SCRIB expression is deregulated in human prostate cancer, and its deficiency in mice promotes prostate neoplasia

Loss of cellular polarity is a hallmark of epithelial cancers, raising the possibility that regulators of polarity have a role in suppressing tumorigenesis. The Scribble complex is one of at least three interacting protein complexes that have a critical role in establishing and maintaining epithelial polarity. In human colorectal, breast, and endometrial cancers, expression of the Scribble complex member SCRIB is often mislocalized and deregulated. Here, we report that Scrib is indispensable for prostate homeostasis in mice. Scrib heterozygosity initiated prostate hyperplasia, while targeted biallelic Scrib loss predisposed mice to prostate intraepithelial neoplasia. Mechanistically, Scrib was shown to negatively regulate the MAPK cascade to suppress tumorigenesis. Further analysis revealed that prostate-specific loss of Scrib in mice combined with expression of an oncogenic Kras mutation promoted the progression of prostate cancer that recapitulated the human disease. The clinical significance of the work in mice was highlighted by our observation that SCRIB deregulation strongly correlated with poor survival in human prostate cancer. These data suggest that the polarity network could provide a new avenue for therapeutic intervention.

Distinct functional and spatial domains are created within epithelial cells owing to the unequal distribution of cellular constituents, referred to as cellular polarity (1). The spatial asymmetry of polarity regulators permits epithelial cells to respond to extracellular cues and maintains the tissue architecture and cellular polarity required for normal physiological function and tissue homeostasis (2, 3). Loss of cell and tissue polarity is a frequent characteristic of epithelial cancers, suggesting a crucial role for polarity mediators in suppressing tumorigenesis (2, 4).

The mammalian polarity program is established by at least three interacting protein complexes (Scrib, Crumbs, and Par) and mediates a plethora of polarization processes, including apical-basal polarity, migration, asymmetric cell division, and planar cell polarity (2, 3). The apical protein modules termed the Crumbs (Pals, PatJ, and Crumbs) and Par (Par3, Par6, and aPKC) complexes act in a mutually antagonistic relationship with the basolateral Scribble module (5). The Scribble module comprises Scrib, Discs large 1–4 (Dlg1–4), and lethal giant larvae 1/2 (Lgl1/2) (5). Together, these complexes mediate a complex series of processes to establish and maintain polarity, including the formation of cell-cell contacts (i.e., adherens and tight junction assembly) (3).

Scrib was originally identified as a neoplastic tumor suppressor in Drosophila (6), whereas human SCRIB was discovered as a target for ubiquitin-mediated degradation by the human papillomavirus (HPV) E6 proteins and E6AP protein ligase (7), which also target human DLG1 (8). Cervical carcinomas are commonly associated with HPV and have been shown to display reduced SCRIB and DLG1 protein expression (9, 10).

Deregulation of Scribble complex members (SCRIB, DLG1, and LLGL2) has been correlated to a number of human epithelial cancers (11–14), although somatic mutations remain to be identified. Human colorectal, breast, and endometrial cancers display mislocalized and deregulated SCRIB expression (14–18), suggesting that SCRIB may be crucial for the homeostatic maintenance of several epithelial tissues. Together, these data highlight the critical function of the Scribble module to maintain epithelial tissue architecture and cellular polarity to prevent tumor onset and progression.

SCRIB has been implicated in the regulation of many cellular processes, including proliferation, differentiation, apoptosis, stem cell maintenance, migration, vesicle trafficking, and apical-basal polarity (5). These events appear to be mediated through several emerging oncogenic signaling cascades, such as the Ras/MAPK, β-catenin/Wnt, planar cell polarity, and Notch pathways, indicating direct mechanisms whereby SCRIB could suppress tumorigenesis (5). SCRIB encodes a large, 220-kDa cytoplasmic scaffolding protein that comprises a large leucine-rich repeat (LRR) region and 4 PDZ domains that regulate protein-protein interactions (5, 7), including SCRIB-LLGL2 (19). Other reported SCRIB binding partners include the tight junction protein ZO-2 (20), the planar cell polarity protein VANGL2 (19), the tumor suppressor adenomatous polyposis coli (APC) (21), and the MAPK signaling molecule ERK (22). In fact, previous in vitro work undertaken in our laboratory and others has indicated that SCRIB may negatively regulate the RAS/MAPK cascade to maintain polarity and tissue homeostasis (22, 23). Furthermore, studies in Drosophila and mammalian cell lines show that SCRIB loss and RAS activation cooperate to promote invasion (23, 24).

Despite these striking human cancer associations and biochemical and functional findings, the tumor-suppressive role of Scrib remains to be determined in vivo. To directly assess this possibility, we generated and characterized global and tissue-specific Scrib–deficient mice. We show that prostate homeostasis is highly sensitive to Scrib deregulation and demonstrate that Scrib heterozygosity is sufficient to cause focal prostate hyperplasia, whereas biallelic loss within prostate epithelium promoted disease progression to low-grade prostate intraepithelial neoplasia (LG-PIN). We observed loss of cellular polarity and elevated proliferation concomitant with elevated Ras/MAPK signaling in Scrib-deficient prostate lesions, suggesting that Scrib negatively regulates the Ras/MAPK cascade to suppress prostate tumorigenesis. In support of this notion, chemical inhibition of the Ras/MAPK cascade demonstrated that this pathway is required for hyperplastic growth in the context of Scrib deficiency. Additionally, we reveal that Scrib loss and oncogenic K-ras cooperate to promote prostate cancer progression using the PBCre⁺ transgenic line to drive Scrib depletion and expression of oncogenic K-ras (LSL-K-ras^G12D) specifically within the prostate. This compound transgenic model recapitulates key pathophysiological aspects of the human disease and exemplifies the tumor-suppressive function of Scrib during prostate cancer initiation and progression. Importantly, we determine that SCRIB is deregulated in human prostate cancer and that aberrant SCRIB localization strongly correlates with poor survival. These studies establish Scrib as a critical barrier to prostate tumorigenesis and present polarity proteins as potential prognostic markers in the clinic.

Generation of Scrib conditional knockout mouse. To test whether deregulation of Scrib expression predisposes to epithelial cancer in vivo, we engineered mice harboring a conditional knockout allele of the Scrib gene (Figure 1A and Supplemental Figure 1, A and B; supplemental material available online with this article; doi: 10.1172/JCI58509DS1). Cre recombinase expression results in the excision of exons 4–13, comprising the majority of the LRR domain, which is required for Scrib function (16, 25, 26). This deletion introduces a nonsense mutation in the nascent Scrib mRNA and a frameshift in the translation reading frame, predicted to produce a small protein fragment (100 aa) with no known functional domains.

Figure 1

Generation and analysis of Scrib-deficient mice. (A) Schematic representation of the WT Scrib allele, targeting construct, and mutated Scrib allele. Excision of the Frt-flanked PGK-Neo cassette in the LoxP-flanked targeting vector was accomplished by crossing Scrib^fl-Neo mice to Actin-FLPe mice, producing the Scrib floxed (Scrib^fl) allele. The insertion of EcoRV and KpnI restriction sites permitted screening for recombinant ES cell clones (see Supplemental Figure 1, A and B). Scrib-KO mice were derived by crossing Scrib^+/fl mice with a germline Cre-deleter mouse. Exons are indicated as boxes and LoxP/FLPe sites as triangles. (B) Photographs of WT, Scrib^+/–, and KO E14.5 embryos. Scale bar: 10 mm. (C) Western blot analysis of embryonic brain lysates (E16.5) confirmed Scrib loss in Scrib^+/– and KO compared with WT embryos.

To examine the effect of ubiquitous Scrib deficiency, we crossed mice harboring a LoxP-flanked Scrib allele (Scrib^+/fl) to a germline Cre-deleter transgenic line (27), resulting in viable Scrib^+/– progeny. To engineer global depletion of Scrib (KO), we intercrossed Scrib^+/– mice. Genotyping of embryos confirmed normal Mendelian ratios, yet Scrib-KO mice died perinatally owing to a severe neural tube closure defect (craniorachischisis) (Figure 1B and Supplemental Figure 1C). All KO embryos displayed a curly tail, eyes open at birth, and gonad morphological abnormalities of both sexes, and 54% (14 of 26) displayed defective abdominal wall closure (gastroschisis) (Supplemental Figure 1, D–F). These characteristics mirror the Circletail (Crc) and rumz Scrib point mutation models, consistent with specific ablation of Scrib gene function (28, 29). Western blot analysis confirmed that Scrib was efficiently depleted in vivo, indicating the Scrib-KO allele is a true null allele (Figure 1C).

Scrib^+/– mice display prostate hyperplasia. Loss of polarity has been implicated as a hallmark of epithelial cancers (4); however, it remains to be determined whether deregulation of polarity regulators is a tumor-initiating event or a consequence of transformation. To address whether reduced Scrib expression is sufficient to drive epithelial tumorigenesis, we initially examined epithelial tissues from a cohort of Scrib^+/– mice at 100 days of age (n = 10). Remarkably, 7 of 10 Scrib^+/– mice were predisposed to prostate hyperplasia (Figure 2), while histologically normal tissues included the intestine, liver, lung, stomach, pancreas, kidney, and bladder (data not shown). PCR analysis of genomic DNA confirmed recombination of the LoxP-flanked Scrib allele in Scrib^+/– prostate tissue (Figure 2A) and quantitative RT-PCR (qRT-PCR) showed a significant, 55% reduction in Scrib mRNA isolated from Scrib^+/– prostate tissue compared with WT prostate (Figure 2B). Immunofluorescence (IF) to detect Scrib revealed that WT epithelium displays uniform Scrib expression along the basolateral membrane that is prominent at tight junctions (Figure 2C). Although Scrib^+/– lesions appeared to display a decrease in Scrib expression, statistical analysis of Scrib intensity revealed this was not significant (Figure 2C and Supplemental Figure 2A). Importantly, no difference in Scrib intensity was established between Scrib^+/– hyperplastic and non-hyperplastic prostate epithelium, indicating that loss of heterozygosity (LOH) had not taken place (Supplemental Figure 2A), and to our knowledge, Scrib LOH has not been reported in the literature.

Figure 2

Scrib heterozygosity predisposes to prostate hyperplasia. (A) PCR analysis of genomic DNA to detect WT (290 bp) and KO (550 bp) Scrib alleles in Scrib^+/– and WT prostate. (B) qRT-PCR for Scrib mRNA confirmed a significant, 55% reduction in Scrib^+/– compared with WT prostates (P = 0.0065, unpaired t test). Error bars indicate SD. (C) Scrib IF staining of WT and Scrib^+/– hyperplastic prostate. (D) Representative H&E images of normal WT and hyperplastic Scrib^+/– prostate (400 days). (E) Phenotype incidence in WT and Scrib^+/– mice at 100, 200, and 400 days (n ≥ 10). Scale bars: 50 μm (larger panels) and 10 μm (insets).

To determine whether Scrib heterozygosity is sufficient to drive neoplastic progression, male cohorts of WT and Scrib^+/– littermates were generated and aged to 200 and 400 days (n ≥ 10). Histological analysis revealed normal branched and single-layered prostate epithelium in WT mice at all time points (Figure 2, D and E). In contrast, all Scrib^+/– mice displayed multifocal prostate hyperplasia (predominantly in the anterior lobe) at 200 and 400 days (100% incidence). Scrib^+/– prostate lesions exhibited lumen overcrowding and a marked increase in mitotic figures (Figure 2D), indicating that Scrib heterozygosity predisposes to prostate hyperplasia. Further investigation of smaller cohorts revealed that multifocal prostate hyperplasia is present in Scrib^+/– mice as early as 50 days (50% incidence), while WT mice displayed normal prostate epithelium (n = 6). Aging to 540 days did not cause neoplastic progression. Multifocal prostate hyperplasia was observed in Scrib^+/– mice at 540 days, with 100% penetrance (n = 8). Indeed, an average of 17%–42% prostatic ducts displayed hyperplastic foci (50–540 days) that contained a similar packing density of cells, illustrating the general multifocal nature of this phenotype (Supplemental Figure 2, B and C). Noticeably, one 540-day-old Scrib^+/– mouse developed a pancreatic preinvasive ductal lesion, and 4 of 8 (50%) were predisposed to lung adenocarcinomas (I. Elsum, unpublished observations). None of these lesions were observed in WT mice at 540 days (n = 5).

Characterization of Scrib^+/– prostate hyperplasia. Assessment of the proliferation marker PCNA by IHC revealed a significant, 4.8-fold increase in the number of PCNA-positive cells in Scrib^+/– prostate hyperplasia compared with WT prostate at 400 days (Figure 3A). However, no change in apoptosis was detected at this time point, as determined by active caspase-3 IHC (Supplemental Figure 2D). To ascertain which population of cells were proliferating, we analyzed basal (cytokeratin-5), and luminal (cytokeratin-8) lineage markers by IHC (Supplemental Figure 2E). Scrib^+/– prostate hyperplastic cells stained positively for cytokeratin-8, whereas the distribution of cytokeratin-5–positive basal cells resembled that in WT prostate tissue. Together, these data indicate that Scrib heterozygosity is an initiating event in murine prostate hyperplastic growth and that Scrib deficiency induces proliferation, reflecting expansion of the luminal cell lineage.

Figure 3

Elevated Ras/MAPK signaling is required for Scrib-deficient prostate hyperplasia. (A) PCNA IHC shows a significant increase in PCNA-positive cells in Scrib^+/– lesions (2.26% ± 0.62%, mean ± SD) compared with WT (0.47% ± 0.09%, P < 0.0001, unpaired t test, 400 days). (B) IHC in Scrib^+/– hyperplastic foci revealed aberrant E-cadherin and p-ERM compared with WT tissue. Arrows indicate aberrant expression. Right column: IF to detect Par3 (red) and pan-Dlg (green) shows that these polarity regulators are mislocalized in Scrib^+/– prostate hyperplasia compared with WT prostate (400 days). N, normal Scrib^+/– tissue; Hyp, hyperplastic Scrib^+/– tissue. Arrowheads indicate Par3 accumulation at tight junctions. (C) IHC revealed a significant elevation in p-ERK–positive nuclei in Scrib^+/– lesions compared with WT (unpaired t test, 400 days). (D) IHC revealed a significant elevation in p-ELK1–positive nuclei in Scrib^+/– lesions compared with WT (unpaired t test, 400 days). (E) Representative H&E images of Scrib^+/– prostates administered vehicle or PD0325901 (20 mg/kg, 5 days on, 2 days off for 3 weeks at 230–260 days of age), and p-ERK staining showing efficient MEK inhibition. (F) Scrib^+/– prostate weight is significantly decreased upon MEK inhibition (0.30 ± 0.04 g, mean ± SD) compared with administration of vehicle (0.45 g ± 0.08, P = 0.0003, Mann-Whitney U non-parametric test), similar to the level in WT prostate (P = 0.4606, Mann-Whitney U non-parametric t test). Scale bar: 50 μm (larger panels) and 10 μm (insets). Error bars indicate SD.

To determine whether epithelial polarity is lost in Scrib^+/– prostate lesions, we assessed the localization of markers for the apical (p-ERM) and lateral (E-cadherin) cell surfaces (Figure 3B). In WT prostate epithelial cells, E-cadherin was detected on the lateral surface and p-ERM was restricted to the apical surface. E-cadherin participates in adherens junctions, whereas p-ERM has been shown to play a role in mediating cell shape, adhesion, and motility (30). In contrast, Scrib^+/– hyperplastic lesions displayed mislocalized E-cadherin, yet no significant difference in E-cadherin intensity was observed between WT and Scrib^+/– prostate epithelium (Supplemental Figure 2F). The enrichment and mis-distribution of p-ERM along the plasma membrane circumference suggest expansion and reorganization of the apical domain (30), consistent with a loss of apical-basal polarity (31). Loss of polarity within Scrib^+/– hyperplastic lesions was further exemplified by IF staining for core apical (Par3) and basolateral (Dlg) components of the polarity program (Figure 3B). In WT prostate epithelium, Par3 localized apically, with prominent accumulation at tight junctions, while pan-Dlg localized to the basolateral membrane. Scrib^+/– hyperplastic lesions displayed the mis-distribution of Par3 throughout the cytoplasm and Dlg along the circumference of the membrane. These data support previous work showing that loss of Scribble complex members is sufficient to mislocalize key polarity proteins and impair tissue and cellular polarity (6, 32–34).

Scrib^+/– prostate hyperplasia is dependent on Ras/MAPK signaling. We and others have previously shown that SCRIB loss stimulates Ras/MAPK signaling in vitro, suggesting that SCRIB negatively regulates the Ras/MAPK cascade to suppress tumorigenesis (22, 23). We hypothesized that Scrib loss deregulates the Ras/MAPK pathway to facilitate prostate hyperplastic growth in mice. To test this hypothesis, we analyzed phosphorylation of the MAPK signaling molecule ERK. As determined by IHC, Scrib^+/– prostate hyperplasia specifically displayed a significant, 3.7-fold increase in nuclear p-ERK expression compared with WT epithelium, indicating that Scrib deficiency elevates Ras/MAPK signaling (Figure 3C and Supplemental Figure 2G). Accordingly, expression of p-ELK1, a downstream transcription factor target of p-ERK (35), was also significantly increased in Scrib^+/– hyperplastic lesions compared with WT tissue (Figure 3D). Activation of the MAPK cascade was further confirmed by elevated cytoplasmic p-MEK expression in Scrib^+/– hyperplastic foci (Supplemental Figure 2H). These data are consistent with the notion that Scrib plays a tumor-suppressive role within the murine prostate by regulating the Ras/MAPK pathway.

To directly test this hypothesis, we treated Scrib^+/– mice with the commercially available MEK inhibitor PD0325901 (20 mg/kg, 5 days on and 2 days off for 3 weeks) at 230–260 days of age (n = 8), when multifocal prostate hyperplasia is usually observed. Histological analysis revealed a marked decrease in hyperplastic foci in Scrib^+/– mice administered PD0325901 compared with the vehicle (Figure 3E). Effective inhibition of p-ERK expression in response to the MEK inhibitor was confirmed by p-ERK IHC (Figure 3E and Supplemental Figure 3A). PD0325901 administration significantly reduced Scrib^+/– prostate weight compared with the vehicle, and the level was similar to that in WT littermate controls (Figure 3F). Scrib^+/– mice receiving PD0325901 also showed a significant, 2.4-fold reduction in PCNA expression compared with those receiving vehicle, indicating that MEK inhibition decreased proliferation in this setting (Supplemental Figure 3B). Notably, no phenotypic difference was detected by H&E analysis in WT littermates receiving the MEK inhibitor compared with the vehicle, consistent with the lack of a significant difference in prostate weight (P = 0.3429, Mann-Whitney U test, n = 4) (Figure 3F and Supplemental Figure 3C). IHC to detect p-ERK confirmed efficient MEK inhibition in WT epithelium (Supplemental Figure 3D).

Taken together, these data indicate that MEK inhibition is sufficient to rescue prostate hyperplasia in the context of Scrib deficiency and that intact Ras/MAPK signaling is required for prostate hyperplastic growth induced by Scrib depletion. Nevertheless, we cannot rule out the possibility that additional genetic/epigenetic changes may contribute to the Scrib-deficient phenotype in this setting.

Biallelic Scrib loss promotes neoplastic tumor progression. We next determined whether complete Scrib inactivation in the prostate could facilitate prostate cancer growth. Given that Scrib-KO mice are neonatal lethal, we employed Cre-LoxP technology to deplete Scrib expression specifically within the prostate epithelium. Scrib floxed mice were crossed to the PBCre transgenic line (36), and male cohorts of PBCre⁺, PBCre⁺;Scrib^+/fl, and PBCre⁺;Scrib^fl/fl mice were established and aged.

We confirmed PBCre-mediated recombination in the prostate by crossing PBCre⁺ mice to the Rosa26 reporter strain (37) and performing LacZ staining (Supplemental Figure 4A). Scrib loss was verified by means of qRT-PCR (Figure 4A). Scrib mRNA was significantly reduced in PBCre⁺;Scrib^+/fl and PBCre⁺;Scrib^fl/fl prostates compared with those of PBCre⁺ controls. Consistent with mRNA expression data, negligible levels of Scrib were detected by IF in PBCre⁺;Scrib^fl/fl prostate epithelium (Figure 4B).

Figure 4

Biallelic Scrib loss causes prostate neoplasia. (A) Scrib mRNA is decreased in PBCre⁺;Scrib^+/fl (45.2%) and PBCre⁺;Scrib^fl/fl (14.5%) compared with PBCre⁺ prostates at 400 days. (B) Scrib IF shows reduced Scrib expression in PBCre⁺;Scrib^fl/fl prostate epithelium (400 days). (C) Representative H&E images of PBCre⁺, PBCre⁺;Scrib^+/fl, and PBCre⁺;Scrib^fl/fl prostates (400 days). (D) Prostate phenotype incidence at 100, 200, and 400 days (n ≥ 10). (E) PCNA IHC shows a significant increase in the number of PCNA-positive cells in PBCre⁺;Scrib^fl/fl mice (4.0% ± 0.24%) compared with PBCre⁺;Scrib^+/fl (2.1% ± 0.82%) and PBCre⁺ mice (0.6% ± 0.38%) at 400 days. (F) p-ERK staining revealed a significant increase in MAPK signaling in PBCre⁺;Scrib^+/fl (7.1% ± 1.64%) and PBCre⁺;Scrib^fl/fl mice (13.1% ± 4.18%) compared with PBCre⁺ controls (1.8% ± 0.58%) at 400 days. (G) p-ELK1 staining shows a significant increase in p-ELK1 expression in PBCre⁺;Scrib^+/fl (1.8% ± 0.25%) and PBCre⁺;Scrib^fl/fl (4.2% ± 1.67%) compared with WT tissue (0.4% ± 0.13%) at 400 days. Arrows indicate positive nuclei. Scale bars: 50 μm (larger panels) and 10 μm (insets). Data are mean ± SD; n = 3; P values represent unpaired t test. Error bars indicate SD.

Histological analysis revealed that PBCre⁺;Scrib^+/fl mice phenocopied Scrib^+/– animals, displaying multifocal prostate hyperplasia at 100, 200, and 400 days (88%, 75%, and 90% incidence respectively; n ≥ 10), indicating that the observed phenotype is intrinsic to the epithelial compartment (Figure 4, C and D). Analysis of PBCre⁺;Scrib^fl/fl prostate tissue revealed that biallelic Scrib loss facilitated neoplastic progression. At 200 and 400 days, PBCre⁺;Scrib^fl/fl mice displayed LG-PIN (14% and 20% incidence, respectively; n ≥ 10) with solid and cribriform intraluminal proliferation of evidently atypical epithelial cells (Figure 4, C and D). Notably, PIN lesions were commonly associated with regions of hyperplasia.

IHC to detect the proliferation marker PCNA demonstrated that PBCre⁺;Scrib^fl/fl prostate lesions display a significant, 1.9- and 6.9-fold increase in PCNA expression compared with PBCre⁺;Scrib^+/fl and PBCre⁺ mice, respectively (Figure 4E). In addition, PBCre⁺;Scrib^fl/fllesions showed a small increase in the number of cells expressing the apoptotic marker active caspase-3; however, this elevation was not significantly different compared with WT and PBCre⁺;Scrib^+/fl mice (Supplemental Figure 4B).

Consistent with the Scrib^+/– model, we observed activation of the Ras/MAPK cascade in prostate-specific Scrib–deficient lesions (Figure 4F). IHC for p-ERK showed a significant, 3.9- and 7.2-fold increase in the number of positive p-ERK nuclei in PBCre⁺;Scrib^+/fl and PBCre⁺;Scrib^fl/fl mice, respectively, compared with PBCre⁺ controls. These data correlated with a significant increase in the downstream target of p-ERK, p-ELK1 (Figure 4G). These results further illustrate that Scrib loss facilitates prostate tumorigenesis through deregulated Ras/MAPK signaling in prostate epithelial cells.

K-ras activation and Scrib loss cooperate to facilitate prostate tumor progression. Loss of scrib and oncogenic ras cooperate to drive invasive cancer in Drosophila (24, 38). In addition, despite the low frequency of RAS mutations in human prostate cancer (39), RAS/RAF signaling is one of the three most common pathways to be altered in both primary and metastatic human prostate cancer tumors (39), and K-RAS deregulation and elevated MAPK signaling are common features (39, 40). To investigate the potential cooperation between Scrib deficiency and oncogenic Ras/MAPK signaling in prostate tumorigenesis, we generated and aged cohorts of male mice in which Scrib loss and hyperactivated K-ras (LSL-K-ras^G12D) were specifically induced in the prostate (n ≥ 10). Several PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ mice (3 of 10) displayed signs of illness and were sacrificed prior to the 400 days end point (Figure 5A). Compared with the PBCre⁺;LSL-K-ras^G12D/+ cohort, which survived to 400 days, PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ mice displayed a small yet significant reduction in average survival (372 days), suggesting that Scrib depletion and K-ras activation may cooperate to accelerate disease progression. Although 2 mice in the PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+ cohort became ill and were sacrificed before 400 days, no statistical difference in average survival was determined compared with single mutant or control cohorts.

Figure 5

Scrib loss and K-ras hyperactivation cooperate to facilitate prostate tumor progression. (A) Kaplan-Meier survival plot shows a significant reduction in PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ average survival (372 days) compared with PBCre⁺;LSL-K-ras^G12D/+ mice (χ² = 4.20, P = 0.0449, log-rank test, n ≥ 10). (B) Representative H&E images of PBCre⁺;LSL-K-ras^G12D/+, PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+, and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ prostates (400 days). (C) Prostate phenotype incidence (400 days, n ≥ 10). (D) IHC to detect p-ERK reveals that the number of p-ERK–positive nuclei in PBCre⁺;LSL-K-ras^G12D/+ (21% ± 11.24%), PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+ (20% ± 11.41%), and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ (17% ± 6.35%) prostate tumors was significantly elevated compared with control prostate (*P ≤ 0.0496), yet no statistical difference was determined between the double mutants and K-ras activation alone (P > 0.6168) at 400 days. (E) IHC to detect p-ELK1 revealed a significant increase in p-ELK1 expression in PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+ (36.6% ± 6.54%), and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ (50.5% ± 0.43%) prostate lesions compared with PBCre⁺;LSL-K-ras^G12D/+ (17.7% ± 0.55%, P ≤ 0.0448) and PBCre⁺ (0.4% ±0.07, P ≤ 0.0052) prostate epithelium at 400 days. Scale bars: 50 μm (larger panels) and 10 μm (insets). In D and E, data are mean ± SD; n = 3; P values represent unpaired t test. Error bars indicate SD.

PCR analysis of DNA isolated from the prostate revealed that recombination of the LoxP-flanked LSL-K-ras^G12D and Scrib alleles occurred in single and double transgenics, respectively, and not in PBCre⁺ and Cre-negative controls (Supplemental Figure 4C). qRT-PCR analysis confirmed a reduction in Scrib mRNA in PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+ and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ mice compared with controls (Supplemental Figure 4D).

Histological analysis of PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+ and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ mice established that compound mutants were susceptible to accelerated tumor progression compared with mice with Scrib loss or K-ras activation alone (Figure 5, B and C). PBCre⁺;LSL-K-ras^G12D/+ prostate glands displayed hyperplasia, LG-PIN, and high-grade PIN (HG-PIN) lesions (9%, 73%, and 18% incidence, respectively), consistent with a similar model (41). PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+ and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ mice displayed HG-PIN (20% and 33% incidence respectively) and well-differentiated adenocarcinoma (20% and 13% incidence, respectively), and 3 of 15 (20%) PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ mice developed poorly differentiated invasive carcinoma (Figure 5, B and C). HG-PIN lesions displayed solid and occasional cribriform patterns with nuclear enlargement/elongation. Well-differentiated adenocarcinomas displayed well-maintained glandular differentiation, while poorly differentiated carcinoma showed more solid nests of tumor cells. All the aforementioned lesions displayed nuclear atypia, mitotic figures, and extensive overcrowding of the lumen as multicellular disorganized layers formed. In addition, all K-ras–induced prostate tumors (PBCre⁺;LSL-K-ras^G12D/+, PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+, and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+) were associated with focal intestinal metaplasia, mirroring previous work (42).

To address the molecular mechanism underlying the cooperation between Scrib deficiency and K-ras activation to promote prostate cancer progression, we performed PCNA IHC (Supplemental Figure 4E). Compared with PBCre⁺ controls, PBCre⁺;LSL-K-ras^G12D/+, PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+, and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ mutants all showed a significant increase in PCNA-positive cells, although Scrib loss did not result in a significant increase compared with K-ras activation alone. In addition, a significant elevation in the expression of the apoptosis marker active caspase-3 in PBCre⁺;LSL-K-ras^G12D/+, PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+, and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ prostate tumors compared with control mice was determined by means of IHC; however, no statistical difference was observed between double mutants and K-ras activation alone (Supplemental Figure 4F). IHC confirmed that all mutants displayed aberrant E-cadherin and p-ERM expression, indicating that polarity was lost (Supplemental Figure 4G). Taken together, these data indicate that aberrant proliferation, apoptosis, and loss of polarity in double mutants do not trigger accelerated tumor progression.

To confirm activation of Ras/MAPK signaling in K-ras mutants and to further understand the mechanism whereby Scrib loss and K-ras activation cooperate to accelerate prostate cancer progression, we analyzed the activity of the Ras/MAPK pathway. IHC to detect p-ERK confirmed activation of the Ras/MAPK cascade in all mutants expressing oncogenic K-ras (Figure 5D). Compared with PBCre⁺ controls, the PBCre⁺;LSL-K-ras^G12D/+, PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+, and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ mutants showed a significant increase in p-ERK expression, yet IHC analysis was not sensitive enough to detect a significant increase in p-ERK expression in double mutants compared with PBCre⁺;LSL-K-ras^G12D/+ mice. Nonetheless, IHC to detect the p-ERK downstream target p-ELK1 revealed a significant increase in PBCre⁺;Scrib^+/fl;LSL-K-ras^G12D/+ (2.1-fold) and PBCre⁺;Scrib^fl/fl;LSL-K-ras^G12D/+ (2.9-fold) mutants compared with PBCre⁺;LSL-K-ras^G12D/+ mice (Figure 5E). These data indicate that Scrib loss and oncogenic K-ras cooperate to promote prostate cancer progression via ELK1 activation, which leads to transcriptional activation of a variety of growth-regulatory genes that include the c-fos proto-oncogene (35). Although p-ELK1 is a direct target of the Ras/MAPK cascade, it is possible that additional pathways that mediate ELK1 activation, including JNK and p38 signaling (43), may also underpin the synergistic relationship between Scrib loss and K-ras activation in this setting.

SCRIB mislocalization predicts poor survival in human prostate cancer. To determine whether SCRIB is deregulated in human prostate cancer, we performed IHC to detect SCRIB on a prostate tissue microarray consisting of more than 2,000 hormone-naive patients with full clinical follow-up data (Supplemental Table 1) and assessed the intensity and mislocalization of SCRIB (Figure 6, A and B, Supplemental Tables 2 and 3, and ref. 44). SCRIB expression was observed in 98% (2,086 of 2,122) of all assessable samples. Non-assessable samples were omitted from our analysis (where clinical data were missing or tissue/tumor sections were absent; n = 1,139). SCRIB intensity significantly correlated with tumor stage, Gleason grade, and prostate-specific antigen (PSA) level (Supplemental Table 2). Despite a small, yet significant difference between low (+1) and moderate (+2) SCRIB expression predicting for poor survival (P = 0.0369, log-rank test), there was no association between moderate (+2) and high (+3) SCRIB expression (Figure 6C). Moreover, negative samples did not statistically correlate with PSA recurrence–free survival compared with positive samples and were predominantly atrophic (n = 31).

Figure 6

Human prostate tissue microarray. Representative images of TMA samples stained by IHC to detect SCRIB indicating the scoring strategy employed to grade SCRIB intensity (0 negative, +1 weak, +2 moderate, and +3 strong) (A) and mislocalization (B). Scale bars: 50 μm (larger panels) and 10 μm (insets). Samples displaying uniform basolateral membrane expression were designated as normal. Mislocalized expression was classified as clustered accumulation of SCRIB (group A: at cell-cell junctions; group B: along the membrane circumference; group A/B displays features of groups A and B). Scale bars: 50 μm (larger panels) and 10 μm (insets). (C) Kaplan-Meier plot indicates no significant difference between overall SCRIB intensity and PSA recurrence–free survival (P = 0.0687, log-rank test). Moreover, negative samples showed no significant difference in PSA recurrence–free survival compared with positive samples (P > 0.2196, log-rank test). (D) Kaplan-Meier plot comparing normal and mislocalized SCRIB expression revealed a significant reduction in PSA recurrence–free survival for all modes of mislocalization (P = 0.0001, log-rank test).

SCRIB mislocalization was observed in 7.1% (151 of 2,122) of all samples and positively correlated with tumor stage, Gleason grade, and PSA level (Supplemental Table 3). Mislocalization was categorized into 3 subgroups: punctate cell-cell junction clustering (A), accumulation along the circumference of the membrane (B), and a combination of features of both categories (A/B) (Figure 6B). Importantly, all modes of mislocalization significantly predicted for poor PSA recurrence–free survival (Figure 6D and Supplemental Table 3). This association may be attributable to the fact that SCRIB mislocalization has been linked to loss of function (17, 26, 45). Taken together, this evidence suggests that mislocalization of the polarity regulator SCRIB correlates with poor survival in human prostate cancer, highlighting the need for further investigation of polarity regulators in prostate cancer.

We have described what we believe to be a novel model of prostate cancer and have demonstrated that while Scrib depletion alone is a poor initiator of prostate neoplasia, following an oncogenic event, loss of Scrib contributes to tumor progression. These data establish a unique link between prostate cancer and loss of an epithelial polarity regulator and demonstrate that loss of cellular polarity is a tumor-initiating event and not simply a consequence of cellular transformation. Mechanistically, we have established that Ras/MAPK signaling is required for Scrib-deficient prostate hyperplasia in vivo using a MEK inhibitor, validating previous in vitro work (22, 23). Recent studies in human keratinocytes suggest that the mechanism whereby SCRIB suppresses RAS/MAPK signaling could result from a direct interaction between SCRIB and ERK, which diminishes ERK activation (22). The emerging complexity of SCRIB-mediated regulation of the RAS/MAPK cascade is further illustrated by SCRIB interactions with RSK2, a negative regulator of the pathway (46), and GIT1, an ARF-GAP that can act as a MEK-ERK scaffold (47, 48).

Furthermore, we demonstrate that Scrib loss and oncogenic K-ras cooperate to accelerate disease progression in mice, illustrating the multistep nature of prostate cancer progression and providing evidence to support published Drosophila in vivo and mammalian in vitro studies (23, 24). Mechanistically, this synergism correlated with elevated p-ELK1 expression, presenting a direct mechanism whereby Scrib loss and oncogenic K-ras synergize to drive prostate cancer progression. While elevated p-ELK1 expression is likely to reflect enhanced Ras/MAPK signaling (35), future work will be necessary to determine whether additional Scrib-mediated pathways, such as JNK, p38, Wnt, and Notch signaling, may also play a role in the cooperative relationship between Scrib loss and K-ras activation to drive murine prostate tumor progression (5). Notably, JNK and p38 signaling have previously been reported to activate ELK1 to mediate transcription of growth regulatory genes (43), and oncogenic Wnt signaling has been shown to cooperate with K-ras activation to accelerate prostate cancer progression in mice (41).

Currently, evidence for the role of SCRIB in human prostate cancer is conflicting. Although SCRIB maps to 8q24, which also harbors c-Myc and prostate stem cell antigen (PSCA) and is frequently amplified in prostate cancer (49), a microarray dataset shows a decrease in Scrib mRNA in human prostate carcinoma (50). Furthermore, integrated genomic profiling has recently shown that 33% (77 of 230, all samples, z-score threshold ± 1.0) of human prostate cancers display deregulated SCRIB expression (39). We have gained key insights into human prostate cancer, where mislocalization rather than loss of SCRIB predicts for poor survival. Given this result, it is tempting to speculate that SCRIB mislocalization reflects aberrant SCRIB function, as suggested by previous studies (17, 25, 26). Functional analysis of LRR SCRIB point mutant proteins shows restriction of SCRIB to the cytoplasm that is associated with loss of function in Caenorhabditis elegans (25), Drosophila (26), mice (29), and a human breast cancer cell line (17). The latter study postulates that the mislocalization of SCRIB disrupts its ability to recruit the β-PIX/GIT1 complex, resulting in Rac activation and tumor growth (17). Consequently, the mislocalization we observe in human prostate cancer is likely to inhibit its tumor-suppressive function. Despite advancements in large-scale sequencing, mutations in SCRIB that could cause its mislocalization remain to be identified in human cancers, suggesting other events may underlie this process.

Consistent with our findings in human prostate cancer, SCRIB is deregulated/mislocalized in human colorectal, breast, and endometrial cancers (14–18). This evidence supports the notion that SCRIB may play a broader role in epithelial tissue homeostasis, and our conditional Scrib-KO mouse model presents a potentially invaluable tool to address this prospect further.

Importantly, we show that Scrib^+/– prostate hyperplasia displayed mis-distribution of several polarity regulators and adhesion proteins, correlating with previous work (16, 32, 34) and raising the possibility that human prostate tumors displaying SCRIB mislocalization could exhibit a general polarity defect. However, it is possible that our study underestimates the number of tumors that have lost polarity, as this event may also stem from additional polarity regulator defects not associated with SCRIB deregulation. Our future work will assess the distribution of core polarity proteins in human prostate cancer and the molecular mechanism for codependent localization of polarity regulators to further understand the link between aberrant cellular polarity and prostate cancer.

Taken together, our data indicate that Scrib deficiency is a poor initiator of cancer and that an additional oncogenic or tumor suppressor mutation is required for Scrib deregulation to contribute to tumor progression. Mechanistic insights into how loss of polarity occurs in prostate epithelium and its impact on tumor onset and progression will improve our understanding of prostate cancer etiology and may facilitate the discovery of innovative prognostic factors and chemotherapeutic avenues involving the polarity network.

View Supplemental data

View Supplemental Excel file 1

Many thanks go to Michael Durrant, Josh Noske, Olivia Cakebread, Samantha Williams, Ryan Galea, and Lorey Smith at the Peter MacCallum Cancer Centre for their technical assistance. D.A. Tuveson acknowledges the support of the University of Cambridge, Cancer Research UK, and Hutchison Whampoa Limited. P.A. Perez-Mancera received financial support from the Ministerio de Educación y Ciencia of Spain and from Fundación Caja Madrid. This work was supported by the Prostate Cancer Foundation of Australia (PCFA) and the National Health and Medical Research Council (NHMRC). P.O. Humbert is supported by an NHMRC career development fellowship.

Address correspondence to: Patrick O. Humbert, Cell Cycle and Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, East Melbourne, VIC 3002, Australia. Phone: 61.0.3.96563526; Fax: 61.0.3.96561411; E-mail: Patrick.Humbert@petermac.org.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J Clin Invest. 2011;121(11):4257–4267. doi:10.1172/JCI58509.

1. Nelson WJ. Adaptation of core mechanisms to generate cell polarity. Nature. 2003;422(6933):766–774.
   View this article via: PubMed CrossRef Google Scholar
2. Dow LE, Humbert PO. Polarity regulators and the control of epithelial architecture, cell migration, and tumorigenesis. Int Rev Cytol. 2007;262:253–302.
   View this article via: PubMed CrossRef Google Scholar
3. Wodarz A, Nathke I. Cell polarity in development and cancer. Nat Cell Biol. 2007;9(9):1016–1024.
   View this article via: PubMed CrossRef Google Scholar
4. Huang L, Muthuswamy SK. Polarity protein alterations in carcinoma: a focus on emerging roles for polarity regulators. Curr Opin Genet Dev. 2010;20(1):41–50.
   View this article via: PubMed Google Scholar
5. Humbert PO, Grzeschik NA, Brumby AM, Galea R, Elsum I, Richardson HE. Control of tumourigenesis by the Scribble/Dlg/Lgl polarity module. Oncogene. 2008;27(55):6888–6907.
   View this article via: PubMed CrossRef Google Scholar
6. Bilder D, Li M, Perrimon N. Cooperative regulation of cell polarity and growth by Drosophila tumor suppressors. Science. 2000;289(5476):113–116.
   View this article via: PubMed CrossRef Google Scholar
7. Nakagawa S, Huibregtse JM. Human scribble (Vartul) is targeted for ubiquitin-mediated degradation by the high-risk papillomavirus E6 proteins and the E6AP ubiquitin-protein ligase. Mol Cell Biol. 2000;20(21):8244–8253.
   View this article via: PubMed Google Scholar
8. Gardiol D, Kuhne C, Glaunsinger B, Lee SS, Javier R, Banks L. Oncogenic human papillomavirus E6 proteins target the discs large tumour suppressor for proteasome-mediated degradation. Oncogene. 1999;18(40):5487–5496.
   View this article via: PubMed Google Scholar
9. Nakagawa S, et al. Analysis of the expression and localisation of a LAP protein, human scribble, in the normal and neoplastic epithelium of uterine cervix. Br J Cancer. 2004;90(1):194–199.
   View this article via: PubMed CrossRef Google Scholar
10. Cavatorta AL, et al. Differential expression of the human homologue of drosophila discs large oncosuppressor in histologic samples from human papillomavirus-associated lesions as a marker for progression to malignancy. Int J Cancer. 2004;111(3):373–380.
    View this article via: PubMed CrossRef Google Scholar
11. Lisovsky M, et al. Cell polarity protein Lgl2 is lost or aberrantly localized in gastric dysplasia and adenocarcinoma: an immunohistochemical study. Mod Pathol. 2009;22(7):977–984.
    View this article via: PubMed CrossRef Google Scholar
12. Lisovsky M, Ogawa F, Dresser K, Woda B, Lauwers GY. Loss of cell polarity protein Lgl2 in foveolar-type gastric dysplasia: correlation with expression of the apical marker aPKC-zeta. Virchows Arch. 2010;457(6):635–642.
    View this article via: PubMed Google Scholar
13. Fuja TJ, Lin F, Osann KE, Bryant PJ. Somatic mutations and altered expression of the candidate tumor suppressors CSNK1 epsilon, DLG1, and EDD/hHYD in mammary ductal carcinoma. Cancer Res. 2004;64(3):942–951.
    View this article via: PubMed CrossRef Google Scholar
14. Gardiol D, Zacchi A, Petrera F, Stanta G, Banks L. Human discs large and scrib are localized at the same regions in colon mucosa and changes in their expression patterns are correlated with loss of tissue architecture during malignant progression. Int J Cancer. 2006;119(6):1285–1290.
    View this article via: PubMed Google Scholar
15. Kamei Y, et al. Human scribble accumulates in colorectal neoplasia in association with an altered distribution of beta-catenin. Hum Pathol. 2007;38(8):1273–1281.
    View this article via: PubMed Google Scholar
16. Navarro C, et al. Junctional recruitment of mammalian Scribble relies on E-cadherin engagement. Oncogene. 2005;24(27):4330–4339.
    View this article via: PubMed CrossRef Google Scholar
17. Zhan L, et al. Deregulation of scribble promotes mammary tumorigenesis and reveals a role for cell polarity in carcinoma. Cell. 2008;135(5):865–878.
    View this article via: PubMed CrossRef Google Scholar
18. Ouyang Z, Zhan W, Dan L. hScrib, a human homolog of Drosophila neoplastic tumor suppressor, is involved in the progress of endometrial cancer. Oncol Res. 2010;18(11–12):593–599.
    View this article via: PubMed CrossRef Google Scholar
19. Kallay LM, McNickle A, Brennwald PJ, Hubbard AL, Braiterman LT. Scribble associates with two polarity proteins, Lgl2 and Vangl2, via distinct molecular domains. J Cell Biochem. 2006;99(2):647–664.
    View this article via: PubMed CrossRef Google Scholar
20. Metais JY, Navarro C, Santoni MJ, Audebert S, Borg JP. hScrib interacts with ZO-2 at the cell-cell junctions of epithelial cells. FEBS Lett. 2005;579(17):3725–3730.
    View this article via: PubMed CrossRef Google Scholar
21. Takizawa S, et al. Human scribble, a novel tumor suppressor identified as a target of high-risk HPV E6 for ubiquitin-mediated degradation, interacts with adenomatous polyposis coli. Genes Cells. 2006;11(4):453–464.
    View this article via: PubMed Google Scholar
22. Nagasaka K, et al. The cell polarity regulator hScrib controls ERK activation through a KIM site-dependent interaction. Oncogene. 2010;29(38):5311–5321.
    View this article via: PubMed CrossRef Google Scholar
23. Dow LE, Elsum IA, King CL, Kinross KM, Richardson HE, Humbert PO. Loss of human Scribble cooperates with H-Ras to promote cell invasion through deregulation of MAPK signalling. Oncogene. 2008;27(46):5988–6001.
    View this article via: PubMed Google Scholar
24. Brumby AM, Richardson HE. scribble mutants cooperate with oncogenic Ras or Notch to cause neoplastic overgrowth in Drosophila. EMBO J. 2003;22(21):5769–5779.
    View this article via: PubMed Google Scholar
25. Legouis R, Jaulin-Bastard F, Schott S, Navarro C, Borg JP, Labouesse M. Basolateral targeting by leucine-rich repeat domains in epithelial cells. EMBO Rep. 2003;4(11):1096–1102.
    View this article via: PubMed CrossRef Google Scholar
26. Zeitler J, Hsu CP, Dionne H, Bilder D. Domains controlling cell polarity and proliferation in the Drosophila tumor suppressor Scribble. J Cell Biol. 2004;167(6):1137–1146.
    View this article via: PubMed Google Scholar
27. Lakso M, et al. Efficient in vivo manipulation of mouse genomic sequences at the zygote stage. Proc Natl Acad Sci U S A. 1996;93(12):5860–5865.
    View this article via: PubMed CrossRef Google Scholar
28. Murdoch JN, et al. Disruption of scribble (Scrb1) causes severe neural tube defects in the circletail mouse. Hum Mol Genet. 2003;12(2):87–98.
    View this article via: PubMed CrossRef Google Scholar
29. Zarbalis K, May SR, Shen Y, Ekker M, Rubenstein JL, Peterson AS. A focused and efficient genetic screening strategy in the mouse: identification of mutations that disrupt cortical development. PLoS Biol. 2004;2(8):E219.
    View this article via: PubMed Google Scholar
30. Fehon RG, McClatchey AI, Bretscher A. Organizing the cell cortex: the role of ERM proteins. Nat Rev Mol Cell Biol. 2010;11(4):276–287.
    View this article via: PubMed CrossRef Google Scholar
31. Xian W, Schwertfeger KL, Vargo-Gogola T, Rosen JM. Pleiotropic effects of FGFR1 on cell proliferation, survival, and migration in a 3D mammary epithelial cell model. J Cell Biol. 2005;171(4):663–673.
    View this article via: PubMed CrossRef Google Scholar
32. Bilder D, Perrimon N. Localization of apical epithelial determinants by the basolateral PDZ protein Scribble. Nature. 2000;403(6770):676–680.
    View this article via: PubMed CrossRef Google Scholar
33. Bilder D. Epithelial polarity and proliferation control: links from the Drosophila neoplastic tumor suppressors. Genes Dev. 2004;18(16):1909–1925.
    View this article via: PubMed CrossRef Google Scholar
34. Bilder D, Schober M, Perrimon N. Integrated activity of PDZ protein complexes regulates epithelial polarity. Nat Cell Biol. 2003;5(1):53–58.
    View this article via: PubMed CrossRef Google Scholar
35. Marais R, Wynne J, Treisman R. The SRF accessory protein Elk-1 contains a growth factor-regulated transcriptional activation domain. Cell. 1993;73(2):381–393.
    View this article via: PubMed Google Scholar
36. Jin C, McKeehan K, Wang F. Transgenic mouse with high Cre recombinase activity in all prostate lobes, seminal vesicle, and ductus deferens. Prostate. 2003;57(2):160–164.
    View this article via: PubMed CrossRef Google Scholar
37. Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat Genet. 1999;21(1):70–71.
    View this article via: PubMed Google Scholar
38. Pagliarini RA, Xu T. A genetic screen in Drosophila for metastatic behavior. Science. 2003;302(5648):1227–123.
39. Taylor BS, et al. Integrative genomic profiling of human prostate cancer. Cancer Cell. 2010;18(1):11–22.
    View this article via: PubMed CrossRef Google Scholar
40. Gioeli D, Mandell JW, Petroni GR, Frierson HF Jr, Weber MJ. Activation of mitogen-activated protein kinase associated with prostate cancer progression. Cancer Res. 1999;59(2):279–284.
    View this article via: PubMed Google Scholar
41. Pearson HB, Phesse TJ, Clarke AR. K-ras and Wnt signaling synergize to accelerate prostate tumorigenesis in the mouse. Cancer Res. 2009;69(1):94–101.
    View this article via: PubMed CrossRef Google Scholar
42. Scherl A, Li JF, Cardiff RD, Schreiber-Agus N. Prostatic intraepithelial neoplasia and intestinal metaplasia in prostates of probasin-RAS transgenic mice. Prostate. 2004;59(4):448–459.
    View this article via: PubMed CrossRef Google Scholar
43. Whitmarsh AJ, Yang SH, Su MS, Sharrocks AD, Davis RJ. Role of p38 and JNK mitogen-activated protein kinases in the activation of ternary complex factors. Mol Cell Biol. 1997;17(5):2360–2371.
    View this article via: PubMed Google Scholar
44. Fleischmann A, et al. Immunological microenvironment in prostate cancer: high mast cell densities are associated with favorable tumor characteristics and good prognosis. Prostate. 2009;69(9):976–981.
    View this article via: PubMed CrossRef Google Scholar
45. Nagasaka K, et al. Human homolog of Drosophila tumor suppressor Scribble negatively regulates cell-cycle progression from G1 to S phase by localizing at the basolateral membrane in epithelial cells. Cancer Sci. 2006;97(11):1217–1225.
    View this article via: PubMed CrossRef Google Scholar
46. Zeniou-Meyer M, et al. The Coffin-Lowry syndrome-associated protein RSK2 is implicated in calcium-regulated exocytosis through the regulation of PLD1. Proc Natl Acad Sci U S A. 2008;105(24):8434–8439.
    View this article via: PubMed CrossRef Google Scholar
47. Yin G, Haendeler J, Yan C, Berk BC. GIT1 functions as a scaffold for MEK1-extracellular signal-regulated kinase 1 and 2 activation by angiotensin II and epidermal growth factor. Mol Cell Biol. 2004;24(2):875–885.
    View this article via: PubMed Google Scholar
48. Audebert S, et al. Mammalian Scribble forms a tight complex with the betaPIX exchange factor. Curr Biol. 2004;14(11):987–995.
    View this article via: PubMed CrossRef Google Scholar
49. Witte JS. Multiple prostate cancer risk variants on 8q24. Nat Genet. 2007;39(5):579–580.
    View this article via: PubMed CrossRef Google Scholar
50. Su AI, et al. Molecular classification of human carcinomas by use of gene expression signatures. Cancer Res. 2001;61(20):7388–7393.
    View this article via: PubMed Google Scholar
51. Jackson EL, et al. Analysis of lung tumor initiation and progression using conditional expression of oncogenic K-ras. Genes Dev. 2001;15(24):3243–3248.
    View this article via: PubMed CrossRef Google Scholar
52. Shappell SB, et al. Prostate pathology of genetically engineered mice: definitions and classification. The consensus report from the Bar Harbor meeting of the Mouse Models of Human Cancer Consortium Prostate Pathology Committee. Cancer Res. 2004;64(6):2270–2305.
    View this article via: PubMed Google Scholar
53. Kinkade CW, et al. Targeting AKT/mTOR and ERK MAPK signaling inhibits hormone-refractory prostate cancer in a preclinical mouse model. J Clin Invest. 2008;118(9):3051–3064.
    View this article via: JCI PubMed Google Scholar
54. Pearson H, McCarthy A, Collins C, Ashworth A, Clarke A. Lkb1 deficiency causes prostate neoplasia in the mouse. Cancer Res. 2008;68(7):2223–2232.
    View this article via: PubMed CrossRef Google Scholar
55. Dow LE, et al. The tumour-suppressor Scribble dictates cell polarity during directed epithelial migration: regulation of Rho GTPase recruitment to the leading edge. Oncogene. 2007;26(16):2272–2282.
    View this article via: PubMed CrossRef Google Scholar
56. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25(4):402–408.
    View this article via: PubMed Google Scholar
57. Bubendorf L, et al. Survey of gene amplifications during prostate cancer progression by high-throughout fluorescence in situ hybridization on tissue microarrays. Cancer Res. 1999;59(4):803–806.
    View this article via: PubMed Google Scholar
58. El Gammal AT, et al. Chromosome 8p deletions and 8q gains are associated with tumor progression and poor prognosis in prostate cancer. Clin Cancer Res. 2010;16(1):56–64.
    View this article via: PubMed CrossRef Google Scholar