FAK deletion accelerates liver regeneration after two-thirds partial hepatectomy

Understanding the molecular mechanisms of liver regeneration is essential to improve the survival rate of patients after surgical resection of large amounts of liver tissue. Focal adhesion kinase (FAK) regulates different cellular functions, including cell survival, proliferation and cell migration. The role of FAK in liver regeneration remains unknown. In this study, we found that Fak is activated and induced during liver regeneration after two-thirds partial hepatectomy (PHx). We used mice with liver-specific deletion of Fak and investigated the role of Fak in liver regeneration in 2/3 PHx model (removal of 2/3 of the liver). We found that specific deletion of Fak accelerates liver regeneration. Fak deletion enhances hepatocyte proliferation prior to day 3 post-PHx but attenuates hepatocyte proliferation 3 days after PHx. Moreover, we demonstrated that the deletion of Fak in liver transiently increases EGFR activation by regulating the TNFα/HB-EGF axis during liver regeneration. Furthermore, we found more apoptosis in Fak-deficient mouse livers compared to WT mouse livers after PHx. Conclusion: Our data suggest that Fak is involved in the process of liver regeneration, and inhibition of FAK may be a promising strategy to accelerate liver regeneration in recipients after liver transplantation.

hepatocarcinogenesis by activation of AKT and ERK 14 . Since FAK is important in cell proliferation, it is reasonable to suspect that a role exists for FAK in liver regeneration. However, such a role has as yet to be determined.
In this study, we investigated the role of FAK in liver regeneration in 2/3 PHx model (resection of two-thirds of the mass of liver tissue). Interestingly, we found that specific deletion of Fak in mouse liver accelerates liver regeneration after 2/3 PHx. Consistently, Fak deletion enhances hepatocyte proliferation prior to day 3 after PHx but attenuates hepatocyte proliferation 3 days after PHx. Intriguingly, we found that the deletion of Fak in mouse liver significantly increases EGFR activation but decreases c-MET activation during liver regeneration. Furthermore, we found that Fak deficiency increases HB-EGF (a ligand of EGFR) during liver regeneration and a specific HB-EGF inhibitor abrogates accelerated liver regeneration and enhanced EGFR activation after PHx. Moreover, we discovered that Fak deficiency increases TNFα expression after PHx and a neutralizing TNFα antibody suppresses accelerated liver regeneration, enhanced HB-EGF expression and EGFR activation in Fak-deficient mice. In addition, more apoptosis was found in Fak-deficient mouse livers compared to WT mouse livers after PHx. In general, our data suggest that Fak deletion accelerates the liver regenerative process by regulating the TNFα / HB-EGF/EGFR axis. Inhibition of FAK may be a promising strategy to accelerate liver regeneration in recipients following liver transplantation.

Results
FAK is activated and induced during liver regeneration after 2/3 PHx. The kinase activity of FAK plays a critical role in its functions 15,16 . Phosphorylation of FAK on Tyr397 is required for its activation 17,18 . To study the role of FAK in liver regeneration, we first examined whether FAK is activated during liver regeneration after 2/3 PHx. We found that phosphorylation of FAK on Tyr397 was significantly increased in the livers of mice which underwent PHx 1 day post-surgery compared to the livers of control mice (Fig. 1A). The activation of FAK reached a peak on day 3 post-surgery and decreased to normal by day 5 (Fig. 1A). Interestingly, total FAK expression was also induced on days 2 and 3 after PHx and decreased to a normal level by day 5 (Fig. 1A). Our data indicate that FAK is induced and activated during liver regeneration.

Deletion of Fak in mouse liver accelerates liver regeneration after 2/3 PHx.
To study the role of FAK in liver regeneration, we performed 2/3 PHx on age-and gender-matched Alb-Cre (Hep WT ) and Alb-Cre; Fak flox/flox (Hep ∆Fak ) mice (Fig. 1B). Mice were sacrificed 1, 1.5, 2, 3, 5, 7 and 14 days after PHx and their livers were collected and analyzed. Intriguingly, we found that the liver regenerative process in Hep ∆Fak mice was significantly accelerated compared to Hep WT mice (Fig. 1B). The relative liver weight versus body weight in Hep ∆Fak mice was increased by 20% compared to Hep WT mice on post-surgical day 2 (Fig. 1B). Three days after PHx, the liver mass in Hep ∆Fak mice recovered to almost 100% while liver mass recovered to only 76% in Hep WT mice (Fig. 1B). The liver mass in Hep ∆Fak mice did not continue to increase and remained at 100% 3 days after PHx, while the liver mass of Hep WT mice continued to grow until reaching 100% 14 days after PHx. These data indicate that a deficiency of Fak in mouse liver accelerates liver regeneration after PHx.
Fak deficiency in mouse liver accelerates hepatocyte proliferation during early liver regeneration after 2/3 PHx. Liver mass is replenished by the replication of hepatocytes 19 . Therefore, dramatically increased hepatocyte proliferation takes place during liver regeneration after PHx 2 . We therefore analyzed proliferation in the livers of Hep WT and Hep ∆Fak mice using Ki67 and BrdU staining. The number of Ki67-and BrdU-positive cells was significantly increased in Fak-deficient livers compared to WT livers by day 2 after PHx ( Fig. 2A-D). However, hepatocyte proliferation significantly decreased by day 3 in Hep ∆Fak mice while hepatocyte proliferation in Hep WT mice reached a peak on day 3. Although hepatocyte proliferation in Hep WT mice also declined after day 3, greater hepatocyte proliferation continued to be observed in Hep WT livers compared to Hep ∆Fak livers even by day 7. These results demonstrate that Fak deficiency accelerates hepatocyte proliferation during liver regeneration.
Fak deficiency in mouse liver accelerates liver regeneration by enhancing EGFR activation following 2/3 PHx. EGFR and HGF/c-MET signaling pathways play key roles in hepatocyte proliferation [3][4][5] . We therefore examined whether Fak deficiency in hepatocytes might affect EGFR and HGF/c-MET signaling. Intriguingly, we found that phosphorylation of EGFR, which leads to EGFR activation, is significantly enhanced in Hep ∆Fak mice compared to Hep WT mice by 1.5-3 days after PHx (Fig. 3A). We did not find a significant difference in total EGFR expression between Hep WT and Hep ∆Fak mice (Fig. 3A), suggesting that Fak deficiency enhances EGFR activation during liver regeneration. We also examined the effect of Fak deficiency on c-MET activation during liver regeneration. Interestingly, phosphorylation of c-MET (on Tyr 1234/1235) during liver regeneration, which causes activation of c-MET 20 , was significantly suppressed in Fak-deficient livers (Fig. 3B).
We have found that FAK mediates the activation of AKT and ERK induced by MET in HCC cells 14 . AKT and ERK could be also activated by EGFR activation 21 . Therefore, it would be interesting to see if activation of AKT or ERK is affected by Fak deficiency during liver regeneration. We examined phosphorylation of AKT and ERK in WT and Fak-KO mice after 2/3 PHX. We found that p-AKT and p-ERK were decreased in Hep ∆Fak mice compared to Hep WT mice by 2-5 days after PHx (Fig. S1). These data suggest that Fak deficiency accelerates liver regeneration not by enhancement of AKT or ERK activation. STAT3, another downstream target of EGFR, has been reported to promote hepatocyte proliferation during liver regeneration 8 . We examined phosphorylation of STAT3 in WT and Fak-KO mice after 2/3 PHX. We found that phosphorylation of STAT3 was increased in Hep ∆Fak mice compared to Hep WT mice by 1-2 days post-PHx. These data suggest that Fak deletion might accelerate liver regeneration by enhancing EGFR/STAT3 activation.
To further determine whether Fak deficiency accelerates liver regeneration by enhancing EGFR activation, we examined whether erlotinib, a specific EGFR inhibitor, attenuates liver regeneration in Fak-deficient mice. We treated Fak-deficient mice with 50 mg/kg erlotinib by oral gavage daily for 3 days starting one day prior to PHx. We found that erlotinib significantly attenuated EGFR activation (Fig. 3C), liver regeneration ( Fig. 3D) and hepatocyte proliferation (Fig. 3E,F) in Hep ∆Fak mice. In general, these data indicate that Fak deficiency accelerates liver regeneration by enhancing EGFR activation.
Fak deficiency in mouse liver enhances EGFR activation by increasing HB-EGF expression after 2/3 PHx. Several EGFR ligands, including EGF, TGFα , heparin binding EGF (HB-EGF) and amphiregulin (ARG) have been shown to activate EGFR during liver regeneration 1 . We therefore examined whether Fak deficiency enhances those EGFR ligands during liver regeneration. Interestingly, we found that mRNA levels of HB-EGF, but not EGF, TGFα nor ARG, were rapidly enhanced by Fak deficiency in mouse livers after 2/3 PHx (Fig. 4A). We confirmed that protein levels of HB-EGF were also rapidly enhanced by Fak deficiency in mouse livers after 2/3 PHx (Fig. 4B). HB-EGF transgenic mice have enhanced hepatocyte proliferation during early liver regeneration while liver regeneration was delayed in HB-EGF-knockout mice after 2/3 PHx 22 . Therefore, Fak deficiency in hepatocytes may enhance EGFR activation through increasing HB-EGF expression after 2/3 PHx. To test this hypothesis, we treated Fak-deficient mice with CRM197, a specific inhibitor of HB-EGF 23,24 , daily for 3 days starting one day pre-PHx. We found that CRM197 significantly attenuated EGFR activation (Fig. 4C), liver regeneration (Fig. 4D) and hepatocyte proliferation (Fig. 4E,F) in Hep ∆Fak mice after PHx. These results suggest that Fak deficiency enhances EGFR activation and accelerates liver regeneration by increasing HB-EGF expression.
Fak deficiency in mouse liver increases HB-EGF expression by enhancing tumor necrosis factor (TNFα) expression after 2/3 PHx. HB-EGF can be induced by TNFα in vascular endothelial cells 25 .
Inhibition of TNFα signaling by TNFα -neutralizing antibodies or genetic deletion of TNF receptor 1 reduced hepatocyte proliferation and liver regeneration 26,27 . We therefore hypothesized that Fak deficiency in mouse liver may increase TNFα , thereby enhancing HB-EGF expression. Indeed, we found that mRNA and protein level of TNFα were significantly enhanced by Fak deficiency in mouse livers during liver regeneration after 2/3 PHx (Fig. 5A,B). To examine whether increased TNFα in Fak-deficient livers enhances HB-EGF expression and liver regeneration, we treated Fak-deficient mice with a TNFα -neutralizing antibody daily for 2 days starting on one day prior to PHx. We found that the TNFα -neutralizing antibody significantly suppressed HB-EGF mRNA expression (Fig. 5B), EGFR activation (Fig. 5C), liver regeneration (Fig. 5D) and hepatocyte proliferation (Fig. 5E,F) in Hep ∆Fak mice. These results indicate that Fak deficiency enhances HB-EGF expression, EGFR activation and accelerates liver regeneration by increasing TNFα expression.
Fak deficiency in mouse liver increases death of hepatocytes after 2/3 PHx. TNFα is mainly produced by Kupffer cells in the liver. Hepatocytes undergoing cell death release interleukin-1 alpha (IL-1α ), which can activate Kupffer cells to produce cytokines and growth factors, including TNFα [28][29][30][31] . FAK plays an important role in promoting cell survival 13 . Therefore, we hypothesized that Fak deletion might increase hepatocyte death, thereby activating Kupffer cells to produce more TNFα after PHx. Indeed we did find more apoptosis in Fak-deficient livers compared to WT livers on days 1 and 1.5 post-PHx (Fig. 6A,B). These results suggest that Fak deficiency enhances TNFα expression by increasing hepatocyte death after PHx.   Pyk2, the other member of the FAK family of cytoplasmic tyrosine kinases, shares significant sequence homology and a similar structural organization as FAK 32 . It has been shown that deletion of FAK can lead to increased expression of endogenous Pyk2, which compensates for Fak functions in embryonic fibroblasts, adult endothelial cells and mammary cancer stem cells [33][34][35] . We therefore examined whether Fak deletion results in compensatory expression of Pyk2 in mouse livers after 2/3 PHx. There were no significant changes in the expression of Pyk2 or p-Pyk2 in Hep ∆Fak mouse livers compared to those of Hep WT mice prior to or after PHx (Fig. S2). These data suggest that Fak deficiency in hepatocytes does not lead to a compensatory expression of Pyk2 during liver regeneration after 2/3 PHx.

Discussion
Understanding the molecular mechanism underlying liver regeneration is important for improving the survival rate of patients after surgical resection or reducing the amount of liver tissue required for liver transplantation. In this study, we found that Fak deletion in hepatocytes accelerates liver regeneration after PHx. These data suggest that FAK inhibits liver regeneration and inhibition of FAK may be a promising strategy to accelerate liver regeneration in the liver transplantation setting.
Activation of EGFR and HGF/c-MET signaling is critical for liver regeneration. We found that Fak deletion significantly increases EGFR activation during liver regeneration. We also found greater HB-EGF expression in Fak-deficient livers compared to WT livers after PHx, and inhibition of HB-EGF abrogates the enhanced EGFR and accelerated liver regeneration induced by Fak deletion. HB-EGF has been shown to play an important role in promoting liver regeneration 22 . These data suggest that Fak deletion increases EGFR activation by enhancing HB-EGF expression in the liver. HG-EGF is produced by monocytes and macrophages. HB-EGF mRNA can be induced rapidly (within 1 hour) by TNFα treatment in vascular endothelial cells 25 . In this study, we found that TNFα levels were significantly higher in Fak-deficient livers compared to WT livers after PHx. Inhibition of TNFα by a neutralizing antibody treatment suppressed HB-EGF mRNA expression, EGFR activation and liver regeneration in Fak-deficient livers. These data suggest that Fak deletion might accelerate liver regeneration by increasing TNFα expression. The molecular mechanism by which HB-EGF is induced by TNFα remains unclear. However, the HB-EGF promoter contains multiple putative binding sites for NF-κ B and c-Jun/AP1 (Fig. S3), and these can be activated by TNFα . Therefore, it is possible that TNFα might induce HB-EGF by activating of NF-κ B and c-Jun/AP1. We intend to study this hypothesis in the near future.
How TNFα production is enhanced by Fak deletion during liver regeneration remains unclear. However, our data indicate that there is more hepatocyte death in Fak-KO mice compared to WT mice after PHx. Because dying hepatocytes release IL-1α , which was indeed higher in Fak-KO livers compared to WT livers (data not shown), and IL-1α activates Kupffer cells to produce TNFα 28-31 , it is reasonable to assume that Fak deletion increases hepatocyte death after PHx and enhances TNFα production. FAK has been shown to play an important role in cell survival in anchorage-dependent cells by binding to the death domain of receptor-interacting protein (RIP) 36,37 . Although FAK deletion does not induce apoptosis under homeostatic conditions 14 , many cytokines, including TNFα , have been induced during liver regeneration after PHx, which may result in increased apoptosis in FAK-null cells. It has been shown that increased expression of FAK partially suppresses TNFα -induced apoptosis in intestinal epithelial cells 38 . Therefore, FAK may suppress TNFα -induced hepatocyte death after PHx and the deletion of FAK would enhance TNFα -induced hepatocyte death, resulting in increased TNFα production and activation of EGFR via HB-EGF. We also found TNFα expression was decreased in Fak-deficient livers 2 days after PHx, suggesting that the regulation of TNFα by deletion of Fak is transient. Similar patterns of hepatocyte death in WT and Fak-deficient livers 2 days after PHx further suggests that deletion of Fak enhances TNFα by increasing hepatocyte death.
The increased hepatocyte proliferation and accelerated liver regeneration are suppressed in Fak-deficient livers by 3 days following PHx. We found that c-MET activation during liver regeneration was significantly suppressed in Fak-deficient livers and the HGF/c-MET pathway plays a critical role in promoting liver regeneration. Therefore, we suggest that increased EGFR activation by Fak deletion is sufficient to overcome the decrease in c-MET activation, thereby accelerating liver regeneration at the early time points. However, inhibition of c-MET activation by Fak deletion might inhibit liver regeneration when EGFR activation is diminished 3 days after PHx. Both the positive and negative effects of Fak deletion in liver regeneration reach a balance and liver mass is maintained in Fak-deficient mouse livers after PHx (Fig. 6C). How c-MET activation is inhibited by Fak deletion remains unclear. c-MET directly interacts with the FERM domain of FAK and phosphorylates FAK in MEFs and HEK293 cells 39 . We previously discovered that c-MET also phosphorylates FAK in mouse liver and HCC cells 14 . We intend to undertake future studies to determine whether there is a feedback loop by which FAK also regulates the activation of c-MET in hepatocytes.
In conclusion, our study shows that Fak deletion accelerates liver regeneration after PHx. Inhibition of FAK may offer an effective strategy to accelerate liver regeneration. FAK inhibition also shows promise in inhibiting HCC development 14 . Therefore, inhibition of FAK might kill the proverbial two birds with one single stone: suppressing tumor cell growth and accelerating normal hepatocyte regeneration. A number of FAK inhibitors have been developed and are being studied in Phase I or Phase II clinical trials for multiple solid tumors 40,41 . These inhibitors might be useful to accelerate liver regeneration, especially in patients following liver transplantation.

Methods
Mice and treatments. All animals received humane care according to the "Guide for the Care and Use of Laboratory Animals" (http://oacu.od.nih.gov/ac_cbt/guide3.htm). The procedures for all animal experiments were approved by the Institutional Animal Care and Use Committee of Loyola University Chicago. The generation and breeding of the Alb-Cre and Alb-Cre; Fak flox/flox mice was described previously 14 . Both Alb-Cre and Alb-Cre; Fak flox/flox mice were in C57BL/6 background.
Immunohistochemical (IHC) staining. IHC staining was performed as previously described 14,44 . Cells with positive staining were scored in at least 5 fields at 400× or 200× magnification and reported as mean ± SD. Five mice were used in each group. Statistical analysis. Statistical analysis was carried out using GraphPad Prism V software. Data are presented as means ± standard deviatiosn (SD). Statistical significance was calculated using Student's t test. P < 0.05 was considered to be significant. Means ± SDs are shown in the Figures where applicable.