If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address reprint request to Silvano Bosari, M.D., Division of Pathology, Department of Medicine, Surgery and Dentistry, University of Milan Medical School, Fondazione IRCCS Ca' Granda, 20122 Milan, Italy
Human Scribble (Scrib) is an evolutionary-conserved cell polarity protein, but its potential role in human cancer is controversial. Herein, we show that Scrib is nearly universally overexpressed in cultured tumor cell lines and genetically disparate cancer patient series compared with matched normal tissues in vivo. Instead of a membrane association seen in normal epithelia, tumor-associated Scrib is mislocalized and found predominantly in the cytosol. Small-interfering RNA silencing of Scrib in model lung adenocarcinoma A549 cells inhibited cell migration in wound-healing assays, suppressed tumor cell invasion across Matrigel-coated inserts, and down-regulated the expression of cell motility markers and mediators of epithelial-mesenchymal transition. These data uncover a previously unrecognized exploitation of Scrib for aberrant tumor cell motility and invasion, thus potentially contributing to disease progression in humans.
Human Scribble (Scrib) is an evolutionary-conserved member of the LAP (PDZ and LRR) gene family
implicates Scrib as a general regulator of directional cell motility, largely via the assembly of multiprotein complexes and the activation of small GTPase signaling at the leading edge of migrating cells.
its role in human cancer, in vivo, has remained vehemently debated. Although initial evidence suggested that Scrib expression is lost during human tumorigenesis, other reports have shown that Scrib levels are increased
for other cell polarity proteins of the Par complex (Par6β and α protein kinase C), which have been associated with both tumor suppression and oncogenic functions.
In this study, we used a global expression strategy to map the distribution of Scrib in human cancer and to test the impact of this pathway on mechanisms of tumor progression.
Materials and Methods
Representative samples of cancers of the colon, liver, prostate, uterus, thyroid, lung, bladder, breast, ovary, and stomach (10 cases each) were selected for Cancer Universal TMA construction. An equal number of cases of lymphomas, tumors of the central nervous system, and testicular cancer were also included in the Cancer Universal TMA. The clinical-pathological characteristics of the patients used in this study are given in Supplemental Table S1 (available at http://ajp.amjpathol.org). For TMA construction of non–small cell lung cancer (NSCLC), 226 patients who underwent surgical resection for curative purposes from September 1999 to December 2004 were selected. Of these patients, 154 (68%) had adenocarcinomas, 64 (28%) had squamous cell carcinomas (SCCs), 5 (2%) had adenosquamous carcinomas, and 3 (1%) had large-cell carcinomas. Twenty paired NSCLC and normal counterparts from this series were selected for mRNA analysis, including 9 adenocarcinomas (45%), 9 SCCs (45%), 1 adenosquamous carcinoma (5%), and 1 large-cell carcinoma (5%).
The breast cancer series used in these studies included five ductal in situ carcinomas, 38 invasive ductal carcinomas, two papillary carcinomas, one mucinous cancer, and one lobular cancer. Three breast specimens collected from plastic surgery procedures, with no evidence of neoplasia, were used as normal controls.
Biopsy specimens from five cervical dysplasias (four cervical intraepithelial neoplasias II and 1 cervical intraepithelial neoplasia III), one carcinoma in situ, and two infiltrating cervical cancers were used. All patients in this series were positive for human papillomavirus (HPV) infection by PCR amplification using HPV-specific primers
For Cancer Universal TMA, one tumoral area from each specimen was punched. For NSCLC and breast cancer TMAs, four tumor areas were selected from each tissue block. Samples of nonneoplastic parenchyma, including 16 cores of normal lung and 20 cores of normal mammary glands, were added to each TMA block. For quality control, a 4-μm-thick section was cut from each TMA block, stained with H&E, and analyzed by immunohistochemistry (IHC).
For antigen retrieval, slides were microwaved for 35 minutes in citrate solution. Sections (4-μm thick) were cut from TMA or full-section blocks and stained with a goat polyclonal antibody specific to Scrib (see Supplemental Figure S1A at http://ajp.amjpathol.org) (1:100, D-8; Santa Cruz Biotechnology, Santa Cruz, CA) for 1 hour at 22°C. IHC was performed using an autostainer (Optimax i6000; Biogenex, San Ramon, CA), with detection of antibody reactivity with a kit (EnVision Detection Kit; Dako, Milan, Italy) using peroxidase-diaminobenzidine as the chromogen. All slides were counterstained with hematoxylin. Immunoreactivity for Scrib in the various samples was evaluated by two pathologists (M.M. and S.B.) and independently scored for membrane-associated or cytoplasmic localization. When discrepancies in Scrib scoring occurred, a consensus interpretation was reached after reexamination. Because Scrib immunoreactivity was either present or absent in all epithelial cells, normal and neoplastic samples were scored as follows (according to immunoreactivity intensity): 0, negative; 1, weak; 2, moderate; or 3, strong.
Cell Lines, Transfections, and Immunoblotting
Human tumor cell lines representative of malignant mesothelioma (MSTO-211H) and cancers of the colon (LoVo and SW1116), breast (MCF-7, MDA-MB231, and SKBR3), lung (A549), stomach (MKN45 and KATO-III), uterus (Hela), prostate (DU145), and liver (HepG2) were purchased from American Type Culture Collection (Manassas, VA) and maintained in culture, as recommended by the supplier. A549 cells (2 × 106 per well in six-well plates) were transfected with 100 pmol of nontargeting or Scrib-directed small-interfering RNA (siRNA) (siGENOME Smart Pool, catalogue numbers M-010500, D-001206; Dharmacon, Lafayette, CO) and 5 μL Lipofectamine 2000 (Invitrogen, Life Technologies Corporation, Carlsbad, CA). In some experiments, transfected A549 cells were harvested after 48 hours and solubilized in 150 μL radioimmunoprecipitation assay buffer plus protease inhibitors (Roche, Basel, Switzerland); aliquots of cell lysates (50 μg) were probed with antibodies (1 μg/μL) to Scrib (Santa Cruz Biotechnology), Snai2 (LifeSpan Biosciences, Seattle, WA), β-catenin (Thermoscientific, Waltham, MA), vimentin (Dako), Zeb-1 (Sigma-Aldrich, Milan, Italy), FAK-Ptk2 (Santa Cruz Biotechnology), or β-actin (Sigma-Aldrich). Reactive bands were visualized with ECL Plus (GE Healthcare, Milan, Italy).
Total RNA, isolated from 20 cases of NSCLC and matched normal tissue using reagent (Trizol; Invitrogen), was reverse transcribed
and analyzed for changes in Scrib mRNA expression relative to β2-microglobulin using TaqMan gene expression assays (Hs00363005_m1 and Hs99999907_m1, respectively; Applied Biosystems, Life Technologies Corporation, Carlsbad, CA), according to the 2-ΔCt formula.
Cell Migration and Cell Invasion Assays
Monolayers of transfected A549 cells (70% confluency) were analyzed for changes in cell migration using a wound-healing assay during a 24-hour interval. Quantification of migration distance was determined as the reduction in the wound's gap using computer software (NIH ImageJ). In other experiments, transfected A549 cells were analyzed for cell invasion across Matrigel-coated 8.0-μm inserts (Transwell; BD Bioscience, Milan, Italy) using serum as the chemoattractant for 24 hours. The cells on the apical surface of each insert were scraped off; and membranes with invaded cells were fixed in 100% methanol, stained with toluidine blue (Sigma-Aldrich), and mounted on microscope slides. Invading cells were counted after photographing the membranes, and both peripheral and central areas of the insert were analyzed.
Differential Scrib expression was analyzed in published microarray data sets using a Web-based data-mining platform (Oncomine; available at http://www.oncomine.org). Relative expression data of Scrib were collected from solid malignancies only, using P ≤ 0.001 as the cutoff (t-statistic provided within the data sets in Oncomine).
All cell-based experiments were performed at least three times in duplicate. Differences among sample groups were analyzed using the unpaired Student's t-test function of GraphPad Prism version 4 software. P < 0.05 was considered statistically significant.
All tumor tissues in the Cancer Universal TMA displayed strong Scrib immunoreactivity, in vivo (IHC intensity >1, Figure 1A ), except for thyroid cancers (IHC intensity = 0.7, Figure 1A). The highest expression of Scrib was observed in cancers of the colon, bladder, prostate, uterus, and ovaries (IHC intensity >2; Figure 1, A and B), thus independently of tumor histotype (see Supplemental Table S1 at http://ajp.amjpathol.org). Scrib was also prominently expressed in all 11 cultured tumor cell lines examined by IHC (Figure 1C).
Analysis of normal pneumocytes or normal bronchial epithelia revealed undetectable to low levels of Scrib, respectively (Figure 2A). In contrast, Scrib was prominently expressed in tumor tissues of a large series (226 patients) of NSCLCs (IHC average immunoreactivity = 1.34; tumor versus normal, P = 1.87 × 10−8; Figure 2B). For NSCLC subtypes, adenocarcinomas expressed comparatively higher levels of Scrib protein (Figure 2B) and Scrib mRNA (Figure 2C) than SCCs [IHC mean immunoreactivity: adenocarcinoma, 1.54; and SCC, 0.83 (P = 2.9 × 10−9); and fold increase of mRNA levels in adenocarcinoma over SCC, 2.5 (P = 0.0082)]. However, either histotype had higher levels of Scrib compared with normal parenchyma (adenocarcinoma versus normal, P = 2.78 × 10−11 and P = 0.0017 for IHC and gene expression analysis, respectively; and SCC versus normal, P = 0.0002 and P = 0.04 for IHC and gene expression analysis, respectively; Figure 2, B and C).
cancers. Although Scrib was detectable in normal mammary glands with a membranous distribution at cell-cell junctions, tumor-associated Scrib was subcellularly mislocalized and exhibited a predominantly diffuse cytoplasmic reactivity (see Supplemental Figure S1D at http://ajp.amjpathol.org), similar to earlier reports
Bioinformatic analysis of published microarray data sets (Oncomine) revealed that Scrib was uniformly expressed at high levels in genetically disparate cancer types, compared with matched normal tissues (see Supplemental Figure S2A at http://ajp.amjpathol.org). Although Scrib deregulation cooperates with mutant k-Ras
to enhance tumorigenesis, in our analysis, Scrib levels in human tumors were independent of mutant k-Ras or HPV protein expression (see Supplemental Figure S2, B and C, at http://ajp.amjpathol.org). Accordingly, strong Scrib immunoreactivity was observed in dysplastic or cancerous cervical lesions, compared with moderate positivity detected in matched nontransformed epithelium (n = 8; see Supplemental Figure S3A at http://ajp.amjpathol.org). In this series and similar to the IHC results obtained with breast cancer, Scrib localization in cervical lesions was primarily diffuse and cytosolic, as opposed to its predominant membrane association observed for the normal mucosa (see Supplemental Figure S3B at http://ajp.amjpathol.org).
Transfection of lung adenocarcinoma A549 cells with Scrib-directed siRNA efficiently silenced the intended target, whereas control siRNA had no effect (Figure 3A). Under these conditions, siRNA depletion of Scrib strongly impaired tumor cell migration in a wound closure assay, in vitro (P < 0.0001, Figure 3B); and profoundly inhibited tumor cell invasion across Matrigel-coated inserts (P = 0.02, Figure 3C). In contrast, control nontargeting siRNA had no effect on Scrib levels (Figure 3A) and did not impair tumor cell migration or invasion (Figure 3, B and C). In addition, siRNA depletion of Scrib in A549 cells resulted in reduced expression of cell motility markers and proteins associated with epithelial-mesenchymal transition, including focal adhesion kinase (Ptk2-Fak), β-catenin, and Snai2 (Figure 3D). In contrast, the expression of vimentin or Zeb-1 was only modestly affected; and a control siRNA had no effect on protein levels (Figure 3D).
In this study, we showed that Scrib, a pivotal cell polarity module
is almost universally overexpressed and subcellularly mislocalized in disparate human cancers, in vivo, independently of oncogenic signaling or viral (ie, HPV) transformation. Nevertheless, Scrib diffusion into the cytoplasm could also be related to saturation of membrane-free positions consequent to Scrib overexpression. Mechanistically, short-term silencing of Scrib in model lung cancer cells profoundly impairs tumor cell migration and invasion and down-regulates several pivotal markers of cell motility and epithelial-mesenchymal transition. These results are at variance with a prevailing model that implicates Scrib in an evolutionary-conserved mechanism of tumor suppression
and suggest that this pathway is broadly exploited during human tumorigenesis, in vivo, enhancing aberrant tumor cell motility and invasion. Thus, the findings potentially contribute to metastatic dissemination and disease progression in humans.
This further contributes to the acquisition of a migratory epithelial-mesenchymal transition–like cellular phenotype. In this context, therapeutic targeting of the Scrib pathway may selectively disrupt pivotal mechanisms of random tumor cell motility and invasion
Scrib protein expression in human breast pathology. A: Validation of Scrib-antibody specificity. Immunohistochemistry with the primary antibody to Scrib was performed in A549 cells transfected with control or Scrib-targeting siRNA. B: Scrib protein expression was evaluated in a series of normal breast glands, ductal carcinoma in situ, and invasive cancers using a tissue microarray platform. C: Quantification of Scrib immunoreactivity (IHC intensity) in normal mammary parenchyma and breast diseases. The number of available areas per tissue category is indicated at the bottom of the bars. D: Representative images of Scrib expression pattern in normal breast glands, ductal carcinoma in situ (DCIS), HR-positive/Her2-negative (HR-POS/Her2-NEG) and HR-negative/Her2-positive (HR- NEG/Her2-POS) invasive cancers. Original magnification x80 (cores) and x400 (enlargments). *P < 0.05; **P < 0.01; ***P < 0.001
Oncomine Cancer Microarray database analysis of Scrib gene expression profiles. Public microarray datasets (Oncomine repository) were queried for expression of Scrib in carcinomas versus normal tissues (A) or in neoplasia with correlated molecular analysis, as k-Ras mutation status (B) or HPV-infection presence (C). Datasets are reported using the name of the involved organ and the name of the author as it appears in Oncomine repository. (ns, not significant; Ca, Carcinoma; Ad-Ca, Adenocarcinoma; SCC, Squamous Cell Carcinoma; SCLC, Small Cell Lung Cancer; HCC, Hepatocellular Carcinoma; AA, Anaplastic Astrocytoma; HOG, Human Oligodendroglioma; GBM, Glioblastoma; DA, Diffuse Astrocytoma; PCA, Prostate cancer; KRAS-WT, k-Ras wild-type; KRAS-Mut, Mutated k-Ras; HPV, Human Papilloma Virus).
A: Scrib expression was evaluated by IHC in cervical lesions and matched normal epithelium in cases of cervical dysplasia (CIN) or invasive carcinomas (SCC) positive for Human Papilloma Virus infection. B: Representative images of cervical lesions and normal epithelium. Original magnification x400. CIN, Cervical intraepithelial neoplasia; ISC, in situ cervical carcinoma; SCC, Squamous cell carcinoma.
Supported by a fellowship of the Doctorate School of Molecular Medicine at Università degli Studi di Milano (A.F.) and in part by grants from the National Institutes of Health (CA140043, CA78810, and CA118005 to D.C.A.) and the Fondazione Berlucchi (S.B.).
D.C.A. and S.B. contributed equally to this work.
Supplemental material for this article can be found at http://ajp.amjpathol.org or at doi: 10.1016/j.ajpath.2011.02.028.