Aberrant activation of hepatocyte growth factor/MET signaling promotes β-catenin–mediated prostatic tumorigenesis

Co-occurrence of aberrant hepatocyte growth factor (HGF)/MET proto-oncogene receptor tyrosine kinase (MET) and Wnt/β-catenin signaling pathways has been observed in advanced and metastatic prostate cancers. This co-occurrence positively correlates with prostate cancer progression and castration-resistant prostate cancer development. However, the biological consequences of these abnormalities in these disease processes remain largely unknown. Here, we investigated the aberrant activation of HGF/MET and Wnt/β-catenin cascades in prostate tumorigenesis by using a newly generated mouse model in which both murine Met transgene and stabilized β-catenin are conditionally co-expressed in prostatic epithelial cells. These compound mice displayed accelerated prostate tumor formation and invasion compared with their littermates that expressed only stabilized β-catenin. RNA-Seq and quantitative RT-PCR analyses revealed increased expression of genes associated with tumor cell proliferation, progression, and metastasis. Moreover, Wnt signaling pathways were robustly enriched in prostate tumor samples from the compound mice. ChIP-qPCR experiments revealed increased β-catenin recruitment within the regulatory regions of the Myc gene in tumor cells of the compound mice. Interestingly, the occupancy of MET on the Myc promoter also appeared in the compound mouse tumor samples, implicating a novel role of MET in β-catenin–mediated transcription. Results from implanting prostate graft tissues derived from the compound mice and controls into HGF-transgenic mice further uncovered that HGF induces prostatic oncogenic transformation and cell growth. These results indicate a role of HGF/MET in β-catenin–mediated prostate cancer cell growth and progression and implicate a molecular mechanism whereby nuclear MET promotes aberrant Wnt/β-catenin signaling–mediated prostate tumorigenesis.

Prostate cancer is the most common malignancy and the second leading cause of cancer mortality in men in the United States (1). Approximately 90% of patients with metastatic castrate-resistant prostate cancer (CRPC) 3 develop distal secondary bone metastasis, and nearly every patient with bone metastasis eventually succumbs to the disease, resulting in 250,000 deaths worldwide each year (2). Emerging evidence has shown the critical role of the interaction between tumor cells and their surrounding microenvironment in prostatic tumorigenesis. Hepatocyte growth factor (HGF) plays a critical role in the regulation of cell growth, cell motility, morphogenesis, and angiogenesis (3). It has been shown that HGF derived from prostate stroma significantly increases proliferation, motility, and invasion of malignant cells through its receptor, Met (4,5). The Met receptor tyrosine kinase (RTK) is encoded by Met, a protooncogene, and has been shown to play a promotional role in the proliferation and progression of a wide variety of human malignancies, including prostate cancer (4,6). The aberrant expression of HGF and Met often correlates with poor prognosis in cancer patients (7). HGF is abundantly expressed in the tumor microenvironment, leading to Met activation and downstream signaling that promotes several properties of tumor progression and metastasis. Up-regulation of Met expression was observed in a majority of metastatic prostate cancer lesions (6, 8 -11). A nuclear form of MET, nMET, has been identified in human CRPC samples (12). Androgen deprivation can induce nMET expression and promotes cell proliferation and stemlike cell self-renewal in androgen-independent prostate cancer cells (12), implicating a novel role of Met in prostate cancer progression and CRPC development.

Met enhances ␤-catenin-mediated tumor progression in a xenograft model
The expression of stabilized ␤-catenin was initiated by ARR2PB, a modified probasin promoter-driven Cre in the above compound mice (36). It has been shown that ␤-catenin is an AR activator and enhances AR-mediated transcription (18,29). To reduce additional factors and specifically assess the effect of Met and stabilized ␤-catenin in prostate tissues, we implanted prostate tissues that were isolated from 3-week-old H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 and Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice under the kidney capsule of naive SCID male mice (Fig. 4A). The tissue grafts were harvested and analyzed after 12 weeks (Fig. 4A). The weight of tissue grafts derived from H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 was significantly higher than those from Ctnnb1 (Ex3)L/ϩ PB-Cre littermates (Fig. 4B). Histological analyses of prostatic graft tissues of the compound mice showed severe pathologic changes resembling prostatic adenocarcinoma (Fig. 4D). Tumor cells showed cellular abnormalities, including loss of normal polarity and an increase in nuclear to cytoplasmic ratio as well as nuclear pleiomorphism (Fig. 4, D1 and D2). An increase in Ki67-positive cells was also observed in prostatic graft samples of the compound mice compared with those of Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice (Fig. 4B). The graft tissues isolated from Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice showed less severe pathologic changes than those of the com-

Dysregulation of Met and ␤-catenin in prostate tumorigenesis
pound mice (Fig. 4, C-C2), resembling PIN lesions. Graft tissues from both genotype mice showed positive staining for ␤-catenin (Fig. 4, panels C6 and C7 and panels D6 and D7). However, only graft tissues from H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ : PB-Cre4 mice showed positive staining for Met (Fig. 4, D4 and D5). These results provided an additional line of evidence demonstrating the promotional role of MET expression in enhancing stabilized ␤-catenin-initiated prostate tumor development and progression.

Conditional expression of Met enhances Wnt signaling to promote cellular survival, proliferation, and migration
In search of the molecular basis for the collaborative role of transgenic Met and stabilized ␤-catenin expression in prostate tumorigenesis, we performed RNA-Seq to examine the global transcriptome profiles in the tumor tissue of different genotype mice. We microscopically confirmed that the tumor tissues used to prepare RNA samples were composed of more than 80% tumor cells. Analyses of the gene expression profiles of H11 L-Met/ϩ : Ctnnb1 (Ex3)L/ϩ :PB-Cre4 compared with Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice yielded 3240 differentially expressed genes (DEGs), of which 894 genes were up-regulated (Ͼ1 log 2 -fold change) and 2346 genes were down-regulated (ϽϪ1 log 2 -fold change) (Table S2). A heat map (Fig. 5A) depicts potential target genes that are associated with prostate differentiation and growth, tumor progression, proliferation, metastasis, and apoptosis within the context of prostate cancer (39 -60) In support of our previous observations, GSEA analyses with hallmark gene sets revealed significant enrichment of the Wnt signaling pathway based on the DEGs of H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 versus Ctnnb1 (Ex3)L/ϩ :PB-Cre4 prostate tumor tissues ( Fig. 5B and Table S3), suggesting a promotional role of Met in ␤-catenin-mediated signaling pathways. Using RT-qPCR, we further investigated the expression of ␤-catenin downstream target genes using RNA samples isolated

Dysregulation of Met and ␤-catenin in prostate tumorigenesis
with those of Ctnnb1 (Ex3)L/ϩ :PB-Cre4 littermates. Using immunohistochemistry, we further demonstrated increased expression of ␤-catenin, cyclin D1, Lef1, and Myc proteins in prostatic tumor samples of H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice compared with the samples of Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice (Fig. 5, D1-D4 and E1-E4). To directly determine the role of Met in ␤-cateninmediated transcription, we performed ChIP-qPCR analyses using the immunoprecipitated genomic DNA samples isolated from prostatic tumor cells of H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 and Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice to examine the occupancy of stabilized ␤-catenin and Met on the mouse Myc locus (Fig. 5F), a bona fide ␤-catenin downstream target gene (61). We observed a significant increase in recruitment of ␤-catenin within both binding sites in the regulatory region of the Myc in tumor samples isolated from H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 compound mice compared with the ones from Ctnnb1 (Ex3)L/ϩ :PB-Cre4 only mice (Fig.  5G). However, there is no significant recruitment with IgG or on the locus of Untr4, used as a negative control (62). A previous report has shown a nuclear localization of Met in prostate tumor cells (12). In this study, we also observed nuclear staining of Met in prostate tumor cells of H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 compound mice (Figs. 2Hٞ and 3D2Ј). To further explore the potential role of Met in the nuclei of prostatic tumor cells in the above mouse models, we further examined the involvement of Met in ␤-catenin-mediated transcription using the above immunoprecipitated genomic DNA samples. Interestingly, we observed the occupancy of Met on both binding sites within the regulatory region of the Myc in tumor samples of H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ : PB-Cre4 compound mice but not in those of Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice (Fig. 5G). The recruitment of Met in ␤-cateninmediated transcriptional complexes on the promoters of ␤-catenin-regulated downstream target genes, Lef1 and Ccnd1, was also identified in the tumor samples of the above compound mice using ChIP-qPCR approaches (Fig. S2). These data suggest a potential role of Met in ␤-catenin-involved transcription complexes, providing new mechanistic insight into the effect of Met in promoting ␤-catenin-mediated tumor growth and progression.

Discussion
The HGF/Met signaling pathway plays a critical role in prostate tumorigenesis. Up-regulation of Met expression appeared in a majority of advanced and metastatic prostate cancer lesions (6, 8 -11). In addition, aberrant activation of Wnt signaling pathways has also been shown to be one of the most frequent abnormalities in advanced human prostate cancer (20). Recent studies from human prostate cancer samples further suggest that these two alterations co-exist in human prostate cancer, particularly in the late stages of disease (32). Particularly, aberrant co-amplification and activation of Wnt/␤-catenin with abnormal MET or HGF activation were also seen in the above-referenced prostate cancer samples, and an inverse correlation exists between these abnormalities and survival rates (Fig. S1). Given this biological significance and clinical relevance, we directly assessed the collaborative role of aberrant activation Met and ␤-catenin in prostate tumorigenesis using a newly generated mouse model, H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4. As described above, conditional expression of the murine Met gene and stabilized ␤-catenin co-occurred in prostatic epithelial cells of the above mice. This clinically relevant mouse model enables us to recapitulate the aberrant activation of Met and ␤-catenin during prostate cancer development and progression. The H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 compound mice showed accelerated prostate tumor development and progression compared with their Ctnnb1 (Ex3)L/ϩ : PB-Cre4 littermates, demonstrating a promotional role of the HGF/Met signaling axis in Wnt/␤-catenin-mediated prostate tumorigenesis. The H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 compound mouse model will be a biologically relevant and useful tool for further characterizing the molecular mechanisms underlying HGF/Met signaling in prostate cancer development, progression, and metastasis.
Emerging evidence has shown the significance of the HGF/ Met signaling pathway in prostate cancer progression and CRPC development (64). HGF plays a critical role in the regulation of cell growth, cell motility, morphogenesis, and angiogenesis (3). The HGF-Met signaling axis is known to be impor-

Dysregulation of Met and ␤-catenin in prostate tumorigenesis
tant to bone remodeling. An increase in expression of both the Met receptor and the HGF ligand has been observed at sites of prostate cancer bone metastasis, suggesting that this pathway may be active during bone metastasis (8). Up-regulation of Met expression has been observed in most metastatic prostate can-cer lesions (6, 8 -11). A significant challenge within the field of prostate cancer has been the lack of clinically relevant models for examining the biological role of Met in prostate tumor progression and metastasis. Therefore, we recently developed H11 Met/ϩ :PB-Cre4 mice, where expression of the mouse Met

Dysregulation of Met and ␤-catenin in prostate tumorigenesis
gene is specifically activated in prostatic epithelial cells through Cre-LoxP-mediated recombination (15). Although activation of murine Met gene expression with the addition of HGF administration induced prostatic intraepithelial neoplasia development, no prostate tumor formation was revealed in H11 Met/ϩ :PB-Cre4 mice. These data suggest that other oncogenic hits may be required in HGF/Met signaling axis-induced prostate tumor formation and progression. Therefore, in this study, we used stabilized ␤-catenin mice (31), a well-established Wnt signaling tumor model, to directly assess the promotional role of Met in prostate tumor formation and progression. We observed the development of more aggressive and invasive prostate tumors in H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 compound mice compared with their Ctnnb1 (Ex3)L/ϩ :PB-Cre4 littermates. However, we did not observe tumor metastasis in the above compound mice. Despite the evidence indicating that HGF/Met activation is closely associated with bone metastases in human prostate cancer, the failure of metastatic prostate tumor development in the above mouse models indicates that many biological differences may exist between human and murine prostate tissues. The data collected in this study have led us to pursue several more in-depth characterizations of the HGF/Met signaling axis.
The Met is an RTK for HGF (3,4,6). It has been shown that Met activation through binding HGF regulates prostate cell proliferation, motility, and invasion (4,5). Low and inconsistent levels of HGF have been reported in mice (65). To address this caveat, we implanted prostatic tissues isolated from either 3-week-old H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 or Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice under the kidney capsules of HGF transgenic SCID male mice (63). We analyzed graft tissues after 8 weeks of implantation, rather than the 12-week period that we routinely use. We observed robust atypical cell growth and pathologic changes resembling HGPIN lesions in grafts derived from H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ : PB-Cre4 mice. In contrast, only a few local LGPIN lesions were revealed in graft tissues of Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice. The above pathological differences clearly reflect a promotional role of transgenic HGF expression in activating Met in inducing oncogenic transformation in the mouse prostate and promoting tumor cell growth. Establishing a new mouse strain with both transgenic HGF and Met expression may more closely mimic the pathologic conditions of human prostate cancer, and this should be developed and further investigated. Particularly, using this double transgenic mouse strain in the presence of other oncogenic hits may produce more aggressive and metastatic prostate tumor phenotypes in future mouse models.
To further understand the molecular mechanism underlying Met-mediated tumor progression and metastasis, we examined the transcriptional profile in prostate tumor tissues isolated from both H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 or Ctnnb1 (Ex3)L/ϩ : PB-Cre4 mice using RNA-Seq approaches. Increased expression of genes related to tumor development and progression was observed in the samples from H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ : PB-Cre4 mice. An enrichment in Wnt signaling was also identified in DEGs between H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 and Ctnnb1 (Ex3)L/ϩ :PB-Cre4 mice, suggesting a promotional role of transgenic Met expression in enhancing ␤-catenin mediated transcription. An increase in ␤-catenin downstream target genes was further shown in tumor samples from the compound mice. Interestingly, a significant increase of One-cut2, a newly defined master regulator in prostate tumor progression (66), was revealed in tumor samples of the compound mice. A previous study has shown a nuclear form of Met in castrated Pten/Trp53 null prostate tumor cells, which can activate Sox9, ␤-catenin, and Nanog transcription factors (12). Using ChIP-qPCR approaches, we observed the recruitment of Met in ␤-catenin-involved transcription complexes on the promoters and other regulatory regions of ␤-catenin-regulated downstream target genes, including Myc, cyclin D1, and Lef1, in prostatic tumor cells of the H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ :PB-Cre4 compound mice. Our data support the previous observation of nuclear Met and provide additional scientific evidence demonstrating a novel role of Met in facilitating ␤-cateninmediated transcription in prostate cancer cells. Identification of the nuclear role of transgenic Met protein beyond its canonical role on the membrane in tumor cells of H11 L-Met/ϩ : Ctnnb1 (Ex3)L/ϩ :PB-Cre4 is novel and interesting. Further characterization of nuclear Met using H11 L-Met/ϩ :Ctnnb1 (Ex3)L/ϩ : PB-Cre4 mice will provide fresh insight into the role and the regulatory mechanism for HGF/Met-mediated oncogenic signaling in promoting prostate tumor development and progression.

Mouse breeding and genotyping
The founder mice were bred with WT C57Bl6/J, and progenies were genotyped to confirm the presence of the transgene. To generate conditional Met transgenic mice, the LSL-Met transgenic mice (H11 Met/ϩ ) were intercrossed with the PB-Cre4 strain (15), carrying the Cre transgene under the control of a modified probasin promoter (ARR2PB) (36,67). Mice homozygous for floxed ␤-catenin exon 3, Ctnnb1 Ex3(L)/Ex3(L) , were obtained from the Jackson Laboratory (Bar Harbor, ME) (strain 004597). All animals used in this study were on a C57BL/6 background, and all experiments were performed in accordance with animal care guidelines approved by the Institutional Animal Care and Use Committee at Beckman Research Institute and City of Hope.

In vivo prostate regeneration assay
Prostatic tissues were collected from 3-week-old mice and implanted under the renal capsule of SCID mice or hHGFtg Dysregulation of Met and ␤-catenin in prostate tumorigenesis SCID mice (35,70). The SCID or hHGFtg SCID mice were sacrificed after 12 or 8 weeks, respectively, and the grafted tissues were collected and used for histological analyses.

Histological analyses and immunostaining
Mouse tissues were fixed and processed as described in our previous study (72). For histological analysis, 5-m serial sections were processed from Clearify to water through a decreasing ethanol gradient, stained with 5% (w/v) Harris hematoxylin and eosin, and processed back to Clearify through an increasing ethanol gradient.
For immunohistochemical assays, 5-m sections were boiled in 0.01 M citrate buffer (pH 6.0) for 20 min after rehydration from Clearify to water, placed in 0.3% H 2 O 2 /methanol for 15 min, and blocked by 5% goat serum or 5% donkey serum. Tissue slides were then exposed to first antibodies in PBS with 1% goat (or donkey serum) at 4°C overnight. The Tissues were then incubated with biotinylated goat antimouse, goat anti-rabbit (Vector Laboratories, BA-1000 or BA-9200), or donkey anti-goat (ab6987, Abcam, Cambridge, MA) at 1:750 dilution for 1 h at room temperature followed by a 45-min incubation with horseradish peroxidaseconjugated streptavidin (Vector Laboratories, SA-5004). Immunostainings were visualized using a DAB kit (Vector Laboratories, SK-4100). Slides were counterstained with hematoxylin, and coverslips were mounted with Permount Mounting Medium (SP15-500, Fisher).

RNA isolation, RNA-Seq, and RT-qPCR
RNA samples were isolated from age-matched mice of different genotypes. The prostate tissues were homogenized in RNA-Bee (TEL-TEST, Inc., Friendswood, TX), and total RNA was isolated as recommended by the manufacturer. The purified RNA libraries were then sequenced using the Illumina HiSeq 2000 at the City of Hope Integrative Genomics Core. Pathway analysis of hallmark gene sets was performed using preranked gene set enrichment analysis (GSEA 4.0.1).

ELISA
Mouse sera were collected from different genotype mice and separated by centrifugation from total blood following cardiac puncture as described previously (38). The serum concentration of HGF was measured by the Quantikine ELISA human HGF immunoassay (DHG00B, R&D Systems). Measurements were performed in in accordance with the manufacturer's instructions.

Statistical analyses
Data are shown as the mean Ϯ S.D. Differences between groups were examined by two-tailed Student's t test or twoway analysis of variance for comparisons among multiple groups. For all analyses, p Ͻ 0.05 was considered statistically significant.