Foxm1b transcription factor is essential for development of hepatocellular carcinomas and is negatively regulated by the p19ARF tumor suppressor

Abstract

Hepatocellular carcinoma (HCC) is a leading cause of cancer-related deaths worldwide. Here, we provide evidence that the Forkhead Box (Fox) m1b (Foxm1b or Foxm1) transcription factor is essential for the development of HCC. Conditionally deleted Foxm1b mouse hepatocytes fail to proliferate and are highly resistant to developing HCC in response to a Diethylnitrosamine (DEN)/Phenobarbital (PB) liver tumor-induction protocol. The mechanism of resistance to HCC development is associated with nuclear accumulation of the cell cycle inhibitor p27(Kip1) protein and reduced expression of the Cdk1-activator Cdc25B phosphatase. We showed that the Foxm1b transcription factor is a novel inhibitory target of the p19(ARF) tumor suppressor. Furthermore, we demonstrated that conditional overexpression of Foxm1b protein in osteosarcoma U2OS cells greatly enhances anchorage-independent growth of cell colonies on soft agar. A p19(ARF) 26-44 peptide containing nine D-Arg to enhance cellular uptake of the peptide was sufficient to significantly reduce both Foxm1b transcriptional activity and Foxm1b-induced growth of U2OS cell colonies on soft agar. These results suggest that this (D-Arg)(9)-p19(ARF) 26-44 peptide is a potential therapeutic inhibitor of Foxm1b function during cellular transformation. Our studies demonstrate that the Foxm1b transcription factor is required for proliferative expansion during tumor progression and constitutes a potential new target for therapy of human HCC tumors.

Figure 1.

Alb–Cre Foxm1b^–/– male mouse livers fail to develop adenomas at either 23 or 33 wk of DEN/PB exposure. Both Foxm1b fl/fl (control) and Alb–Cre Foxm1b^–/– (experimental) mice were subjected to either 6, 23, or 33 wk of Diethylnitrosamine (DEN)/Phenobarbital (PB) exposure (Tamano et al. 1994; Sargent et al. 1996; Kalinina et al. 2003) and livers were dissected, paraformaldehyde fixed, paraffin embedded, and microtome sections were prepared. Liver sections were histologically stained with Hematoxylin and Eosin (H&E) and hepatocyte DNA replication was determined by immunofluorescent detection of BrdU incorporation. (A–J) Alb–Cre Foxm1b^–/– male mice are resistant to developing adenomas or hepatocellular carcinomas (HCC) after either 23 or 33 wk of DEN/PB treatment. Arrows indicate margins of adenomas (C,D) or HCC (I,J) in either H&E- or BrdU-stained Foxm1b fl/fl livers after 23 or 33 wk of DEN/PB treatment. (E,F) Photograph of Foxm1b fl/fl mouse livers depicting HCC tumors after either 33 wk (E, male) or 50 wk (F, female) of DEN/PB exposure, whereas Alb–Cre Foxm1b^–/– mice are resistant to liver tumor induction. (K,L) Same microscope field shows α-fetoprotein (AFP) and BrdU-positive immunofluorescent cells in the Foxm1b fl/fl HCC liver tumors, which identified proliferating AFP-positive HCC cells. (M) Graph of mean number of BrdU-positive tumor cells per 1000 cells (±S.D.) in the male Foxm1b fl/fl liver tumors induced by 23 or 33 wk of DEN/PB exposure. (N) Graph of mean number of BrdU-positive cells per 1000 hepatocytes (±S.D.) in nontumor regions of livers from male Foxm1b fl/fl or Alb–Cre Foxm1b^–/– mice either untreated or after 6, 23, or 33 wk of DEN/PB exposure. We calculated the mean number (±S.D.) of BrdU-positive hepatocyte nuclei per 1000 cells or 200× field by counting the number of BrdU-positive hepatocyte nuclei using five different liver sections from three distinct male mice at the indicated times of DEN/PB exposure. Magnification: A–D, 100×; G–J, 50×, K,L, 400×.

Figure 2.

Alb–Cre Foxm1b^–/– mouse hepatocytes exhibit no elevation in apoptosis and increased hypertrophy in response to DEN/PB treatments. (A–C) Alb–Cre Foxm1b^–/– and Foxm1b fl/fl hepatocytes exhibited no differences in apoptosis levels following DEN/PB treatment. Florescent micrograph of TUNEL assay (A,B; 100×) demonstrated similar apoptosis levels in Alb–Cre Foxm1b^–/– and Foxm1b fl/fl control after 23 wk of DEN/PB exposure. (C) Graph of number of apoptotic cells (TUNEL positive) per 1000 hepatocytes (±S.D.) in nontumor regions of livers from male Foxm1b fl/fl or Alb–Cre Foxm1b^–/– mice after either 6, 23, or 33 wk of DEN/PB exposure or untreated. (D–H) Foxm1b-deficient hepatocytes became more polyploid than control hepatocytes after 23 wk of DEN/PB exposure. A centromere-specific mouse FISH probe was used to show that Alb–Cre Foxm1b^–/– hepatocyte nuclei possessed an increase in the number of hybridizing chromosomes compared with control hepatocyte nuclei at 23 wk of DEN/PB treatment. Shown is a high-power magnification of hepatocytes, in which the nuclei were counterstained with DAPI (630×; D,E) or visualized by Laser Confocal microscopy (F,G; bar, 2 μm). (H) Graph of mean number of DAPI-stained hepatocyte nuclei per 200× field (±S.D.) in nontumor regions of livers from male Foxm1b fl/fl or Alb–Cre Foxm1b^–/– mice either untreated or after 6, 23, or 33 wk of DEN/PB exposure. Diminished number of hepatocyte nuclei per field indicates increased hepatocyte hypertrophy. We calculated the mean number (±S.D.) of TUNEL or DAPI-positive hepatocyte nuclei per 1000 cells or 200× field by counting the number of positive hepatocyte nuclei using five different liver sections from three distinct male mice at the indicated times of DEN/PB exposure.

Figure 3.

Hepatocyte nuclear expression of Foxm1b protein after DEN/PB liver tumor-induction protocol. Liver sections from Foxm1b fl/fl and Alb–Cre Foxm1b^–/– mice either untreated or treated with DEN/PB for either 6, 23, or 33 wk were immunohistochemically stained for nuclear expression of Foxm1b protein. Abundant nuclear staining of Foxm1b protein is induced as early as 6 wk after DEN/PB exposure in Foxm1b fl/fl hepatocytes surrounding the portal vein (PV, C), but not in hepatocytes near the central vein (CV). (C,E,G) High levels of nuclear Foxm1b protein persisted in hyperproliferative hepatic adenomas and HCC (margins of tumor indicated by arrows). (B,D,F,H) As expected, nuclear staining of Foxm1b protein was not found in Alb–Cre Foxm1b^–/– hepatocytes at any of the time points following DEN/PB treatment or in untreated liver (A,B). (PV) Portal vein; (CV) central vein. Magnification, 200×.

Figure 4.

Alb–Cre Foxm1b^–/– livers exhibit normal expression of GST-pi and CAR following DEN/PB treatment. Alb–Cre Foxm1b^–/– and Foxm1b fl/fl livers isolated from male mice after 23 wk of DEN/PB exposure were immunohistochemically stained with antibody specific to GST-pi, which is an early marker for altered enzyme foci in response to DEN tumor initiation (Hatayama et al. 1993). Both Alb–Cre Foxm1b^–/– and Foxm1b fl/fl hepatocytes were strongly immunostained for GST-pi after 23 wk of DEN/PB treatment (C–F), but its expression was not detected in untreated control Foxm1b fl/fl mouse liver (A,B). (G) Western blot analysis with liver protein extracts demonstrates that hepatic expression of GST-pi protein was induced as early as 6 wk following DEN/PB treatment, and that its hepatic expression continued following 23 wk of DEN/PB exposure. (H,I) Normal hepatocyte nuclear levels of the CAR nuclear receptor were found in male Alb–Cre Foxm1b^–/– mice following DEN/PB treatment. Magnification: A,C,E, 50×; B,D,F,H,I, 200×.

Figure 5.

Alb–Cre Foxm1b^–/– hepatocytes exhibit persistent increase in nuclear protein levels of Cdk inhibitor p27^Kip1 following DEN/PB exposure. (A,B) Persistent nuclear accumulation of p27^Kip1 protein was found only in Alb–Cre Foxm1b^–/– liver sections at 36 h after partial hepatectomy (PHx), which still exhibited detectable BrdU incorporation (Wang et al. 2002a). (C–J) Liver sections from Alb–Cre Foxm1b^–/– and Foxm1b fl/fl male mice after either untreated or after 6, 23, or 33 wk of DEN/PB exposure were used for immunohistochemical staining to examine for hepatocyte nuclear expression of p27^Kip1 protein. (I,J) Liver sections from Alb–Cre Foxm1b^–/– male and female mice after 50 wk of DEN/PB exposure were immunostained for nuclear levels of p27^Kip1 protein. (C,D) No difference in hepatocyte nuclear staining of p27^Kip1 protein was found in liver sections from untreated Alb–Cre Foxm1b^–/– and Foxm1b fl/fl mice. (C–J) Hepatocyte nuclear staining of p27^Kip1 protein was sustained only in male Alb–Cre Foxm1b^–/– hepatocytes after 6, 23, or 33 wk of DEN/PB exposure (E–J) or female Alb–Cre Foxm1b^–/– hepatocytes after 50 wk of DEN/PB treatment (K,L). (E,G) Margins of hepatic adenoma (Ad) or hepatocellular Carcinoma (HCC) are indicated by arrows. (M,N) Graph of percent p27^Kip1-positive hepatocyte nuclei per 200× field liver section during tumor progression. Number of hepatocyte nuclei per 200× section was determined by DAPI staining in adjacent sections. We calculated the mean ± S.D. percent of p27^Kip1-positive hepatocyte nuclei per 200× fields by counting hepatocyte nuclei positive for p27^Kip1 staining using five different liver sections from three distinct male or female mice that were either untreated or after 6, 23, 33, or 50 wk of DEN/PB exposure. Magnification: A–J, 200×.

Figure 6.

The Cdk inhibitor p27^Kip1 protein associates with Foxm1b through the Cdk–Cyclin complexes and inhibits its transcriptional activity. (A) Western blot analysis of liver protein extracts isolated from either untreated (Un) or DEN/PB-treated mice. Liver protein extracts prepared from two distinct mice following either no treatment or 6, 23, and 33 wk of DEN/PB exposure were used for Western blot analysis with antibodies specific to either p27^Kip1, Cdc25B, or Cdc25C proteins. Expression levels of Cdk2 were used as a loading control. (B) Drawing depicting mechanisms regulating Foxm1b transcriptional activity. Schematically shown is the Foxm1b winged helix DNA-binding domain (WHD), the C-terminal transcriptional activation domain (TAD; Ye et al. 1997), the Foxm1b LXL motif (639–641) that recruits either the Cdk2–Cyclin E/A (S phase), or Cdk1–Cyclin B (G2 phase) complexes. Foxm1b transcriptional activity requires binding of the Cdk–Cyclin complexes, which is necessary for efficient phosphorylation of the Foxm1b Cdk 596 site that recruits the CREB Binding Protein (CBP) histone acetyltransferase (Major et al. 2004). (C) The p27^Kip1 protein associates with GFP–Foxm1b protein through interaction with the Cdk–Cyclin complexes. We performed Co-IP assays with protein extracts prepared from U2OS cells that were transiently transfected with CMV p27^Kip1 and CMV expression vectors containing either wild-type GFP–Foxm1b or GFP–Foxm1b L641A mutant protein that fail to recruit the Cdk–Cyclin complexes. Protein extracts were immunoprecipitated with p27^Kip1 antibody followed by Western blot analysis with GFP monoclonal antibody. We included a control lane containing 1/10 of the extract used in the Co-IP experiment. (D) The p27^Kip1 protein inhibits Foxm1b-transcriptional activity in cotransfection assays. We transiently transfected U2OS cells with the 6× Foxm1b-TATA-luciferase reporter plasmid (Rausa et al. 2003; Major et al. 2004) and the CMV Foxm1b either with or without the CMV p27^Kip1 expression vector. Transfected cells were harvested at 48 h after transfection and processed for dual luciferase assays to determine Foxm1b transcriptional activity. Transfections were performed twice in triplicate and used to calculate the percent of wild-type Foxm1b transcriptional levels (±S.D.).

Figure 7.

p19^ARF tumor suppressor associates with Foxm1b in liver protein extracts and the p19^ARF 26–44 sequences are sufficient to inhibit Foxm1b transcriptional activity. (A) Expression of p19^ARF tumor suppressor is transiently induced after 6 wk of DEN/PB exposure. We performed Western Blot analysis with liver extracts prepared from two distinct mice following either no treatment (Un) or 6, 23, and 33 wk of DEN/PB exposure with a p19^ARF (p19) antibody. Expression levels of Cdk2 were used as a loading control. (B) p19^ARF protein associates with endogenous Foxm1b protein in liver extracts from mice treated for 6 wk with DEN/PB. Co-IP assays were performed with liver protein extracts prepared from Foxm1b fl/fl and Alb–Cre Foxm1b^–/– mice following either 6 or 23 wk of DEN/PB treatment. As a positive control, we also performed Co-IP experiments with protein extracts prepared from mouse embryo fibroblasts (MEFs) that were cultured in vitro for 12 passages to induce endogenous protein expression of the p19 tumor suppressor (Kamijo et al. 1997). The protein extracts were first immunoprecipitated with p19 antibody and then analyzed by Western blot analysis with a mouse Foxm1b antibody. (C) Drawing depicting functional domains of the Foxm1b and p19^ARF tumor suppressor proteins. Schematically shown is the Foxm1b winged helix DNA-binding domain (WHD), the C-terminal transcriptional activation domain (TAD; Ye et al. 1997), and the C-terminal region (688–748) required for p19^ARF (p19) binding. Schematically shown are the p19 nucleolar localization sequence (NrLS) and the p19 Mdm2 and Foxm1b-binding sites (Weber et al. 2000). (D) p19^ARF 26–37 and Foxm1b 688–748 residues are essential for protein association. To identify p19 protein sequences that are essential for association with Foxm1b protein, we performed Co-IP assays with protein extracts prepared from U2OS cells that were transiently transfected with CMV green fluorescent protein (GFP) Foxm1b fusion protein and with p19 expression vectors. These included expression vectors containing either wild-type p19 protein or N-terminal deletion mutants of the p19 protein (Δ1–14, Δ15–25, Δ26–37, Δ26–37 + Δ1–14) that were fused with an HA epitope tag (Weber et al. 2000). The p19 protein was immunoprecipitated from transfected protein extracts with HA antibody, followed by Western blot analysis with a monoclonal antibody specific to the GFP protein to detect the GFP–Foxm1b fusion protein. (E) The p19^ARF sequences 26–44 are sufficient to interact with Foxm1b protein. To identify p19 protein sequences that are sufficient for association with Foxm1b protein, we performed Co-IP assays with protein extracts prepared from U2OS cells that were transiently transfected with CMV GFP–Foxm1b fusion protein and expression vector containing V5 epitope tagged p19^ARF 26–44 or p19^ARF 26–55 sequences (see Materials and Methods). The p19 protein was immunoprecipitated from transfected protein extracts with V5 epitope antibody, followed by Western blot analysis with GFP monoclonal antibody. (F) The p19 protein inhibits Foxm1b transcriptional activity in cotransfection assays. We transiently transfected U2OS cells with the 6× Foxm1b-TATA-luciferase reporter plasmid (Rausa et al. 2003; Major et al. 2004) and the CMV Foxm1b wild type (1–748) and with either CMV wild-type p19 or the indicated N-terminal mutant T7-p19 or V5-p19 26–44 or 26–55 expression vector. We also performed transfection with CMV Foxm1b (1–688) that removed 60 amino acids from the C terminus. Transfected cells were harvested at 48 h after transfection and processed for dual luciferase assays to determine Foxm1b transcriptional activity as described in the legend for Figure 6D.

Figure 8.

The p19^ARF tumor suppressor targets Foxm1b protein to the nucleolus. (A–D) U2OS cells were transfected with HA-p19ARF and GFP–Foxm1b expression vectors, demonstrating that the HA-tagged p19 (A) was able to target nuclear fluorescence of wild-type GFP–Foxm1b fusion protein (D) to the nucleolus (B,C). (E–H) Nucleolar targeting of GFP–Foxm1b wild-type protein was found in cotransfections with CMV expression vectors containing mutant p19^ARF proteins (Δ1–14, Δ15–25, 26–44, or 26–55) that were still able to associate with Foxm1b protein. (I) Nucleolar fluorescence was found with CMV GFP–p19^ARF 26–44 or GFP–p19^ARF 26–55 proteins. (J) Nuclear fluorescence was found with CMV wild-type GFP–Foxm1b and expression vector containing mutant p19^ARFΔ26–37 protein that failed to interact with Foxm1b. (K) Transfection of CMV wild-type p19 expression vector was unable to elicit nucleolar targeting of GFP–Foxm1b 1–688 protein, which failed to bind to p19 protein. (L) Treatment of U2OS cells for 24 h with the TRITC fluorescently tagged (D-Arg)₉-p19^ARF 26–44 peptide demonstrated that this p19^ARF peptide is transduced into the cell and is localized to the nucleolus.

Figure 9.

A membrane-transducing (D-Arg)₉ p19^ARF 26–44 peptide significantly reduces both Foxm1b transcriptional activity and Foxm1b-induced cell colony formation on soft agar. (A) The (D-Arg)₉-p19^ARF 26–44 peptide is an effective inhibitor of Foxm1b transcriptional activity. Treatment of U2OS cells with 12 μM of the (D-Arg)₉-p19^ARF 26–44 peptide that were transfected with CMV-Foxm1b expression vector and the 6× Foxm1b-TATA-luciferase plasmid resulted in significant reduction in Foxm1b transcriptional activation of its target reporter gene. Transfected cells were harvested at 48 h after transfection and processed for dual luciferase assays to determine Foxm1b transcriptional activity as described in the legend for Figure 6D. (B) The CMV-TETO GFP–Foxm1b U2OS clone C3 cell line displayed Doxycycline-inducible expression of the GFP–Foxm1b fusion protein. We used the tetracycline (TET)-regulated T-REx System (Yao et al. 1998) to produce the U2OS Clone 3 cell line as described in Materials and Methods, which induces intermediate levels of the GFP–Foxm1b protein in response to Doxycycline treatment. (C–H) The (D-Arg)₉-p19^ARF 26–44 peptide significantly diminished the ability of induced GFP–Foxm1b to stimulate colony formation of the U2OS clone C3 cells on soft agar. Doxycycline induced Foxm1b–GFP expression stimulates anchorage-independent growth in the U2OS clone C3 cell line (F,G) as assessed by propagation for 2 wk on soft agar (Conzen et al. 2000), whereas the (D-Arg)₉-p19^ARF 26–44 peptide significantly inhibited colony formation of U2OS cells on soft agar (E,H). (I) Quantitation of Foxm1b-induced formation of U2OS cell colonies on soft agar treated or not treated with the (D-Arg)₉-p19^ARF 26–44 peptide. We counted the number of U2OS colonies of the indicated treatments from four to five different 100× fields and determined the mean number of cell colonies (±S.D.). (J) Model for Foxm1b transcriptional inhibition by p19ARF protein during liver tumor initiation, progression, and carcinogenesis in response to DEN/PB liver tumor induction. Foxm1b protein diminishes nuclear expression of the Cdk inhibitor p27^Kip1 protein and regulating expression of Cdk-1 regulator Cdc25B phosphatase.