
Wnt/β-catenin signaling regulates cell proliferation in a number of different cell types. To explore whether loss of β-catenin affects hepatocyte proliferation after hepatectomy, DNA replication was examined by BrdU incorporation 24 to 96 hours after hepatectomy. In control mice, the number of BrdU-labeled cells peaked at 36 to 48 hours after hepatectomy, and significantly decreased after 72 hours (Fig. 4A, B, D). In contrast, BrdU labeling peaked at 48 to 72 hours after hepatectomy in β-catenin knockout mice (Fig. 4A, C, E), indicating that the absence of β-catenin delays DNA replication in hepatocytes by 12 to 24 hours. Immunohistochemical analysis showed that BrdU-positive hepatocytes were widely distributed throughout the liver lobules in both β-catenin knockout and control mice. Thus, although we find a temporal delay in DNA replication in β-catenin knockout mice compared with controls, the spatial distribution of replicating cells was unaffected in mutant mice.
Expression of Putative β-Catenin Target Genes in β-Catenin Knockout Mice.
To explore the underlying mechanism for delayed DNA replication in β-catenin knockout mice, we used quantitative reverse transcription PCR to investigate expression of three putative β-catenin downstream genes, Ccnd1 (encoding cyclin D1), Myc, and Egfr (encoding epidermal growth factor receptor) (Fig. 5). As previously reported, Ccnd1 expression was induced 24 hours after hepatectomy in control mice.24, 25 However, Ccnd1 expression was significantly lower in β-catenin knockout mice 36 to 96 hours after hepatectomy. β-Catenin knockout mice tended to show lower expression of Myc at 1 hour after hepatectomy, but the differences were not statistically significant. Expression of Egfr was not altered in β-catenin knockout mice.
Expression of Cell Cycle–Related Proteins in β-Catenin Knockout Mice.
We examined expression of cell cycle–related proteins during regeneration to further characterize delayed DNA replication in hepatectomized β-catenin knockout mice (Fig. 6). Cyclin D1 was induced 36 to 48 hours after hepatectomy in control mice. In contrast, cyclin D1 was not induced in β-catenin knockout mice at any time point analyzed, consistent with the pattern of Ccnd1 expression observed in these mice. Cyclin E was induced 36 to 48 hours postprocedure in control mice, but showed delayed and weaker expression in β-catenin knockout mice. Cyclin A and proliferating cell nuclear antigen peaked at 48 hours in control mice, but peaked at 72 hours in knockout mice. The 24-hour delay parallels the delay in DNA replication.
Recent studies have significantly improved our understanding of the molecular mechanisms underlying liver regeneration.1–3 Activation of multiple signaling cascades is required for proper regeneration, and intricate interactions between signaling pathways regulate this process. β-Catenin signaling has been implicated in cell proliferation in a number of regenerating tissues, including the liver.17, 18 To investigate the role of this pathway in regenerating liver, we determined whether the level of β-catenin/TCF signaling activity changes during liver regeneration. Previous studies have reported widespread nuclear localization of β-catenin between 5 minutes and 6 hours after hepatectomy in rats,17 indicating that β-catenin/TCF signaling is activated at high levels during early stages of liver regeneration. We monitored activity of this pathway by determining the expression levels of the ubiquitous TCF/LEF target gene Axin2 and the level of β-galactosidase activity in TOPGal mice. Although Axin2 expression is moderately activated after 12 hours in control mice, we failed to detect an early induction of Axin2 or widespread upregulation of LacZ expression in TOPGal mice during liver regeneration. Together with the unchanged levels of β-catenin protein, these data suggest that the β-catenin/TCF pathway is only modestly induced during liver regeneration.
To directly address the question of whether β-catenin is required for liver regeneration after hepatectomy, we analyzed transgenic mice in which β-catenin has been eliminated in hepatocytes. We have previously reported that hepatocyte-specific loss of β-catenin with its target genes does not impair organ development but results in reduced liver mass.16 Comparison of liver mass recovery between control and β-catenin knockout mice showed no apparent difference in timing or extent of liver mass restoration. However, these results conflict with a previous study in rats.18 In this study, a significant delay in recovery of liver mass after hepatectomy was observed on inhibition of β-catenin by injection of anti-β-catenin morpholino oligonucleotides.18 One possible explanation for this discrepancy might be that the anti-β-catenin morpholino oligonucleotides used in this experiment also might suppress β-catenin function in nonparenchymal liver or extrahepatic cells not targeted in our β-catenin knockout mice. Future studies addressing the function of β-catenin in non-hepatocytes would be required to resolve this question.
β-Catenin has been shown to directly control cell proliferation by regulating the expression of cell cycle components, including Ccnd1 and Myc.26–28 We found that both DNA replication and cell cycle progression was delayed by 12 to 24 hours in β-catenin knockout mice. Quantitative PCR showed that induction of Ccnd1 was significantly impaired in β-catenin knockout mice, resulting in reduced expression of the cyclin D1 protein. Because cyclinD1 regulates the G1/S cell cycle transition, the absence of cyclin D1 induction is likely responsible for the delay in DNA synthesis in β-catenin knockout mice. Intriguingly, this delay in DNA synthesis did not affect the recovery of liver mass in mutant mice. A possible explanation for the uncoupling of cell proliferation and mass recovery in β-catenin knockout mice may be that β-catenin plays a minor role in the initiation of protein synthesis. A previous study showed that even in the absence of hepatocyte replication, Skp2 knockout mice recovered liver mass by enlargement of hepatocytes after hepatectomy.29 This finding supports our observation that cell cycle progression and liver mass recovery are not necessarily linked.
Staining for β-galactosidase activity in TOPGal liver revealed that hepatocytes with active TCF/LEF-dependent transcription are confined to the pericentral areas of the liver lobules both at baseline and during liver regeneration. TOPGal mice appear to be a sensitive tool to identify TCF/LEF transcription active cells, as β-galactosidase-positive cells are observed whereas nuclear β-catenin expression, widely recognized hallmark of active Wnt/β-catenin signaling,8 is not detectable in quiescent liver. The localization of LacZ-positive hepatocytes in TOPGal mice is also concordant with the expression pattern of Axin2.14 Furthermore, suppression of the Wnt/β-catenin signaling pathway results in loss of pericentral gene expression.14, 16, 30 Thus, consistent with our results in the TOPGal mice, these studies imply that Wnt/β-catenin signal is predominantly active in pericentral hepatocytes.
Considering the restricted distribution of hepatocytes displaying active TCF/LEF-dependent transcription in liver lobules, if β-catenin/TCF signaling directly activates hepatocyte proliferation, proliferating cells should be observed primarily in the pericentral areas. However, virtually all hepatocytes throughout the liver lobules undergo proliferation in response to partial hepatectomy. One possible explanation for the discrepancy between the distribution of TCF/LEF transcriptional activity versus proliferating hepatocytes is that β-catenin/TCF signaling affects hepatocyte proliferation in a paracrine fashion. If this is the case, β-catenin/TCF signaling would activate expression of certain target genes, presumably coding for soluble proteins, in hepatocytes surrounding the central veins. Secretion of these proteins may then regulate the proliferation of TCF/LEF transcriptionally inactive hepatocytes within the entire lobule. However, no evidence has been presented for the presence of growth-promoting secreted molecules regulated by β-catenin/TCF signaling in the liver.
An alternative explanation would be that β-galactosidase staining of TOPGal mouse liver might only identify cells with high TCF/LEF-dependent transcriptional activity. Lower-level TCF/LEF transcriptional activity, undetectable by β-galactosidase staining in TOPGal mice, might be widely present in liver lobules. The putative presence of low-level Wnt/β-catenin/TCF signaling activity would explain the observation that many of the β-catenin downstream genes, including CYP1A2, CYP2E1, and Rnase4, are expressed outside the areas marked by β-galactosidase activity in TOPGal mice.14, 16, 30 Benhamouche et al.14 recently proposed the presence of a Wnt/β-catenin signaling gradient along the portocentral axis. According to this hypothesis, the differences in Wnt/β-catenin signaling activity contribute to the formation of distinct functional zones in liver lobules. Therefore, levels of TCF/LEF transcription below the limit of detection of nuclear β-catenin or β-galactosidase staining in TOPGal mice may regulate hepatocyte proliferation throughout liver lobules. In addition, other signaling pathways induced by hepatectomy could affect the expression of β-catenin target genes. Indeed, recent studies indicated that known direct β-catenin target genes, Ccnd1, Myc, and Cryptdin, show temporally and spatially distinct expression patterns in colorectal adenomas in which β-catenin/TCF signaling is constitutively upregulated because of Ctnnb1 mutations.31, 32 These results indicate that β-catenin target genes, even if they are direct transcriptional targets, are not regulated solely by β-catenin/TCF signaling but are controlled by synergistic interaction with other signals. Similar interactions between β-catenin/TCF and other signaling pathways likely modulate hepatocyte proliferation in the regenerating liver.
In summary, our study shows that β-catenin plays an important role in directing the proper proliferative response of hepatocytes during liver regeneration. Importantly, however, the portocentral gradient of β-catenin/TCF signaling activity appears to be maintained throughout the regeneration process, and the magnitude of the TCF/LEF-dependent transcription does not directly correlate with hepatocyte proliferation. Our results suggest that β-catenin/TCF signaling is likely to regulate hepatocyte proliferation in cooperation with other signaling cascades.
We thank David Cano for sharing reagents, John P Morris IV for critical reading of the manuscript, and the members of the Hebrok laboratory for stimulating discussions.