The direct effect of Focal Adhesion Kinase (FAK), dominant-negative FAK, FAK-CD and FAK siRNA on gene expression and human MCF-7 breast cancer cell tumorigenesis

Background Focal adhesion kinase (FAK) is a non-receptor tyrosine kinase that plays an important role in survival signaling. FAK has been shown to be overexpressed in breast cancer tumors at early stages of tumorigenesis. Methods To study the direct effect of FAK on breast tumorigenesis, we developed Tet-ON (tetracycline-inducible) system of MCF-7 breast cancer cells stably transfected with FAK or dominant-negative, C-terminal domain of FAK (FAK-CD), and also FAKsiRNA with silenced FAK MCF-7 stable cell line. Increased expression of FAK in isogenic Tet-inducible MCF-7 cells caused increased cell growth, adhesion and soft agar colony formation in vitro, while expression of dominant-negative FAK inhibitor caused inhibition of these cellular processes. To study the role of induced FAK and FAK-CD in vivo, we inoculated these Tet-inducible cells in nude mice to generate tumors in the presence or absence of doxycycline in the drinking water. FAKsiRNA-MCF-7 cells were also injected into nude mice to generate xenograft tumors. Results Induction of FAK resulted in significant increased tumorigenesis, while induced FAK-CD resulted in decreased tumorigenesis. Taq Man Low Density Array assay demonstrated specific induction of FAKmRNA in MCF-7-Tet-ON-FAK cells. DMP1, encoding cyclin D binding myb-like protein 1 was one of the genes specifically affected by Tet-inducible FAK or FAK-CD in breast xenograft tumors. In addition, silencing of FAK in MCF-7 cells with FAK siRNA caused increased cell rounding, decreased cell viability in vitro and inhibited tumorigenesis in vivo. Importantly, Affymetrix microarray gene profiling analysis using Human Genome U133A GeneChips revealed >4300 genes, known to be involved in apoptosis, cell cycle, and adhesion that were significantly down- or up-regulated (p < 0.05) by FAKsiRNA. Conclusion Thus, these data for the first time demonstrate the direct effect of FAK expression and function on MCF-7 breast cancer tumorigenesis in vivo and reveal specific expression of genes affected by silencing of FAK.


Background
Focal adhesion kinase (FAK) is a 125 kDa non-receptor tyrosine kinase localized at the focal adhesions [1], which are the contact points between cells and extracellular matrix and are the sites of intense tyrosine phosphorylation [2]. FAK is tyrosine phosphorylated in response to a number of stimuli, including clustering of integrins [3], plating on fibronectin or collagen [4,5], and in response to a number of mitogenic agents [6]. FAK is involved in regulation of different cellular processes, such as cell spreading, adhesion, motility, proliferation, and survival [7]. Although several studies supported that FAK plays a role in breast carcinogenesis [8][9][10][11], the direct and specific role of FAK up and down-regulation on breast cancer tumorigenesis in vivo and genes expression profiling effected by FAK silencing are not understood.
FAK was originally identified as a major tyrosine phosphorylated protein in cells transformed by v-Src and associated with c-Src [12,13]. FAK is overexpressed in invasive and metastatic tumors [14], and the FAK gene is also amplified in many types of tumors [15], suggesting a role for FAK in adhesion or survival in tumor cells. In cancer cells, attenuation of FAK expression induces detachment and apoptosis [16], suggesting that a FAK-dependent signal is required for tumor cell growth. Furthermore, an activated form of FAK leads to resistance to anoikis [17], and FAK degradation is associated with apoptosis [18,19]. The C-terminal domain of FAK, called FAK-CD is analogous to murine FAK-related non-kinase (FRNK) [16], and has been shown to cause increased cell rounding, detachment, and apoptosis when transduced into breast and colon cancer cells [20][21][22].
Immunohistochemical analysis of FAK expression demonstrated up-regulation of FAK in 88% of invasive and metastatic breast tumors [23]. The up-regulation of FAK occurred at early stages of breast carcinogenesis [24], as FAK overexpression was detected ductal carcinoma in situ (DCIS) that precedes tumor cell invasion and metastasis [25]. FAK overexpression highly correlated with microvessel density, metastasis, and angiogenesis [26]. However, the studies on the role of FAK in breast tumorigenesis in vivo have been mostly limited to immunohistochemical studies of tumor biopsies. The recent study using Cre/loxP recombination system to disrupt FAK function in the mammary epithelium demonstrated that FAK is required for the transition of premalignant hyperplasia to carcinomas and their subsequent metastasis [27].
To determine the direct role of FAK in breast tumors in vivo, we created stable clones of human breast cancer cells overexpressing FAK or dominant-negative FAK-CD using the Tet-ON system and studied these cells in a nude xenograft model. In addition, we employed RT-PCR Low-density Array and Affymetrix analyses to reveal genes directly affected by FAK up or down-regulation. To the best of our knowledge, this is the first study of gene expression profiles affected by FAK regulation in MCF-7 breast cancer cell model.

Construction of FAK and FAK-CD-TRE-2 plasmids
The HA-tagged FAKcDNA fragment that was subcloned from pcDNA3 plasmid into BamHI and EcoRV sites of TRE-2-hyg plasmid. FAK-CD, FAK-C-terminal domain of FAK (677-1052 amino acids) was obtained by PCR. The PCR fragment was subcloned into NheI and EcoRV sites of TRE-2 plasmid. The FAK and FAK-TRE-2 plasmid sequences were confirmed by sequencing in the ICBR facility (UF, Shands Cancer Center).

Generation of stable human doxycycline-inducible breast cancer MCF-7 cells, overexpressing FAK or dominantnegative FAK-CD
The first step was to create stably transformed MCF-7 cells by transfecting with neomycin-resistant pTet-ON regulator plasmid, encoding rtTA protein (reverse tTA, tetracycline-controlled transactivator). The stably transformed MCF-Tet-ON clones were selected by cultivation in EMEM, containing 500 μg/ml Geneticin, G418. The Tet-ON MCF-7 cells were selected and used for transfection of FAK and FAK-CD-TRE-2-hyg plasmids with Lipofectamin 2000 (Invitrogen Inc) according to the manufacturer's protocol. The transfected cells with TRE-2-hyg, FAK-TRE-2hyg and FAK-CD-TRE-2-hyg plasmids were maintained in a medium with G418 (0.2 mg/ml) and hygromycin (0.1 mg/ml). The dose-and time-dependent experiments on stably transfected Tet-ON cells showed a maximal induction of FAK and FAK-CD at 2 μg/ml doxycycline at 4 days of cell cultivation. The pooled population of cells with maximal induction of FAK and FAK-CD by 2 μg/ml doxycycline used for the study.

Generation of MCF-7 cells stably expressing FAKsiRNA
The pRNAT-H1.4-hyg plasmid was kindly provided by Dr. K. Brown (University of Florida, Gainesville). For generation of siRNA construct, the GenScript software was used. The primers for generating the hairpin construct containing FAKsiRNA were the following (with siRNA sequences shown in bold): FAKsiRNA#1 oligonucleotides were 5'-CGCGTCG-TAATACTCGCTCCATTGCACCTT GATATCCGGGTGCA ATGGAGCGAGTATTATTTTTTCCAAC-3' and complementary oligonucleotide. For FAKsiRNA#2, oligonucleotides were the following: 5'-CGCGTCGTA AATGCC TTGATGTACATCTTTGATATCCGAGATGTAC ATCAAGG CATTTATTTTT TCCAAC-3' and the complementary oligonucleotide. For control siRNA, scrambled FAK siRNA or firefly luciferase oligonucleotides were used. Oligonucleotides for control siRNA (scrambled FAK siRNA#1 sequence) were generated with GenScript software and oligonucleotides were the following: 5'-CGCGTCGC CACTTACGATAGCTCCATTCT TG ATATCCGGAAT-GGAGCTATCGTAAGTGGTTTTTTCCAAC-3' and the complementary oligonucleotide. Firefly luciferase siRNA was also used as a control, and the oligonucleotides were 5'-CGCGTCGTCGAAGTACTCAGCGTAAGTTGATAT CCGC TTACGCTGAGTACTT CGATTTTTTCCAAC-3' and the complementary oligonucleotide. The oligonucleotide duplexes were subcloned into MluI and XhoI sites of pRNAT-H1.4-hyg plasmid. The resulting siRNA DNAs were used for transfecting cells with Lipofectamin 2000 (Invitrogen Inc) according to the manufacturer's protocol. The MCF-7 cells were transfected with control vector, control siRNA, FAKsiRNA#1, and FAKsiRNA#2 and after several weeks growing in the presence of hygromycin (100 μg/ml), stable clones were collected and analyzed by Western immunoblotting with FAK antibodies. The stable cell line with highest reduction of FAK expression was used for the study.

Western Blotting
Western Blotting was performed as described before [28].

Immunohistochemistry staining of xenograft tumors
Tumors from untreated and treated with doxycyclinetreated mice were removed and fixed in 4% formaldehyde solution immediately after surgical resection. The fixed samples were analyzed in the Immunohistochemistry Core Facility (UF, Department of Pathology). FAK staining was performed, as described previously with FAK 4.47 antibodies [29].
Cell growth in vitro 2 × 10 5 cells were plated on 6-well plates. Cell growth was determined by counting cells on hemocytometers. Trypan blue exclusion assay was used for detecting viable cells.

RNA isolation
Total cellular RNA was isolated from cultured cells with a NucleoSpin RNA II Purification Kit (Clontech Laboratories, Inc.) according to the manufacturer's protocol.

Taq Man Low Density Array Real time PCR assay
Customized Taq Man Low density arrays with 44 different genes (Table 1) and a GAPDH probe as a normalization control were obtained from Applied Biosystems. The isolated RNA was used for PCR reaction as described in the manufacturer's protocol. The ABI PRISM 7700 cycler's software calculated a threshold cycle number (Ct) at which each PCR amplification reached a significant threshold level. The relative quantity, RQ, was calculated and statistical analysis was performed with Student's ttest.

Microarray and statistical data analyses
RNA was isolated from MCF-7, MCF-7-Vector, MCF-7-Control luciferase siRNA, FAKsiRNA#1, and FAKsiRNA#2 samples using Clontech Labs, Inc. Kit. The cDNA preparation, probe labeling, hybridization and image analysis of the arrays were carried at ICBR Core Facility (UF) according to the manufacturer's recommendations (Affymetrix, Santa Clara, CA, USA), using Affymetrix Human Genome U133A GeneChip containing 47000 transcripts. For the data analysis, we treated siRNA#1, siRNA#2 as siRNA group, and MCF7, MCF-7-Vector and MCF-7-Control siRNA as control group. We used with 4 replicates per group in the analysis. Affymetrix Microarray software was used to analyze the data.
Statistical tests were performed using BioConductor statistical software http://www.bioconductor.org/ [30]. Data pre-processing and normalization were performed using the Affy package [31]. Raw data were normalized by Robust Multichip Analysis (RMA) approach. The local pooled error (LPE) method [32] was applied to detect the genes which are significant differentially expressed between FAK siRNA and control (MCF-7-Vector and Control siRNA) samples, and the resampling technique was used to control the false-discovery rate (FDR). In combination with a resampling FDR correction, the LPE method was shown to outperform other 2-sample comparison methods [32]. We used local-pooled-error (LPE) approach to evaluate the level of significance for each gene's differential expression. The LPE estimation is based on pooling errors within genes and between replicate arrays for genes in which expression values are similar.
The p-values from LPE were used as first criteria to define the significant gene set. The complete data are uploaded to NCBI website, Accession number GSE11581. Differentially expressed genes were ranked by the p-values, and the genes with p < 0.05 were considered as differentially expressed genes at a statistically significant level. For those significant genes, different level of fold-change were used to select the significant differentially expressed genes. Cluster analysis was performed on these selected differentially expressed genes using hierarchical clustering with the complete linkage method on a similarity matrix built with Pearson correlation coefficient. The results of the cluster analysis were displayed at heat map. The data are deposited in NCBI database (NCBI Accession number GSE11581).

Soft agar growth assay
Cells cultivated for 6 days with or without doxycycline (2 μg/ml) were plated in 0.3% agar (with or without doxycycline) on the plates, containing 0.5% agar (with or without doxycycline). The plates were cultivated on the plates for 2-3 weeks. The colonies were counted on stained with crystal violet plates under the microscope. Samples were assayed in duplicates.
Adhesion Assay 96-well plates coated with collagen were blocked in 0.5% BSA in the EMEM medium. Then 2 × 10 4 of cells were plated on collagen-treated plates and incubated at 37°C for 37 minutes in CO 2 incubator. Cells were washed with PBS and fixed with 3.7% formaldehyde for 10 minutes. After washing with PBS, cells were stained with crystal violet (5 mg/ml in 2% ethanol). Then 2% SDS was added and plate was read at 590 nm to detect adhesion of cells.

Tumor growth in nude mice
Female nude mice (4-5 weeks old) were ordered from Harlan Laboratories Inc. All experiments were performed according to guidelines of the approved IACUC protocol. To supplement the estrogens for MCF-7 proliferation each nude mouse was implanted with a 1.5 mg of 17β-estradiol pellet (Innovative Research of America, Sarasota, FL, USA). A week after the pellet implantation, 5 × 10 6 of MCF-7-Tet-ON cells stably expressing FAK-TRE-2 or FAK-CD were subcutaneously injected into the mice. Mice were divided into two groups: the first group did not receive doxycycline in the drinking water, and the second group received doxycycline (2 mg/ml) in the water. For FAKsiRNA experiments, 5 × 10 6 of MCF-7 cells were injected subcutaneously into mice with implanted 1.5 mg of 17β-estradiol pellet and tumor growth was observed. The tumors were measured using calipers and the tumor volume was calculated using the formula volume, V = L × W 2 /2, where Llong diameter and W-short diameter.

Statistical analyses
Student's t test was performed to determine significance. The difference between data with p < 0.05 was considered significant.

Doxycycline induces expression of FAK and FAK-CD in MCF-7-Tet-ON breast cancer cells
To understand the biologic role of FAK in breast cancer cells, stable clones overexpressing FAK with a HA-tag were created from MCF-7 breast cancer cells using the Tet-ON system. The MCF-7-Tet-ON cells stably expressing FAK or FAK-CD were cultivated in the presence of doxycycline and Western blotting was performed with FAK antibodies ( Figure 1). Expression of FAK ( Figure 1A) and FAK-CD ( Figure 1B) was induced by doxycycline compared to the cells without doxycycline. Western blotting analysis with HA-tag antibody confirmed expression of HA-tagged FAK ( Figure 1A). Doxycycline-induced FAK was activated, resulting in increased Y397 autophosphorylated FAK and increased phosphorylation of FAK substrate, Y118-paxillin ( Figure 1A).

A B C D
To determine whether exogenous HA-tagged FAK localized to focal adhesions, immunostaining with HA-tag antibody was performed. Doxycycline-inducible FAK mainly localized to the cytoplasm, perinuclear area and the focal adhesions in MCF-7-Tet-ON-FAK cells (Fig. 1C, lower panels), while control MCF-7-Tet-ON-TRE-2 cells were negative by immunostaining with HA-tag antibody ( Figure 1C, upper panels). Induction of FAK with doxycycline did not change stress fiber formation, compared to the cells without doxycycline, as detected by staining actin with phalloidin ( Figure 1C).
To detect the effect of exogenous FAK-CD on FAK localization in cells with doxycycline-inducible FAK-CD overexpression, we stained FAK with the N-terminal domain antibody ( Figure 1D). Doxycycline caused displacement of FAK from the focal adhesion and induced cell rounding ( Figure 1D). To quantify the cell rounding, we counted the percent of rounded cells caused by doxycycline-inducible FAK-CD ( Figure 2A). Rounding in the presence of doxycycline was induced more than 5 times in the MCF-7 cells overexpressing FAK-CD. In contrast, there was no cell rounding caused by overexpression of FAK compared to the TRE-2 stable control cells ( Figure 2A). Thus, doxycycline-induced FAK-CD caused changes in cell morphology resulting in cell rounding.

Doxycycline-inducible FAK increases MCF-7 cell growth, adhesion and soft agar colony formation, while FAK-CD decreases these cellular processes
MCF-7 cells expressing TRE-2, FAK-TRE-2 or FAK-CD-TRE-2 were grown without and with doxycycline for 8 days, and viable cells were counted ( Figure 2B). Control MCF-7-Tet-ON-TRE-2 cells that did not significantly change cell growth in vitro ( Figure 2B). In contrast, cells with doxycycline-inducible FAK expressed significantly (p < 0.05) increased cell growth at 6-8 days, while cells with induced FAK-CD had significantly (p < 0.05) decreased cell growth ( Figure 2B). To detect the effect of overexpressed FAK and FAK-CD on cell adhesion, we plated MCF-7 cells on collagen and performed adhesion assay ( Figure 2C). Doxycycline-inducible FAK resulted in induction of cell adhesion compared to the cells without doxycycline, while FAK-CD induction resulted in significantly decreased cell adhesion ( Figure 2C).
To detect whether FAK and FAK-CD affected anchorageindependent growth, cells were treated with doxycycline (2 μg/ml) for 6 days, resuspended in 0.3% agar, with or without doxycycline, and cells were cultivated for 3 weeks on soft agar, and colonies were counted. Doxycyclineinduced FAK resulted in an increased number of colonies compared to cells without doxycycline, while FAK-CD resulted in decreased colony formation ( Figure 2D).
Thus, overexpression of FAK increases breast cancer cell growth, adhesion and soft agar colony formation in vitro, while dominant-negative FAK, FAK-CD decreases these cellular processes.

Doxycycline-inducible FAK induces tumor growth, while doxycycline-inducible FAK-CD reduces the tumor growth in nude mice in vivo
To determine role of FAK and FAK-CD in breast tumorigenesis, we injected MCF-7-Tet-ON-FAK, FAK-CD and TRE-2-vector stable cell lines into two groups of nude mice given drinking water either with or without doxycycline and observed tumor growth. Doxycycline-induced FAK significantly increased tumor volume, more than 2.9 times that of control TRE-2 mice ( Figure 3A

TaqMan Low Density Array analysis in Tet-inducible MCF-7-Tet-ON-FAK cell lines and tumors demonstrates specific overexpression of FAK
To determine the molecular mechanism of FAK and FAK-CD induction, we performed TaqMan low-density array analysis in Tet-inducible MCF-7 cell lines. We compared the expression of genes inside each isogenic cell line grown with and without doxycycline. We compared expression of 44 genes relative to GAPDH control in all cell lines (Additional file 1). The expression levels of expression of 39 genes with detectable mRNA levels that were differentially affected by Tet-inducible FAK or FAK-CD-induction are shown in Table 1 Figure  3B) and by immunohistochemical staining ( Figure 3D).
Among the genes, specifically affected by FAK or FAK-CD overexpression in the presence of doxycyclin, we found that the DMP1 mRNA expression was > 1.

FAK siRNA causes cell rounding, decreases cell growth in vitro and inhibits tumorigenesis in vivo
To test the effect of decreased FAK expression and compare with the FAK-CD-induced expression, we generated stable MCF-7 cell lines, expressing vector alone, Control siRNA, FAKsiRNA#1 and FAKsiRNA#2. All stably transfected cells were GFP-positive showing effective expression of protein, since vector had GFP-tag. Western blotting with FAK antibody demonstrates that both FAK-siRNA #1 and #2 cells efficiently decreased expression of FAK in MCF-7 cells, while MCF-7, MCF-7-Vector or MCF-7-Control siRNA did not decrease FAK expression ( Figure  4A). FAKsiRNA caused change in cell morphology and significantly increased cell rounding compared with MCF-Vector or Control siRNA ( Figure 4B). In addition, FAK-siRNA significantly decreased cell growth in vitro com-pared with MCF-7, Vector and Control siRNA cells ( Figure  4C). Thus, FAK siRNA caused cell rounding and decreased cell growth in vitro.
To test the effect of down-regulated FAK on tumorigenesis, we injected MCF-7, MCF-7-Vector, MCF-7-Control siRNA, MCF-7-FAKsiRNA and FAKsiRNA#2 cells in the nude mice and observed tumor growth. Both FAKsiRNA #1 and #2 decreased tumor cell growth compared to MCF-7 and MCF-7-Vector and MCF-7-Control siRNA cells (Figure 5A, left upper panel). The significantly decreased tumor size in case of FAK siRNA#1 and #2 is shown on Figure 5A, right panels. Both FAK siRNA #1 and #2 significantly decreased tumor volume versus control MCF-7, MCF-7-Vector and MCF-7-Control siRNA cells ( Figure 5A, left lower panel). We performed analysis of FAK expression in these tumor samples by Western blotting ( Figure  5B). Tumors from both FAKsiRNA#1 and FAKsiRNA#2 had significantly decreased FAK expression compared with tumors from MCF-7, MCF-7-Vector and MCF-7-Control siRNA group ( Figure 5B). Thus, down-regulation of FAK by FAK siRNA significantly decreased tumor growth in vivo compared to MCF-7 and MCF-7-Vector and MCF-7-Control siRNA cell lines.

Affymetrix Microarray and TaqMan Low-Density RT-PCR array assays reveal more than 4300 genes down-and upregulated in MCF-7-FAKsiRNA cell lines
In addition, to better understand molecular mechanism of FAK down-regulation by FAKsiRNA, we performed gene expression of MCF-7, MCF-7-Vector, MCF-7-Control siRNA, FAKsiRNA#1 and FAKsiRNA#2 cell line samples by microarray analysis using Affymetrix Human Genome U133 GeneChip Array, covering the whole human genome with >47000 transcripts. Among all genes on the Chips, we revealed over than 4300 genes (all gene profiles, affected by FAKsiRNA are downloaded to NCBI databases with [Accession number GSE11581]) were significantly with p < 0.05 either up-or down-regulated in FAKsiRNA#1 and FAKsiRNA#2 cell lines versus control group (MCF-7, MCF-7-Vector, MCF-7-Control siRNA cell lines). The heatmap of 4300 genes that were significantly down-regulated and up-regulated (p < 0.05) by FAKsiRNA in MCF-7 cells is shown on Figure 5C. 169 genes were more than 2-fold down or up-regulated (p < 0.05), 32 genes were more than 3-fold down-or up-regulated (p < 0.05), and 12 genes were > 4-fold (p < 0.05) affected by FAKsiRNA. We found that expression of FAK was 1.5-fold down-regulated (p < 0.05) in the pooled FAKsiRNA#1 and FAKsiRNA#2 group compared to control group (Table 2). TaqMan Real-time PCR array assay demonstrated the same 1.5 fold decrease of FAKmRNA by both FAKsiRNA compared to control MCF-7 and MCF-7-Vector cells, confirming and validating microarray data.
Importantly, Pyk2 and several other genes were not significantly different by both microarray and Real-time-PCR, additionally validating both methods.

Discussion
FAK controls survival and is activated at the early stages of breast carcinogenesis [25]. FAK expression was demonstrated in ductal carcinoma in situ (DCIS) tumors. The study suggested that FAK overexpression occurred in preinvasive, DCIS tumors preceding tumor metastasis. In addition, we have shown that FAK formed a protein complex with vascular endothelial receptor-3 protein, VEGFR-3 in breast cancer cells [33] suggesting its critical role in breast lymphogenesis and angiogenesis.
Thus, in the present study we performed analysis of induced FAK expression and induced dominant-negative FAK, FAK-CD in breast carcinogenesis using MCF-7-Tet-ON model. We generated MCF-7-tet-inducible cells, stably transfected with Tet-inducible FAK or Tet-inducible FAK-CD plasmids and performed analysis of this induced expression on the cellular processes in vitro and tumorigenesis in vivo. Increased expression of FAK increased cell growth, adhesion and soft agar colony formation in vitro.
In contrast, induced expression of FAK-CD decreased these cellular processes. We inoculated these cells in nude mice and demonstrated increased tumorigenesis in the case of induced FAK and reverse processes in the presence of induced FAK-CD. Expression of FAK-CD caused cell rounding, which is explained by exogenous localization of FAK-CD in the focal adhesion and displacement of FAK from the focal adhesion sites [20]. The data are consistent with our previous report, where we have shown that exogenous FAK-CD can inhibit FAK functions and cause cell rounding and apoptosis of BT474 breast cancer cells [34]. FAK-CD or FRNK has been shown to decreased cell motility of AU-565 breast cancer cells [35]. In this report, we for the first time analyzed the direct effect of FAK and FAK-CD Tet-inducible expression on gene expression and cellular processes in MCF-7 line and in breast xenograft models. The increased tumorigenesis was accompanied by increased FAK mRNA and protein levels. Real-time PCR analysis demonstrated specific increased FAK mRNA in MCF-7-Tet-ON-FAK cells. Importantly, the expression of homologous Pyk-2 was not increased and was even decreased in MCF-7-Tet-ON-FAK cells indicating FAKindependent regulation of Pyk-2 in MCF-7 cells.
Although, the recent report demonstrated increased expression of Pyk-2 and FAK in tissues from early and advanced breast cancers suggesting importance of Pyk-2 pathway in breast tumorigenesis [10], the down-stream signaling mediated by FAK and Pyk-2 kinases is different. The functional differences between Pyk-2 and FAK kinases are supported by the recent report on the structural differences between C-terminal FAT domains of FAK and Pyk-2 and differences in association and phosphorylation of focal adhesion protein, paxillin [36]. In this study we show that silencing of FAK with two different FAKsiRNA in MCF-7 stable cell line resulted in decreased breast tumorigenesis in vivo and decreased FAK expression in the tumor samples. We revealed significant differences in gene expression affected by FAK silencing or FAK up-regulation in MCF-7 cells. Thus, FAK is critical for breast cell survival and tumorigenesis. The models can be used for targeted therapy and for studies of FAK inhibitors.
Intriguingly, MCF-7-Tet-ON-FAK cell line and tumors grown in the presence of doxycycline had decreased DMP1 (cyclin D binding protein 1) mRNA and protein levels. DMP1 is a transcription factor which binds to cyclin D and when overexpressed induces cell cycle arrest [37]. DMP1 can bind Arf1 (p14 Arf known as an inhibitor of Mdm-2 and stabilizer of p53) promoter and activate its transcription, thus regulating Arf-p53 pathway. It is known that loss of DMP1 caused spontaneous tumorigenesis in mice and death by 24 months of age from different forms of cancer [38]. Dmp1 -/mice among phenotypic abnormalities had also poor mammary development [37]. The Dmp1 + /tumors often retain wild type allele of DMP1, thus DMP1 is haplo-insufficient for tumor suppression [37]. Overexpression of cyclin D1, which is found to be overexpressed in 60-80% of breast cancer tumors, inhibits the transcriptional activity of DMP1 and antagonizes its function [37]. It was shown that overexpression of FAK increased expression of cyclin D1 [39], which contributed to increased expression of cellular proliferation. The function of human DMP1 protein (the transcription factor that is involved in the oncogenetumor suppressor signaling) is an unexplored area in human cancer, and it remains to discover its post-translational modifications and identification of DMP1-protein binding partners [37]. Thus, FAK-DMP1-cyclin D1 linked pathway can be a novel mechanism regulating intracellular functions and carcinogenesis, and decreased DMP1  [20,28], Tet-inducible MCF-7-Tet-ON-FAK-CD cells showed decreased viability and growth on soft agar. Interestingly, TaqMan analysis demonstrated significantly decreased AKT level in MCF-7-Tet-ON-FAK-CD tumors. These data are consistent with our previous report, when adenoviral FAK-CD decreased AKT in breast cancer cell lines [28]. We have shown that AKT increased survival of the breast cancer cell line [28]. Thus, down-regulation of AKT by doxycycline-induced FAK-CD can explain decreased tumorigenesis in these tumors.
We demonstrated that silencing of FAK with two FAK siRNA decreased tumorigenesis in MCF-7 xenograft model that was accompanied by decreased FAK expression in tumor samples. Importantly, we performed Affymetrix chip microarray analysis and for the first time demonstrated more than 4300 genes significantly up-and down-regulated, and >160 genes that are >2-fold down or up-regulated (p < 0.05) in FAKsiRNA#1 and #2 samples group compared to control group that provides basis for future mechanistic detail study of FAKsiRNA-directed gene expression in breast cancer cells. The most important that both microarray data and TaqMan Real-time PCR array data demonstrate significantly decreased expression of FAK in FAKsiRNA cell lines compared to the control group, and these data together with decreased tumorigenesis in vivo support the critical role of FAK signaling in breast tumorigenesis.
The data on 169 genes that were 2-fold significantly up or down-regulated by FAKsiRNA in MCF-7 breast cancer model provide a basis for detail mechanistic study of FAK down-stream signaling during breast tumorigenesis. Several genes that were up-regulated by FAKsiRNA, like MAP2K5, mitogen-activated protein are connected with the FAK signaling pathway. We have shown that FAK upregulated ERK1/2 in the stress conditions [28] in breast cancer cells. MAP2K5 (MEK5) kinase has been shown to be correlated with metastasis in prostate tumors [40] and has been involved in breast carcinogenesis [41]. Other important genes down-regulated by both FAKsiRNA#1 and #2 was PEG10, paternally expressed 10 gene that is directly involved in tumorigenesis [42]. PEG10 is a c-Myc target gene in cancer cells and has been shown to be activated during breast tumorigenesis, expressed in 55% of ductal carcinoma in situ and 32% of invasive ductal carcinoma [43]. Thus, down-regulation of PEG10 by FAK-siRNA is consistent with decreased tumorigenesis in these cells. Another important gene that was down-regulated by FAKsiRNA is PSAT1, phosphoserine aminotransferase. Recently, it has been shown that PSAT1 overexpression stimutated cell growth of colon cancer cells [44]. High PSAT1 mRNA levels were associated in breast cancers with poor clinical response to endocrine therapy [45]. Thus, PSAT1 functions as pro-survival and proliferative factor in tumorigenesis.
Among up-regulated genes (>3-fold) by FAKsiRNA was thiredoxin interacting protein (TXNIP) that is known to be involved in apoptosis [46]. In pancreatic Miapaca-2 cells, overexpression of TXNIP resulted in increased basal apoptosis and increased sensitivity to cisplatin and oxaliplatin [46]. We also performed Affymetrix array analysis on Tet-ON-FAK MCF-7 cells and found that thiredoxin interacting protein (TXNIP) was 2.7 fold decreased in the presence of doxycyline, supporting that TXNIP is regulated by FAK and plays role in breast tumorigenesis. We revealed several genes that were inversely regulated in MCF-Tet-ON-FAK with induced FAK and in MCF-7 cells with silenced RNA, 16 genes were >1.5 fold in inversely regulated, p < 0.05 (not shown), providing a basis for the mechanism of their regulation. Another gene that was ~5 fold up-regulated by FAKsiRNA is transcription factor AP-2 alpha (TFAP2A). Activation and expression of AP-2 was associated with increased apoptosis and inhibiton of cell growth [47]. In addition, immunohistochemical staining studies showed that loss of transcription factor AP-2 correlated with disease progression from normal breast to invasive breast cancer disease [48]. Another group demonstrated that reduced expression of AP-2 transcription factor associated with aggressive breast cancer [49]. Down-regulation of AP-2 with siRNA led to enhanced breast cancer tumor growth and reduced chemotherapyinduced cell death [50]. Thus, these few examples of down-regulated and up-regulated genes can explain reduced tumorigenesis in FAKsiRNA MCF-7 model and suggest that FAKsiRNA can be used as a therapy approach.
The presented Tet-regulated FAK-CD, dominant negative FAK, breast cancer cell model can be compared with Tetinducible dominant-negative c-Src model [51]. Similar to dominant-negative c-Src-induced model of Tet-ON MCF-7 cells, cells with induced FAK-CD had decreased cell adhesion and viability and reduced tumorigenesis, consistent with our data on cooperative survival signaling of FAK and Src in colon cancer cells [22].

Conclusion
In conclusion, we have shown for the first time that Tetinducible FAK increased cell growth, adhesion and soft agar colony formation in vitro, while it increased breast tumorigenesis in vivo. In contrast, Tet-inducible FAK-CD and FAKsiRNA blocked these processes in vivo. TaqMan Low Density Array identified several specific genes affected by modulation of FAK expression and function, such as DMP1 and AKT1. In addition, Affymetrix analysis revealed more than 4300 genes that were affected by FAK-siRNA in MCF-7 xenograft tumors. This system demonstrates the basis for future studies on the role of FAK and downstream signaling in breast tumorigenesis and the biology of breast cancer cells. Thus, FAK is a promising target for future breast cancer therapy.