Gli1 enhances migration and invasion via up-regulation of MMP-11 and promotes metastasis in ERα negative breast cancer cell lines
Gli1 is an established oncogene and its expression in Estrogen Receptor (ER) α negative and triple negative breast cancers is predictive of a poor prognosis; however, the biological functions regulated by Gli1 in breast cancer have not been extensively evaluated. Herein, Gli1 was over-expressed or down-regulated (by RNA interference and by expression of the repressor form of Gli3) in the ERα negative, human breast cancer cell lines MDA-MB-231 and SUM1315. Reduced expression of Gli1 in these two cell lines resulted in a decrease in migration and invasion. Gli1 over-expression increased the migration and invasion of MDA-MB-231 cells with a corresponding increase in expression of MMP-11. Silencing MMP-11 in MDA-MB-231 cells that over-expressed Gli1 abrogated the Gli1-induced enhancement of migration and invasion. Sustained suppression of Gli1 expression decreased growth of MDA-MB-231 in vitro by increasing apoptosis and decreasing proliferation. In addition, silencing of Gli1 reduced the numbers and sizes of pulmonary metastases of MDA-MB-231 in an in vivo experimental metastasis assay. In summary, Gli1 promotes the growth, survival, migration, invasion and metastasis of ERα negative breast cancer. Additionally, MMP-11 is up-regulated by Gli1 and mediates the migration and invasion induced by Gli1 in MDA-MB-231.
KeywordsBreast cancer Gli1 Gli-mediated transcription MMP-11 Invasion Migration Metastasis
Analysis of variance
Complementary deoxyribonucleic acid
Ductal carinoma in situ
Dulbecco’s Modification of Eagle’s Medium
Epidermal growth factor
Extracellular signal regulated kinase
Estrogen receptor α
Fetal bovine serum
Insulin like growth factor
Mouse mammary tumor virus
Polymerase chain reaction
Quantitative, reverse transcription PCR
Ribosomal protein large 0
Short hairpin ribonucleic acid
Small interfering ribonucleic acid
Transforming growth factor β
Gli1 is a zinc finger transcription factor and a member of the vertebrate Gli family. The Gli transcription factors, which in addition to Gli1 include Gli2 and Gli3, are effectors of the Hedgehog signaling pathway. The Gli transcription factors coordinately regulate Gli-mediated transcription ; however, Gli1 functions as the terminal and critical activator of Gli-mediated transcription . Although Gli1 was first identified through its role as a transcriptional mediator of Hedgehog signaling, it is now known that Gli1 expression and activity are also modulated by other signaling pathways, including Transforming Growth Factor β (TGFβ), Ras/ERK, and Wnt [3, 4, 5, 6].
Gli1 plays an important role in the initiation and progression of several types of cancer. Gli1 was first identified by its amplification in human glioma . Additionally, Gli1 drives the development of cancers associated with Gorlin’s syndrome, which include basal cell carcinomas, medulloblastomas and rhabdomyosarcomas and result from inappropriate activation of Hedgehog signaling by mutations of pathway members . Gli1 activity has also been shown to promote the growth, migration, invasion and/or metastasis of several other cancer types, including cancers of the prostate and pancreas [9, 10, 11].
Several lines of evidence indicate that Gli1 contributes to breast cancer development and progression. Conditional expression of Gli1 under the regulation of the MMTV promoter induces mammary carcinomas in transgenic mice. These Gli1-induced carcinomas are Estrogen Receptor (ER) α negative and have features reminiscent of the basal subtype of human breast cancers . Over-expression of Gli1 in human breast cancer tissue and cell lines is well-documented [13, 14]. In addition, Gli1 expression has been shown to be an indicator of a poor prognosis in human breast cancers [14, 15, 16]. Our prior work found that nuclear localization of Gli1 protein indicates a poor prognosis in women with ERα negative and triple negative breast cancers (i.e., those breast cancers lacking ERα, Progesterone Receptor (PR) and amplification of Her2/neu ), but not in those with ERα positive cancers .
Although it has been shown that Gli1 predicts a poor outcome in ERα negative and triple negative breast cancers, the functions of Gli1 in these cancer types have not been extensively studied. To examine the role of Gli1 in the progression of ERα negative breast cancers, we studied whether modulation of Gli1 expression would influence the migration, invasion and metastasis of cell lines representative of ERα negative breast cancer. In addition, we found that Gli1 up-regulates the expression of the matrix metalloproteinase (MMP) MMP-11 in these breast cancer cell lines and studied its contribution to Gli1-induced migration and invasion. Herein, we provide evidence that Gli1 promotes the migration, invasion, and metastasis of ERα negative breast cancer.
Materials and methods
Cell lines and culture conditions
The MDA-MB-231 (231) breast cancer cell line was a gift from Dr. Janet Price, MD Anderson Cancer Center (Houston, TX), and was maintained in Dulbecco’s Modification of Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS). The SUM1315 cells were a gift of Dr. Stephen Ethier, Karmanos Cancer Institute (Detroit, MI), and maintained in Ham’s F-12 media supplemented with 5% FBS, Insulin (10 μg/ml), and Epidermal Growth Factor (EGF) (20 ng/ml). MCF10A cells were purchased from the American Type Culture Collection and were maintained as previously described .
Production of viral supernatants and transduction
For expression of Gli1, the pLJD-HA-Gli1 retroviral construct and its corresponding empty vector control were gifts of Dr. Michael Ruppert . Retroviral supernatants were prepared from Bing cells, and 231 and MCF10A cells were transduced as described previously . After transduction, cells were mass selected with 500–700 μg/ml G418 for at least 2 weeks. For silencing Gli1 expression, lentiviral supernatants were prepared using the pLKO.1-puro constructs containing shRNA targeting human Gli1 (TRCN0000020484, TRCN0000020488; Mission shRNA plasmid DNA, Sigma-Aldrich, Inc.) or a non-targeting control (NT) shRNA (Sigma-Aldrich, Inc.), as per the manufacturer’s protocol. Transduction of 231 cells was followed by mass selection in 20 μg/ml puromycin for at least 3 days. For expression of Gli3R, the pLenti6/U6-Gli3R lentiviral construct and its corresponding vector control (pLenti6/U6-β-galactosidase) were gifts of Dr. Bradley Yoder, University of Alabama at Birmingham (Birmingham, AL). Lentiviral supernatants were prepared from 293T cells, and 231 cells were transduced as described above.
Quantitative, reverse transcription PCR
RNA was extracted and subjected to DNase pretreatment prior to cDNA synthesis (High-Capacity cDNA Reverse Transcription kit, Applied Biosystems, Inc.). Primer and probe sets for Gli1, MMP-11 and Ribosomal Protein Large 0 (RPLP0) were purchased from Applied Biosystems (TaqMan Gene Expression Assays-on-Demand, Hs00171790_m1, Hs00968295_m1 and Hs99999902_m1, respectively). Gene assays were performed using TaqMan Universal PCR Master Mix (Applied Biosystems, Inc.). Fluorescent signal data were collected by the Applied Biosystems Prism 7700 Sequence Detection System. RPLP0 was used as the internal reference. Relative expression values were calculated using the ΔΔCT method.
Western blot analysis
For detection of Gli1 and cleaved poly (ADP-ribose) polymerase (PARP), cells were lysed in RIPA lysis buffer (50 mM Tris–HCl at pH 7.4, 2 mM dithiothreitol, 1% Triton X-100, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, and 7× protease inhibitor cocktail (Roche Applied Science)). Fifty micrograms of total protein was separated on 4–15% gradient gels and transferred to nitrocellulose membranes. The membranes were blocked with 5% dry milk in 0.1% PBS-T for 1 h at room temperature, followed by incubation with the following primary antibodies individually at 4°C overnight: a mouse monoclonal antibody to HA (1:1,000 dilution, clone 16B12, Covance, Inc.), a rabbit polyclonal antibody to Gli1 (1:2,500 dilution, a gift of Dr. Michael Ruppert), a mouse monoclonal antibody to cleaved PARP (1:1,000 dilution, Asp214, Cell Signaling Technology), or a mouse monoclonal antibody to β-actin (1:20,000 dilution, Sigma-Aldrich, Inc.). Secondary detection was achieved with horseradish peroxidase (HRP)-conjugated secondary antibodies (Biorad Laboratories) and chemiluminescent HRP substrate.
For detection of secreted MMP-11, 1.5 × 106 cells were plated in a 60-mm dish. After attachment, the cells were washed with cold phosphate buffered saline (PBS) and 2.0 ml of serum-free DMEM was added to each plate. Forty-eight hours later, the medium was collected and centrifuged. The volume of medium loaded for protein separation was normalized to the numbers of cells on each plate at the time of media collection. A lysate of a human cell line containing a high level of MMP-11 protein (ThermoScientific, MS-1035-PCL) was included as a positive control. The remaining procedure was as described above except that the membranes were incubated with mouse monoclonal antibody to MMP-11 (1:500 dilution, clone SL3.05, Neomarkers/ThermoFisher Scientific).
Transwell migration and invasion assays
For transwell migration assays, 1 × 105 cells were plated in 24-well inserts (8-μm pore size, BD Biosciences). For SUM1315 cells, cells were incubated in medium with 2% FBS on the transwell inserts (i.e., top well), and medium containing 5% FBS was placed in the bottom well. After 24 h, cells on the upper surface of the transwell filter were removed, the filters were stained with hematoxylin and eosin, and the number of cells that migrated through the filters was counted in 5, random 200× microscopic fields per filter. For 231 cells, the same procedure was utilized except that the top well contained DMEM medium supplemented with 1% FBS and the bottom well contained DMEM with 20% serum. For MCF10A cells, the top well contained maintenance medium with 2% horse serum and the bottom well contained maintenance medium with 5% horse serum. For invasion assays, the same experimental procedures as for transwell migration assays were used except that the 24-well inserts were coated with growth factor-reduced Matrigel (8-μm pore size, BD Biosciences).
Transfection of siRNA
To silence endogenous Gli1 expression in breast cancer cell lines, 50 nM of 2 different siRNAs (Stealth RNAi siRNA sets, Invitrogen) targeting Gli1 were transfected using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s protocols. Experiments were performed 48–72 h after transfection. To silence MMP-11 expression, 50 nM of a mixture of four different siRNAs (Smartpool siRNA, Dharmacon) targeting MMP-11 were transfected using Dharmafect transfection reagent (Dharmacon), as per the manufacturer’s instructions. Experiments were performed 72 h after transfection.
To measure cell growth, 5,000 cells per well were plated in a 96 well plate, and MTT assay was performed using CellTiter 96® Non-Radioactive Cell Proliferation Assay (Promega), according to the manufacturer’s protocols.
Immunostaining for Ki-67
Cells on coverslips were fixed in 70% ethanol, dehydrated in graded alcohols, permeabilized in acetone, rehydrated, and treated with 3% hydrogen peroxide followed by incubation with 3% goat serum. Cells were incubated with a rabbit polyclonal antibody to Ki-67 (dilution 1:50, Zymed/Invitrogen) for 1 h. Secondary detection was accomplished using a streptavidin–biotin system, as previously described . The percentage of Ki-67 labeled cells was determined after counting of cells in random, 200× microscopic fields. A minimum of 500 cells was counted.
Treatment with HPI-1
HPI-1 (Hedgehog Pathway Inhibitor-1) was purchased from TimTec, Inc. and dissolved in dimethyl sulfoxide (DMSO). Cells were treated with 5 and 10 μM HPI-1 in DMEM containing 0.5% FBS. The medium was changed every other day.
Experimental (tail vein) metastasis assay
231 cells were transduced to express shRNA targeting Gli1 (shGli1) or the NT control and maintained in selection medium for 10 days. Adherent, viable cells were washed and suspended in sterile Hank’s balanced salt solution. 2.5 × 105 cells were injected intravenously into the lateral tail veins of 3 to 4-week-old, female, athymic mice (nu/nu-Foxn1, Harlan Labs) to evaluate lung colonization. The five groups of animals consisted of parental, non-transduced cells and cells transduced with shGli1-1, shGli1-2, a combination of shGli1-1/2, and the NT control. After approximately 6 weeks, the mice were euthanized and the lungs were removed and fixed in neutral-buffered formalin. The number of surface metastases to the lungs was determined by examination under a dissecting microscope, and the diameter of each metastasis was estimated with an ocular micrometer. The volume of each metastasis was calculated using the formula (4/3)πr3 .
Comparison of migration, invasion, optical density, absorbance (MTT assay), Ki-67-labeling, and the numbers and volumes of metastases was accomplished using the unpaired t test with Welch’s correction or one-way ANOVA with Tukey post-test. Only P values less than 0.05 were regarded as statistically significant.
Over-expression of Gli1 promotes migration and invasion of MDA-MB-231 breast cancer cells
Reduction of Gli1 expression and activity decreases migration and invasion of ERα negative breast cancer cells
Gli1 is a major activator of Gli-mediated transcription, and it is also a direct target of Gli-mediated transcription in most cell types . Therefore, another approach to reducing Gli1 expression is to inhibit Gli-mediated transcription. The repressor form of Gli3 (Gli3R) is an inhibitor of Gli-mediated transcription. Gli3R is derived from full-length Gli3 after proteolytic cleavage and lacks amino acids C-terminal to the zinc finger domain . To reduce Gli-mediated transcription and Gli1 expression, Gli3R was expressed in 231 cells (231-Gli3R) by lentiviral transduction. Overexpression of Gli3 in comparison to the vector control (231-Vector) was confirmed by QRT (Fig. 2d). Correspondingly, expression of Gli1 was inhibited by approximately 50% in 231-Gli3R cells (Fig. 2e), and transwell migration and invasion of 231 cells was significantly inhibited by expression of Gli3R (P = 0.039 and P < 0.001, t test, respectively) (Fig. 2f, g). Expression of Gli3R for the time period required for the migration and invasion assays had no effect on cell growth (Supplemental Data, Fig. 2c) Therefore, by modulating expression of Gli1 via a variety of approaches, we have demonstrated that Gli1 promotes the migration and invasion of 231 cells.
Gli1 regulates the expression of MMP-11 in ERα negative breast cancer cells
Gli1-induced migration and invasion is mediated by MMP-11
To determine whether Gli1 enhances migration and invasion by upregulating MMP-11 in 231 cells, we silenced MMP-11 expression in 231-Gli1 cells using siRNA (Fig. 4d) resulting in a reduction in secretion of MMP-11 protein in 231-Gli1 cells to a level similar to that in 231 cells without over-expression of Gli1 (Fig. 4e, f). Transwell migration and invasion assays were performed for 24 h. Migration and invasion of 231-Gli1 cells with reduction of MMP-11 expression was decreased to the level exhibited by 231-Vector cells (P = 0.006 and P = 0.001, respectively, t test) (Fig. 4g, h). In addition, silencing MMP-11 in 231 cells which do not over-express Gli1 (231-vector) also resulted in a decrease in migration and invasion (P < 0.001 for both, t test) (Supplemental Data, Fig. 4). Our data provide evidence that MMP-11 is important for the migration and invasion of 231 cells, and its increase resulting from over-expression of Gli1 mediates the promotion of migration and invasion induced by Gli1.
Sustained inhibition of Gli1 expression and activity reduces growth of MDA-MB-231 breast cancer cells
To achieve long term inhibition of Gli1 by another mechanism, we treated 231 cells continuously for 8 days with the small molecular inhibitor of Gli-mediated transcription, HPI-1. HPI-1 was identified as inhibitor of Gli-mediated transcription by a high throughput screen in NIH3T3 cells . We confirmed the ability of HPI-1 to inhibit Gli-mediated transcription by demonstrating a dose dependent decrease in Gli1 expression after treatment with HPI-1 for 48 h (Fig. 5g). Then, we treated 231 cells with HPI-1 for 3 and 8 days and assessed cell growth by MTT assay. After 3 days of treatment, there was no decrease in cell growth (Fig. 5h), similar to the results seen with suppression of Gli1 expression by shRNA, siRNA and Gli3R. However, on day 8, cell growth was significantly reduced (P < 0.001, ANOVA) (Fig. 5h), similar to the effect of sustained suppression of Gli1 expression by shRNA (Fig. 5c).
Long-term silencing of Gli1 expression inhibits experimental metastasis of MDA-MB-231 breast cancer cells
Our data indicate that Gli1 is important for the growth, migration, invasion and metastasis of ERα negative breast cancer. Our findings are supported by the work of others also demonstrating that knockdown of Gli1 by RNA interference in two ERα negative breast cancer cell lines, MDA-MB-231 and SKBR3, reduced both cell growth and invasion . Silencing of Gli1 was also shown to inhibit the metastasis of the MDA-MB-435 cancer cell line, but whether this cell line represents ERα negative breast cancer or melanoma is currently debated [34, 35, 36]. Therefore, our data provide new evidence that Gli1 is important for the metastasis of ERα negative breast cancer. In addition, we report that Gli1 expression results in an increase in MMP-11, which mediates the pro-migratory and pro-invasive activities of Gli1 in the ERα negative MDA-MB-231 breast cancer cell line.
Metastasis is a multi-step process that requires cancer cells to detach from the main tumor, to migrate and invade through stroma and intravasate, to survive in the circulatory system and arrive at a secondary site, and to extravasate, invade and grow at the secondary site. In the tail-vein metastasis assay, only the later steps in this process are tested, specifically survival in the circulatory system, and extravasation, invasion and growth at the secondary site. Our in vitro data indicating that Gli1 is important for cell growth and survival, migration and invasion suggest that Gli1 promotes metastasis by advancing several processes—by promoting cell survival and growth in the lung and by promoting migration and invasion through the vascular wall and into the surrounding pulmonary parenchyma. The smaller sizes of the metastases with Gli1 silencing indicate the importance of Gli1 in promoting growth, while the smaller number of metastases may result from inhibition of the pro-survival effect of Gli1 or its role in promoting migration and invasion from the vasculature and into the lung parenchyma.
The cell lines included in this study, 231 and SUM1315, are two of the few breast cancer cell lines that are capable of metastasis in animal models. Expression of Gli1 mRNA is higher in these two cancer cells (Supplemental Fig. 1) than in MCF10A cells, which are derived from benign breast and are non-tumorigenic and non-metastatic. While Gli1 is relatively high in these metastatic lines, it is difficult to directly correlate the level of endogenous Gli1 in a cancer cell line and its relative migratory, invasive or metastatic capability because many different genes and complex molecular pathways contribute to their invasive and metastatic phenotypes. Gli1 is one of many molecules that regulate migration, invasion and metastasis. This is underscored by the many different genes that have been identified as being important in breast cancer metastasis, including those that direct metastases to the lung and brain [37, 38]. Of those 13 genes included in the gene expression signature associated with metastasis to the lung and those 11 genes associated with metastasis to the bone [37, 38], only one, CXCR4, has thus far been identified as a target of Gli1-mediated transcription. However, several of the many other genes that contribute to the metastasis of breast cancer are known to be up-regulated by Gli1. These include Osteopontin, Snail, Platelet-derived Growth Factor, and MMP-9 [9, 19, 39, 40, 41, 42, 43].
In this study, we demonstrate that MMP-11 is also up-regulated by Gli1 and that it contributes to the migration and invasion induced by Gli1 in 231 cells. Therefore, a reduction in MMP-11, and possibly other Gli1 targets, inhibits migration and invasion relatively rapidly. Unlike most MMP family members, MMP-11 does not cleave major components of the extracellular matrix . Several potential substrates of MMP-11 have been identified and include Collagen VI, Laminin Receptor, α1-Proteinase Inhibitor, and Insulin Like Growth Factor (IGF) Binding Protein 1 [30, 45, 46, 47]. MMP-11 was first identified from a cDNA library of breast carcinoma tissues and its expression is up-regulated in invasive breast cancers compared to ductal carcinoma in situ (DCIS) . MMP-11 is expressed at a high level in the fibroblastic stroma of breast cancers, but is also expressed in the cancer epithelial cells [27, 48, 49, 50]. Expression of MMP-11 in either the stromal or epithelial compartment in breast cancer is predictive of a poor disease outcome .
MMP-11 has been previously shown to contribute to the migration and adhesion of a hepatocellular carcinoma cell line, but the underlying molecular mechanism was not elucidated . A possible mechanism arises from the observation that over-expression of MMP-11 in MCF7 cells resulted in activation of Erk1/2 and Akt . Activated Akt regulates several proteins that are involved in cell migration and invasion, either by direct phosphorylation of these proteins or by modulation of their upstream regulators . The Erk/MAPK pathway regulates different processes involved in cell motility, including focal adhesion disassembly and the activity of the Rho family of small GTPases, which participate in cell migration and invasion . The mechanism through which MMP-11 activates Erk1/2 and Akt has not been specifically elucidated; however, it has been demonstrated that MMP-11 can release extracellular IGF-1 bound to IGF Binding Protein 1, which is a substrate of MMP-11 . IGF-1, by signaling through the IGF-1 receptor can activate both Erk1/2 and Akt [54, 55]. Stimulation of IGF-1 signaling by MMP-11 in MDA-MB-231 cells was shown to account for, at least in part, an increased tumor take and growth of xenografts of 231 cells overexpressing MMP-11 and to enhance the expression of molecules involved in cell migration and invasion .
Our present study also demonstrates that sustained inhibition of Gli1 for 8–10 days, rather than a short-term inhibition, is required to decrease the growth of 231 cells. This finding suggests that Gli1 regulates the production of one or more pro-survival signals, possibly secreted proteins, which require several days after down-regulation of Gli1 to become sufficiently diminished in quantity or activity that cell viability can no longer be maintained. Alternatively, prolonged culture may alter cellular function/homeostasis such that the cells become more susceptible to the effects of decreased Gli1 expression and activity.
Our data in the ERα negative and triple negative 231 and SUM1315 cell lines also demonstrate the potential for targeting Gli1 and the use of inhibitors of Gli-mediated transcription in the treatment of ERα negative and triple negative breast cancers. Triple negative breast cancers are particularly aggressive and currently lack effective targeted therapies . Therefore, there is a critical need for the development of new therapeutic strategies, such as targeting Gli1 via antagonists of Gli-mediated transcription, for the treatment of breast cancer. Further work in both in vitro and in vivo models of ERα negative and triple negative breast cancer will be required to establish the efficacy of targeting Gli1 using small molecule Gli1 antagonists.
We thank Drs. Janet Price, Stephen Ethier, Michael Ruppert and Bradley Yoder for their contributions of reagents for use in this work. This work was funded by the American Cancer Society (RSG-05-207-01-TBE), the Susan G. Komen Foundation (BCTR0707453), the National Institutes of Health (1R03CA130057), the Department of Defense (BC083907), and the National Foundation for Cancer Research.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.