Potential conflict of interest: Nothing to report.
Article first published online: 4 DEC 2012
Copyright © 2012 American Association for the Study of Liver Diseases
Volume 56, Issue 6, pages 2344–2352, December 2012
How to Cite
Borude, P., Edwards, G., Walesky, C., Li, F., Ma, X., Kong, B., Guo, G. L. and Apte, U. (2012), Hepatocyte-specific deletion of farnesoid X receptor delays but does not inhibit liver regeneration after partial hepatectomy in mice. Hepatology, 56: 2344–2352. doi: 10.1002/hep.25918
Supported by National Institutes of Health grant P20 RR021940 (U. A. and G. L. G.), AASLD/ALF Liver Scholar Award (U. A.), and National Institutes of Health grant R01 DK031343 and a University of Kansas endowment (G. L. G.).
- Issue published online: 4 DEC 2012
- Article first published online: 4 DEC 2012
- Accepted manuscript online: 22 JUN 2012 08:46AM EST
- Manuscript Accepted: 10 JUN 2012
- Manuscript Received: 17 FEB 2012
- National Institutes of Health. Grant Number: P20 RR021940
- National Institutes of Health. Grant Number: R01 DK031343
Farnesoid X receptor (FXR), the primary bile acid–sensing nuclear receptor, also plays a role in the stimulation of liver regeneration. Whole body deletion of FXR results in significant inhibition of liver regeneration after partial hepatectomy (PHX). FXR is expressed in the liver and intestines, and recent reports indicate that FXR regulates a distinct set of genes in a tissue-specific manner. These data raise a question about the relative contribution of hepatic and intestinal FXR in the regulation of liver regeneration. We studied liver regeneration after PHX in hepatocyte-specific FXR knockout (hepFXR-KO) mice over a time course of 0-14 days. Whereas the overall kinetics of liver regrowth in hepFXR-KO mice was unaffected, a delay in peak hepatocyte proliferation from day 2 to day 3 after PHX was observed in hepFXR-KO mice compared with Cre− control mice. Real-time polymerase chain reaction, western blot and co-immunoprecipitation studies revealed decreased cyclin D1 expression and decreased association of cyclin D1 with CDK4 in hepFXR-KO mice after PHX, correlating with decreased phosphorylation of the Rb protein and delayed cell proliferation in the hepFXR-KO livers. The hepFXR-KO mice also exhibited delay in acute hepatic fat accumulation following PHX, which is associated with regulation of cell cycle. Further, a significant delay in hepatocyte growth factor–initiated signaling, including the AKT, c-myc, and extracellular signal-regulated kinase 1/2 pathways, was observed in hepFXR-KO mice. Ultraperformance liquid chromatography/mass spectroscopy analysis of hepatic bile acids indicated no difference in levels of bile acids in hepFXR-KO and control mice. Conclusion: Deletion of hepatic FXR did not completely inhibit but delays liver regeneration after PHX secondary to delayed cyclin D1 activation. (HEPATOLOGY 2012;56:2344–2352)
Liver regeneration is regulated by a complex network of signals involving cytokines, chemokines, growth factors, and nuclear receptors. 1-3 Previous studies have demonstrated that farnesoid X receptor (FXR), the primary bile acid sensing nuclear receptor, plays a critical role in stimulation of liver regeneration following partial hepatectomy (PHX). 4 FXR is highly expressed in the liver and intestines and plays a central role in maintaining bile acid homeostasis. 5, 6 Deletion of FXR in the whole body as in FXR knockout (FXR-KO) mice results in massive disruption in bile acid homeostasis leading to higher total bile acids, moderate liver injury, increased oxidative stress, and development of spontaneous liver cancers. 7-9 FXR-KO mice also exhibit significant inhibition of liver regeneration following PHX, highlighting the role of FXR in regulation of liver cell proliferation and tissue repair. 4, 10
Recent studies using chromatin immunoprecipitation (ChIP) sequencing have revealed that FXR binds to a distinct set of genes in the liver compared with its binding in the intestine. 11 Further, it is known that coordination between hepatic and intestinal FXR signaling is necessary for maintaining bile acid homeostasis in the body, and the intestinal FXR-mediated pathway is more critical in suppressing bile acid synthesis than a hepatic FXR-mediated pathway. 12 It is clear that FXR signaling in the liver and intestines is distinct, and results obtained from whole body modulation of FXR need to be further refined to understand the underlying molecular mechanisms.
In the current study, we determined the contribution of hepatic FXR to liver regeneration following PHX using hepatocyte-specific FXR knockout (hepFXR-KO) mice. Our data indicate that liver regeneration is regulated by hepatic FXR, but the intestinal FXR likely contributes to regulation of hepatocyte proliferation following PHX. We provide evidence that the gut-liver FXR signaling axis plays a critical role in liver regeneration.
Materials and Methods
Animals, Surgeries, and Tissue Harvesting.
hepFXR-KO mice (FXRloxP/loxP; albumin-cre+) and control mice (FXRloxP/loxP; albumin-cre−) were generated as described before using the cre-lox P system. 12 Two- to 4-month-old male control and hepFXR-KO mice were used in these studies. All animals were housed in facilities accredited by the Association for Assessment and Accreditation of Laboratory Animal Care at the University of Kansas Medical Center under a standard 12-hour light/dark cycle with access to chow and water ad libitum. The Institutional Animal Care and Use Committee at the University of Kansas Medical Center approved all studies.
PHX surgeries were performed as previously described. 13 All surgeries were performed between 9:00 AM and 11:00 AM. Mice were killed at 0, 1, 2, 3, 5, 7, and 14 days post-PHX via cervical dislocation under isofluorane anesthesia, and livers were collected. The liver weights and body weights at the time of animal sacrifice were used to calculate the liver-to-body-weight ratios. The results obtained were the mean of 3-5 different animals per time point.
Part of liver tissue was fixed in 10% neutral buffered formalin for 48 hours, further processed to obtain paraffin blocks and 4-μm-thick sections. A piece of liver was frozen in optimal cutting temperature compound and used to obtain fresh frozen sections. Approximately 100 mg liver tissue was used to prepare fresh nuclear and cytoplasmic protein extracts using the NE-PER Nuclear and Cytoplasmic Extraction Kit (Pierce, Rockford, IL). The remaining liver tissue was frozen in liquid N2, and stored at −80°C until used to prepare RIPA protein extracts.
All primary and secondary antibodies used for western blot analysis were obtained from Cell Signaling Technologies (Danvers, MA) and biotinylated secondary antibodies for immunohistochemistry were purchased from Jackson Immunoresearch (West Grove, PA). The antibodies used and their catalogue numbers are provided in Table 1.
|Name||Cell Signaling Technologies Catalog Number|
Protein Isolation and Western Blotting.
Western blot analysis was performed as reported without any modifications. 14 Briefly, RIPA protein extracts were prepared from frozen liver tissues, and protein concentration was estimated using the bicinchoninic acid method (Thermo Fisher). Fifty micrograms of total protein per well was used for western blot analysis.
Proliferating Cell Nuclear Antigen Immunohistochemistry and Oil Red O Staining.
Real-Time Polymerase Chain Reaction.
Total RNA was isolated from control and hepFXR-KO livers using Trizol method according to the manufacturer's protocol (Sigma, St. Louis, MO) and converted to complementary DNA as described. 14 Messenger RNA (mRNA) levels of various genes was determined using TaqMan-based real-time polymerase chain reaction (PCR) analysis using commercially available TaqMan Gene Expression Assays (Life Technologies [previously Applied Biosystems], Carlsbad, CA) on an Applied Biosystems Prism 7300 Real-Time PCR Instrument as described. 14 Rplp0 gene expression in the same samples was used for data normalization.
Bile Acid Analysis.
Bile acids were analyzed by ultraperformance liquid chromatography/mass spectroscopy (Waters, Milford, MA) as described. 16 Briefly, individual bile acid was separated by a 100 × 2.1 mm (Acquity 1.7 μm) Ultraperformance liquid chromatography BEH C-18 column. The flow rate of the mobile phase was 0.3 mL/minute with a gradient ranging from 2% to 98% aqueous acetonitrile containing 0.1% formic acid in a 10-minute run. MS was operated in a negative mode with electrospray ionization. The source temperature and desolvation temperature were set at 120°C and 350°C, respectively. The capillary voltage and the cone voltage were set at 3.0 kV and 28 V, respectively. Nitrogen was applied as the cone gas (10 L/hour) and desolvation gas (700 liters/hour). Argon was applied as the collision gas.
Data presented in the form of bar graphs show the mean ± SD. To determine statistically significant difference between groups, paired Student t test was used. The difference between groups was considered statistically significant at P < 0.05.
Delayed Peak Proliferation in hepFXR-KO Mice.
The liver weight to body weight ratio analysis in control and hepFXR-KO mice indicated no overall difference in the rate of liver regrowth following PHX over the time course of 0-14 days (Fig. 1A). To determine the change in hepatic cell proliferation following PHX, we stained liver sections from control and hepFXR-KO mice for PCNA and determined the percentage of cells in cell cycle (Figs. 1B and 2). PCNA analysis revealed a delay in peak of cell proliferation in hepFXR-KO mice. In control mice, peak cell proliferation was observed at day 2 after PHX, which was delayed to day 2 after PHX in the hepFXR-KO mice. Cell proliferation declined significantly in both genotypes at day 7 after PHX but a moderate increase in cell proliferation was observed in the hepFXR-KO mice at day 14 after PHX.
Decreased Cyclin D1 Expression in hepFXR-KO Mice After PHX.
To determine the mechanism underlying delayed cell proliferation in hepFXR-KO mice, we determined levels of core cell cycle proteins including cyclin D1, CDK4, and phosphorylated Rb (phospho-Rb). A marked increase in cyclin D1 protein was observed at day 1 after PHX in control mice, which remained increased until day 7 after PHX. A similar increase in cyclin D1 protein expression was observed in hepFXR-KO mice at day 2 after PHX. However, a marked reduction of cyclin D1 protein expression in the livers of hepFXR-KO mice was shown at day 2 compared with control mice (Fig. 3B). The decrease in cyclin D1 protein was accompanied by decreased cyclin D1 gene expression at day 2 after PHX in hepFXR-KO mice (Fig. 3A). Both mRNA and protein expression of cyclin D1 increased at day 3 after PHX and was similar to that observed in control mice. Western blot analysis revealed that levels of CDK4, the catalytic partner of cyclin D1, were moderately decreased at days 2 and 3 after PHX in hepFXR-KO mice compared with control mice (Fig. 3B). These changes were correlated with complete lack of phospho-Rb protein, necessary for initiation of cell cycle progression, in the hepFXR-KO mice at days 1 and 2 after PHX (Fig. 3B). We observed increased phospho-Rb levels at day 3 after PHX in hepFXR-KO mice, coinciding with increased cell proliferation at that time point.
Interestingly, at day 1 after PHX in hepFXR-KO mice, we observed the presence of significant cyclin D1 protein and CDK4, which form the active complex required for phosphorylation of Rb. However, we could not detect any phospho-Rb protein at day 1 after PHX in hepFXR-KO mice by western blot. To decipher these data, we performed co-immunoprecipitation studies on cyclin D1 and CDK4 in control and hepFXR-KO mouse livers at day 2 after PHX (Fig. 3C,D). The data indicate a decreased association of cyclin D1 with CDK4 in hepFXR-KO mice compared with the control mice at day 1 after PHX. These data indicate that lack of phospho-Rb and cell proliferation at day 1 after PHX, despite the presence of cyclin D1 and CDK4, is due to decreased association of these two components of the phosphorylation machinery.
Cell cycle inhibitors such as p21 and p27 are known to inhibit cyclin D1-CDK4 complex activity resulting in decreased phosphorylation of Rb. Therefore, we determined levels of p21 and p27 protein in control and hepFXR-KO mice over a time course after PHX. Whereas the data revealed a difference in the pattern of p21 and p27 expression following PHX between control and hepFXR-KO mice, no correlation was found in the delayed cell proliferation in hepFXR-KO mice livers and the expression of p21 or p27.
Delayed Activation of Hepatocyte Growth Factor–Mediated Signaling in hepFXR-KO Mice After PHX.
Hepatocyte growth factor (HGF) is one of the primary mitogens critical for proper liver regeneration. HGF binds to its receptor c-Met and initiates a cascade of promitogenic signaling. We determined the changes in HGF signaling axis in control and hepFXR-KO mouse livers following PHX. A marked increase in HGF mRNA was seen at day 2 after PHX in the control mice consistent with previous reports. In contrast, HGF mRNA peaked at day 3 after PHX in the hepFXR-KO mice (Fig 4A). Furthermore, a significantly higher protein expression of c-Met was observed at days 3 and 5 after PHX in the hepFXR-KO livers compared with control mice (Fig. 4B). Moreover, we determined the activation of three major downstream MAPK mediators—namely, extracellular signal-regulated kinase 1/2 (ERK-1/2), AKT, and p38 kinases—and the data indicated a marked increase in activated (i.e., phosphorylated) forms of AKT and ERK-1/2 kinases in hepFXR-KO mice at day 3 after PHX. We did not observe any difference in p38 kinase expression and activation between control and hepFXR-KO mice at any time point (Fig 4C). Interestingly, an increase in both total and phosphorylated forms of AKT protein was observed after PHX, which was significantly higher in hepFXR-KO mice (Fig. 4E). The ratio of phosphorylated to total AKT indicated a significant activation at day 1 after PHX in control mice, which was delayed until day 3 after PHX in the hepFXR-KO mice (Fig. 4F). Finally, we observed a delayed increase in two well-known promitogenic genes, c-Myc (Fig. 4B) and c-Jun (Fig. 4D), at day 3 after PHX in hepFXR-KO mice.
Delayed Acute Fat Accumulation Following PHX in hepFXR-KO Mice.
Previous studies have demonstrated that acute fat accumulation in hepatocytes of the regenerating livers is critical for proper hepatocyte proliferation during liver regeneration. 17, 18 Because we observed changes in fat droplet accumulation in hepFXR-KO mice, we further determined the degree to which FXR deficiency affects acute fat accumulation. The results indicate an extensive lipid droplet accumulation in control mouse livers at day 1 after PHX, which disappeared at later time points. Fat accumulation was mostly absent in the hepFXR-KO mouse livers at day 1 after PHX, but increased moderately at day 2 after PHX. The hepFXR-KO mouse livers have extensive fat accumulation at day 3 after PHX, which is delayed for day 2 compared with the control mice (Fig. 5A). To further investigate this delay in acute fat accumulation in hepFXR-KO mice, we determined the mRNA expression of several genes involved in fat mobilization and metabolism. Significant changes were shown in two genes associated with increased fat accumulation following PHX: CFD and aP2 (Fig. 5B,C). The data indicate that CFD, a gene associated with fat uptake in the liver following PHX, increased significantly in control mice at day 1 after PHX, but not in hepFXR-KO mice at any of the time points studied. Further, mRNA levels of aP2, a gene associated with adipocyte fat mobilization, were increased significantly in the hepFXR-KO mice at days 2 and 3 after PHX coinciding with increase in fat accumulation.
No Significant Difference in Bile Acid Levels in Control and hepFXR-KO Mice After PHX.
One of the main physiological changes observed in whole body FXR-KO mice is the 4-fold increase in total bile acid concentration, which remain elevated in the FXR-KO mice after PHX. To study whether the delay in liver regeneration in hepFXR-KO mice is related to change in bile acid concentration or isoforms, we determined hepatic concentrations of major hepatic bile acids, including α-MCA, β-MCA, ϵ-MCA, T- β-MCA, and TCA. However, no significant difference was observed in these bile acid levels between control and hepFXR-KO mice (Fig. 6).
FXR is a nuclear receptor involved in a variety of pathophysiological processes, including bile acid homeostasis, pathogenesis of diseases (insulin resistance, metabolic syndrome, and hepatocellular carcinoma), and liver regeneration. 5, 6, 19-21 Previous studies have demonstrated that FXR plays a critical role in liver regeneration after PHX as well as after chemical induced injury. 4, 10 FXR is expressed both in the gut and in the liver, and it is known that coordinated gut-liver signaling by FXR is crucial for many of its effects, including regulation of bile acid synthesis. 12 Recent studies have demonstrated that hepatic FXR and gut FXR regulate a distinct set of target genes. Further, development of organ-specific knockout mice has allowed investigators to determine the relative contribution of hepatic versus gut FXR in the regulation of pathophysiological processes. 11 In the present study, we determined the role of hepatic FXR in regulating liver regeneration after PHX. Our data indicate that hepatic FXR plays a limited role in regulation of liver regeneration and the regenerative defects observed in whole of FXR-KO mice are likely due to disruption of gut-liver FXR signaling.
Our studies demonstrated a 24-hour delay in peak proliferation following PHX in the hepFXR-KO mice. This delay was secondary to decreased cyclin D1 expression in the hepFXR-KO mice. Interestingly, ChIP sequencing analysis of FXR in the liver indicated that FXR has a potential binding site on the cyclin D1 promoter (data not shown). However, we did not observe a decrease in cyclin D1 expression at 1 day after PHX, but observed a significant decline in cyclin D1 expression only at day 2 after PHX in hepFXR-KO mice. These data suggest that FXR may not regulate initial activation of cyclin D1 but may be involved in sustained cyclin D1 activation following PHX.
HGF is a primary mitogen for hepatocytes and plays a critical role in hepatocyte proliferation during regeneration. 2, 22 HGF binds to its receptor c-Met, which activates a plethora of downstream promitogenic kinase signaling. 23 Our data indicate a significant delay in HGF expression and downstream signaling consistent with delayed cell proliferation in the hepFXR-KO mice. However, the exact role of FXR in regulating HGF signaling currently remains unknown. Previous ChIP sequencing studies on FXR target genes in the liver indicate that c-Met is a potential target of FXR in the liver. 11 Other studies have shown that HGF can stimulate its own gene expression via binding to c-Met and downstream activation of the AP1 transcription factor (a complex of c-Fos and c-Jun). 24, 25 Our data indicate a delayed increase in both c-Met and c-Jun (a part of AP1) in hepFXR-KO mice after PHX coinciding with increased HGF gene expression. It is plausible that the delay in HGF expression observed in the hepFXR-KO mice may be secondary to delayed activation of c-Met-AP1 signaling axis resulting in delayed transcriptional upregulation of HGF. Additionally, HGF expression can be stimulated by cytokines such as interleukin-6, which is also a putative target of FXR. 11
A serendipitous observation in our studies was the delayed fat accumulation in the hepFXR-KO mice after PHX. Previous work has demonstrated the acute fat accumulation within first 24 hours after PHX is essential for timely cell proliferation and liver regeneration. 17, 18 This acute fat accumulation is shown to be associated with genes such as aP-2, CFD, and others. Our results indicate that delay in fat accumulation in hepFXR-KO mice coincides with delayed cell proliferation and is consistent with delayed induction of aP-2 and CFD, the genes involved in fat accumulation after PHX. These data suggest that FXR plays a crucial role in hepatic fat accumulation after PHX and supports the energy requirement for the ongoing cell proliferation. However, ChIP sequencing analysis did not show any of these genes as potential targets of FXR, indicating an indirect role of FXR in hepatic fat accumulation.
Further, we investigated the degree to which delayed regeneration in hepFXR-KO mice is due to changes in bile acids. Previous studies have shown that whole body FXR-KO mice have 4-fold higher total bile acids, which further increase after PHX. 4, 7 Interestingly, hepFXR-KO mice do not have the increase in bile acids observed in whole body FXR-KO mice. 12 Our data indicated no difference in changes in hepatic bile acids in control versus hepFXR-KO mice, ruling out the involvement of bile acids in the observed delay in regeneration.
In conclusion, our study indicates that FXR plays a multifaceted role in the regulation of hepatocyte proliferation, including stimulating cyclin D1, modulating HGF signaling, and mobilization of acute hepatic fat accumulation necessary for proper liver regeneration. These data indicate that regulation of liver regeneration after PHX by FXR is likely dependent on the gut-liver FXR signaling axis and not on the hepatic FXR alone. These studies highlight the complex multipathway signaling involved in regulation of liver regeneration after PHX and suggest a strong role for metabolic signals in the initiation and termination of liver regeneration.