The human breast cancer cell lines Hs578T, MDA-MB231 and T47D were obtained from American Type Culture Collection ATCC, Manassas, VA). Cell culture was carried out as recommended by ATCC. Hs578T and MDA-MB231 cells were grown in DMEM supplemented with 10% calf bovine serum. T47D cells were grown in RPMI-1640 medium supplemented with 10% calf bovine serum and 0.2 units/mL insulin. The human mammary epithelial cell line AG1132B was obtained from Coriell Institute for Medical Research (Camden, NJ). Cell culture was carried out as recommended by the supplier.

The MDA-MB231 and T47D cell lines were transfected with siRNA-RASSF1C and control plasmids as previously described [20]. Since the shRNA plasmids used in this study would target both RASSF1A and RASSF1C mRNAs, we utilized breast cancer cells that express RASSF1C but not RASSF1A. Cells were plated at 20,000 and 50,000 cells per well in the appropriate medium with 10% calf serum in 24 and 6 well culture dishes, respectively. After 24 hr, the cells were transfected with 1 μg/ml plasmid DNA using Lipofectamine (Invitrogen, Carlsbad, CA) using recommended conditions. 48 hr post-transfection, cells were collected and were used for RNA extraction. 200 ng of total RNA was used to prepare cDNA with the Omniscript kit (Qiagen, Valencia, CA). 100 ng of the reverse transcriptase reaction was used for PCR using the HotStart master mix (Qiagen, Valencia, CA) and PCR reactions were run with the following conditions: 95°C 1 min, 60°C for 30 sec, and 72°C for 30 sec for 35 cycles.

Cells were treated with either siRNA-RASSF1C (suppresses RASSF1C expression) or control plasmid for 48 hr, and cell proliferation was measured by the alamar blue assay as previously described [18,19].

MDA-MB231, T47D, and AG1132B cells were treated with either siRNA-RASSF1C (suppresses RASSF1C expression) or control plasmid for 18 hr. 3H-thymidine was then added and cells were labeled for 6 hr before cultures were terminated and 3H-thymidine incorporation was assayed as previously described [19].

In order to over-express RASSF1C cDNA in human breast cancer cells in a regulated fashion, we chose to use a doxycycline (dox)-inducible Murine Leukemia Virus based retroviral vector that was developed in house (Chen et. al., manuscript in preparation). Using the YFP-RASSF1C plasmid [18,19] as a template, the full-length HA-RASSF1C coding sequence was cloned using the forward primer: 5' GGG TCG ACG CGG CCG CC ATGTACCCATACGATGTTCCGGATTACGCT GGC GAG GCTGAAACACCTT 3' and the reverse primer: 5' CGG GAT CC TCA CCC AAG GGG GCA GGC GTG CAG 3', and the HA-IGFBP-5 coding sequence was cloned using the forward primer: 5' GGG TCG ACG CGG CCG CC ATGTACCCATACGATGTTCCGGATTACGCT GGC TCC TTC GTG CAC TGC GA 3', and the reverse primer: 5' CGG GAT CC TCA CTC AAC GTT GCT GCT GTC GAA 3' flanked by NotI and BamHI restriction enzyme sites, respectively. The NotI and BamHI sites in the pGYT plasmid were used to insert the RASSF1C cDNA sequence down stream of the TetO and mammalian promoter. The correct cDNA sequence and orientation were confirmed by sequencing the pGYT-HA-RASSF1C plasmid. This plasmid then was used to produce VSV-G pseudotyped MLV-based vector as described [20].

MDA-MB231 and T47D breast cancer cell lines were seeded at 1 × 10⁵ cells/well in 6-well plates. After 24 hr of incubation, the cells were transduced with MLV-based vectors rtTA-GYT (vector without transgene), rtTA-GYT-GFP, and rtTA-GYT-HA-RASSF1C with different MOI in 6-well plates, using 2 or 3 serial infection cycles as described [20]. After 1-4 days, cells were treated with up to 1 × 10^-6 M doxycycline (dox) for 48 hr. HA-RASSF1C expression was assessed by Western blot analysis using anti-HA antibody. We tested the MLV-GFP vector expression in human breast cancer cell lines and demonstrated that a 10 fold induction of GFP expression can be achieved with a dox concentration of 1 ug/ml.

Total RNA from human cell lines was isolated from confluent cultures using the Absolutely RNA Microprep Kit (Stratagene, La Jolla, CA). 1 μg of total RNA was used to set up reverse transcriptase (RT) reactions using the superscript kit (Qiagen, Valencia, CA) and the RT reactions were subsequently used to set up real-time PCR reactions using 1 μl of RT as a template. 1 μg of total RNA was used to perform reverse transcription reactions (RTs) and 1 ul of the RT reaction was used to set up qRT-PCR reactions in triplicates using RASSF1A and RASSF1C specific primer. RASSF1A forward primer is 5'GGCGTCGTGCGCAAAGGCC and RASFF1C forward primer: is 5'CTGCAGCCAAGAGGACTCGG 3', and the reverse primer for both RASSF1A and 1C is 5'GGGTGGCTTCTTGCTGGAGGG 3' [2]. The real-time PCR reactions were set up using SYBR green PCR master mix (Bio-Rad, Hercules, CA) and the PCR reactions were run using the Opticon 2 PCR machine (Bio-Rad). The PCR reactions were run using the following protocol: 1. incubate at 95°C for 10 min, 2. incubate at 95°C for 15 sec, 3. incubate at 60°C for 30 sec, 4. incubate at 72°C for 30 sec, 5. go to step 2 for 39 more cycles, 6. melting curve from 60°C to 95°C, read every 1.0°C.

Western blot analysis was carried out using the Odyssey^® Infrared System (LI-COR Biosciences, Lincoln, NE). Anti-caspase 3, P-ERK1/2, total ERK1/2, and -GHR antibodies were purchased from Santa Cruz Biotechnology (1: 250-1000, Santa Cruz, CA), the anti-HA antibody was purchased from Covance (1:1000; Berkeley, CA), the anti-CXCR4 antibody was purchase from Millipore (1:500; Temecula, CA), and the anti-trans glutaminase 2 (1:1000; TGM2) antibody was purchased from Sigma (Saint Luis, MO). Fluorescently labeled secondary antibodies were purchased from LI-COR Biosciences (1:5000). Cells were collected and lysed in RIPA buffer supplemented with protease inhibitors. 30-40 ug of protein was separated on 12% SDS-PAGE gels and transferred to nitrocellulose membranes. The membranes were blocked overnight at 4°C in TBST and dried milk. Incubation with antibodies was carried out in Odyssey^® Infrared System blocking buffer (diluted 1:10 in PBST blocking buffer).

Hybridization of 12 μg of labeled cRNA to an Affymetrix U133 plus 2.0 chip was carried out in triplicates and data analyses were carried out at the UCLA Microarray facility core, Department of Pathology. The control sample is RNA from T47D cells stably transduced with MLV-backbone (T47D-BB) and the experimental sample is RNA from T47D cells stably transduced with MLV-RASSF1C (T47D-1C). Prior to RNA isolation, T47D-BB and T47D-1C cells were treated with 1 μg/ml doxycycline for 48 hr. Data analysis was performed using dChip [21]. Thresholds for selecting significant genes were set at a relative difference > = 1.5-fold (or/and 2-fold), absolute signal difference > = 50, and p < 0.05. Genes that met all three criteria were considered as significant changes. Comparison results with False Discovery Rate (FDR) <5% was considered as a valid analysis. The microarray data has been deposited in the Gene Expression Omnibus (GEO) data base and the accession number is GSE24473.

Caspase 3 gene primers were purchased from realtimeprimers.com (Elkins Park, PA). Other gene primers were as follows: Cyclophilin forward primer 5' GCATACAGGTCCTGGCATCT3 and reverse 5'TCTTGCTGGTCTTGCCATTC3'; CREG forward primer: 5'GTGCCCTATTTCTACCTGAGCC 3' and reverse primer 5'AGCATTATGTGAACACAAAGGGG 3'; CXCR4 forward primer: ATGAAGGAACCCTGTTTCCGT and reverse primer 5' AGATGATGGAGTAGATGGTGGG 3'; EGR1 forward primer: 5'ATGAAGGAACCCTGTTTCCGT 3'and reverse primer 5' ATGATGGAGTAGATGGTGGG 3'; GHR forward primer 5' GCTGCTGTTGACCTTGGC 3' and reverse primer 5'ACCTCATCTGTCCAGTGG 3'; MRAS forward primer: 5'TTTGTGCCTGACTATGACCCC 3' and reverse primer 5' TGACGGAGTAGACGATGAGGA 3'; and HOXA1 forward primer: 5' CGCCTCAATACATTCACCACT 3', and reverse primer 5' CCAGCAGGACTGACCTGTTTT 3'. The RT-PCR reactions were carried out in triplicate and the fold change was calculated using the 2^-ΔΔC_T method [22].

Breast cancer cells were plated at 5000/well in 96-well plates 24 hrs before infection.

Cells were incubated with 8 μg/ml hexadimethrine bromide for two hours before virus particles were added. Cells were infected with Mission^® non-target shRNA control transduction particles or with multiple Mission^® lentiviral shRNA transduction particles (NMID: NM_007182) for silencing RASSF1C (Sigma, St. Louis, MO). Since the lentiviral shRNA Transduction Particles (NMID: NM_007182) used in this study would target both RASSF1A and RASSF1C, we used breast cancer cells that express RASSF1C but not RASSF1A. The infections were carried out using an MOI of at as outline in the supplier manual. Infected cells were selected in media containing 2 μg/ml puromycin for 2-4 weeks and then cells were harvested. Knockdown validation of RASSF1C expression was assessed by qRT-PCR using RASSF1C specific primers.

Caspase 3 activity was assayed using the Apo3/7 caspase activity assay (Promega, Madison, WI). Cells were plated in 96-well plates at 5000 cells/well and the next day cells were treated with doxycycline, DMSO, etoposide at 45 umol/ml, or doxycycline and etoposide for 48 hr before cells were assayed for caspase 3 activity. Etoposide was purchased from Sigma and diluted in DMSO (Sigma) to a concentration of 45 mM and doxycycline was purchased from Invitrogen (Carlsbad, CA).

Breast cancer cells stably over-expressing RASSF1C were incubated for 14 days in presence of 1 μg/ml doxycycline before cells were used to isolate genomic DNA for DNA fragmentation analysis using an Apoptotic DNA Ladder Kit (Roche Applied Science, Mannheim, Germany). Apoptotic DNA ladder corresponding to genomic DNA isolated from lyophilized apoptotic U937 cells that were treated with 4 ug/ml camptothecin for 3 hrs that were provided with the kit used as a positive control for apoptosis.

The 24-well plate BD BioCoat™ Matrigel™ Invasion Chamber (BD Biosciences. Bedfor, MA) was used to co-culture T47D breast cancer cells with human stromal cells, Hs27a according to the user manual. The Hs27a cells were seeded at 25,000 cells per well in the 24-well BD Falcon™ TC Companian Plate in DMEM supplemented with 10% calf bovine serum. T47D cells were seeded at a density of 25,000 cells in the Matrigel chambers, T47D-BB and T47D-1C were grown in RPMI-1640 medium supplemented with 10% calf bovine serum and 0.2 units/mL insulin. Then the chambers containing the T47D-BB and T47D-1C were transferred to the well containing the Hs27a stromal cells and incubated for 22 hours. MDA-MB231 and T47D cells were also seeded at a density of 25,000 cells in the Matrigel chambers and the chambers were transferred to wells containing either 40 ng/ml SDF-1 conditioned medium or control medium lacking SDF-1 and incubated for 22 hours. The cells in the lower surface of the membrane were fixed methanol and stained with 1% Toluidine blue per the user manual instructions. The stained membranes were photographed through the microscope and invading cells were counted.

Data are presented as mean values ± SEM and analyzed with Student's t-test. Values ≤0.05 were considered significant.

Because RASSF1C and RASSF1A are structurally similar, but appear to have opposing effects, it is possible that they may interact and modulate each other's effects. Therefore, prior to silencing RASSF1C mRNA, the endogenous RASSF1A and RASSF1C mRNA levels were measured in MDA-MB231 and T47D breast cancer cells. RASSF1C is readily detectable, while RASSF1A is barely detectable in both cell lines (Figure 1A).

Next, expression of RASSF1C was silenced with small interfering RNA (siRNA) technology. The siRNA-RASSF1C plasmid used in this study is one of three RASSF1C siRNA plasmids [18] that we previously demonstrated to consistently reduce HA-RASSF1C protein expression compared to non-target siRNA oligos as judged by Western blot analysis using anti-HA antibody [19].

Cells transfected with siRNA-RASSF1C plasmid showed a significant decrease (p < 0.05) in cell proliferation compared to cells transfected with control plasmid as judged by the alamar blue and the ³H-thymidine incorporation assays (Figure 1B and 1C). To confirm that the inhibitory effect of RASSF1C siRNA on cell number correlated with reduction of RASSF1C mRNA, RASSF1C mRNA levels were measured in MDA-MB231 and T47D cultures treated with siRNA-RASSF1C or control plasmid. Figure 1D shows that transient transfection with siRNA-RASSF1C reduced RASSF1C mRNA levels in these breast cancer cells. We have also confirmed our plasmid silencing data using Mission^® lentiviral shRNA transduction particles to silence RASSF1C expression in T47D cells (Figure 1E). These findings suggest that RASSF1C appears to be important in promoting breast cancer cell growth.

To further elucidate the function of RASSF1C and demonstrate that RASSF1C is not a tumor suppressor, we carried out RASSF1C over-expression studies in breast cancer cells using a tet-inducible Mouse Leukemia Virus-based retroviral vector to express HA-tagged RASSF1C fusion protein. Cells were stably transduced with MLV-backbone (BB) or MLV-RASSF1C (1C) as outlined in Materials and Methods. Western blot analysis using an anti-HA-tag antibody to detect the HA-RASSF1C fusion protein verified that RASSF1C was over-expressed in cells transduced with the MLV-RASSF1C vector following treatment with 1 μg/ml doxycycline for 48 hr (Figure 2A). Over-expression of RASSF1C did not inhibit cell proliferation. Instead, it consistently resulted in a small but reproducibly and statistically significant increase in cell proliferation of Hs578T, MDA-MB231, and T47D cells stably transduced with MLV-HA-RASSF1C (1C) compared to an empty MLV-backbone (BB) as demonstrated by ³H-thymidine cell proliferation assays (Figure 2B and 2C). These findings demonstrate that RASSF1 C over-expression does not inhibit breast cancer cell growth and may suggest a potential role of RASSF1C in promoting cancer cell growth and progression.

We assessed the expression of RASSF1A and RASSF1C in primary mammary epithelial (AG1132B) cells and established breast cancer cell lines using quantitative real time PCR (qRT-PCR). The RT-PCR reactions were carried out in triplicates and the fold change was calculated using the 2^-ΔΔC_T method [22]. Interestingly, RASSF1C expression was at least 6 fold higher and RASSF1A was at least 2.5 fold lower in the breast cancer cell lines (Hs578T, MDA-MB231, and T47D) compared to the normal mammary epithelial cells, AG1132B) (Table 1). The elevated expression of RASSF1C detected in established breast cancer cell lines compared to primary cells is indeed consistent with our hypothesis that RASSF1C, unlike RASSF1A, is a potential growth and survival factor in breast cancer.

The observed increase in cell number in breast cancer cells over-expressing RASSF1C predicted that over-expression of RASSF1C might either down-regulate the expression of cell growth inhibiting/pro-apoptotic genes or up-regulate the expression of cell growth promoting/anti-apoptotic genes. Affymetrix-microarray analysis was performed using T47D cells over-expressing RASSF1C to answer this question. The control sample was RNA from T47D cells stably transduced with MLV-backbone (T47D-BB) and the experimental sample was RNA from T47D cells stably transduced with MLV-RASSF1C (T47D-1C). Prior to RNA isolation, T47D-BB and T47D-1C cells were treated with 1 μg/ml doxycycline for 48 hr. Data analysis was performed using the dChip program [21] and the thresholds for selecting significant genes were set at a relative difference > = 1.5-fold (or/and 2-fold), absolute signal difference > = 50, and p < 0.05. Genes that met all three criteria were considered as significant changes. Comparison results with False Discovery Rate (FDR) <5% was considered as a valid analysis.

We found that RASSF1C over-expression modulated the expression of a number of genes that are involved in cancer development, cell growth/proliferation, cell cycle, cell death, and apoptosis (Table 2). RASSF1C down-regulated several pro-apoptotic and tumor suppressor genes, including Bcl2-associated protein (BAX) [23], Caspase 3 [24], disabled homolog 2 (DAB2) [25], epithelial membrane protein 1(EMP-1) [26], insulin-like growth factor binding protein 3 (IGFBP-3) [27], mitochondrial tumor suppressor 1 (MTUSI) [28], ring finger protein 182 (RFN182) [29], SRY (sex determining region Y)-box 9 (SOX9) [30], sushi-repeat-containing protein, X-linked (SPRX) [31], transglutaminase 2 (TGM2) [32], and transmembrane protein 158 (TMEM158, also known as Ras-induced senescence 1(Ris-1)) [33]. RASSF1C also up-regulated several growth promoting genes that include apolipoprotein E (APOE) [34], carboxypeptidase E (CPE) [35], chemokine (C-X-C motif) receptor 4 (CXCR4) [36], human growth hormone receptor (GHR) [37], homeobox A1 (HOXA1) [38], muscle RAS (MRAS) oncogene homolog [39], SPANX family member A1 (SPANXA1), and SPANXB1 [40]. The RASSF1C target genes identified in this study are consistent with a potential growth promoting role for RASSF1C in breast cancer cells. We then selected several RASSF1C target genes and confirmed the microarray results using RT-PCR analysis (Table 3). We also show that changes in mRNA levels of caspase 3, CXCR4, GHR, and TGM2 are indeed translated to a change in protein expression in T47D cells (Figure 3A). We also found that T47D cell over-expressing RASSF1C displayed higher levels of phosphorylated ERK1/2 compared to control cells (Figure 3B). It should be noted that total ERK1/2 levels were the same in both T47D-BB and T47D-1C (Figure 3C). In addition, we show that silencing of endogenous RASSF1C expression in T47D cells resulted in an increase in caspase 3 and a decrease in CXCR4 mRNA expression (Table 4).

Because RASSF1C over-expression up-regulates the expression of CXCR4, a key metastasis gene, we carried out an in vitro invasion assay to determine if T47D cells over-expressing RASSF1C and grown in the presence of SDF-1 (the ligand for the CXCR4 receptor) were more invasive than control cells. For the invasion assays, T47D cells over-expressing RASSF1C were co-cultured with the human stromal cell line Hs27a, which secretes SDF-1, or were grown in conditioned serum-free medium containing SDF-I. Figure 4A shows that T47D-1C were significantly more invasive than the T47D-BB control cells. Interestingly, silencing of RASSF1C in T47D cells using lentiviral shRNA transduction particles significantly reduces T47D cell invasion/migration compared to cells infected with lentiviral shRNA control transduction particles (Figure 4B), further supporting that RASSF1C may promote breast cancer cell invasion/migration.

In addition to T47D cells, we also show that MDA-MB231 cells over-expressing RASSF1C were more invasive than the control cells (Figure 4C). All together, our novel findings suggest that RASSF1C may promote breast cancer cell invasion/migration perhaps in part through the up-regulation of the expression of the CXCR4 gene.

Etoposide is a chemotherapy agent that is known to induce apoptosis through activation of caspase 3. Since over-expression of RASSF1C down-regulates caspase 3 expression, we hypothesized that over-expression of RASSF1C would reduce the amount of active caspase 3 that is induced by etoposide. We tested this hypothesis by measuring the amount of caspase 3 produced in response to etoposide by T47D breast cancer cells that either over-express or normally express RASSF1C. RASSF1C over-expressing cells (T47D-1C) exhibit reduced caspase 3 activity compared to cells that do not over-express RASSF1C (T47D-BB) when treated with etoposide (Figure 5). To further show that prolonged over-expression of RASSF1C does not promote apoptosis in breast cancer cells, DNA fragmentation analysis was carried out using genomic DNA isolated from breast cancer cells treated with 1 μg/ml doxycycline for 14 days. Over-expression of RASSF1C did not induce DNA fragmentation in breast cancer cells (Figure 6). These findings further support our hypothesis that RASSF1C plays a role in promoting breast cancer cell growth, and it may allow cancer cells to evade killing by chemotherapy agents.