Abstract
Background/Aim: NOTCH-regulated ankyrin repeat protein (NRARP) has been implicated in crosstalk between NOTCH and wingless-type mouse mammary tumor virus integration site (WNT) signals during development. Our study aimed to clarify its role in breast cancer cells. Materials and Methods: Public microarray data were used to analyze gene expression in human and rat breast cancer. A short interfering RNA was introduced into MCF7 and T47D human breast cancer cells for NRARP silencing. Gene expression was analyzed by quantitative polymerase chain reaction. Results: The NRARP transcript was commonly overexpressed in various rat mammary cancer models. In addition, a subset of human breast cancer also expressed high levels of NRARP transcript, which correlated positively with up-regulation of cell proliferation-related genes. Silencing of NRARP suppressed the growth of MCF7 and T47D cells and lowered the expression of cell cycle-related genes in MCF7 cells. Conclusion: NRARP may stimulate cell proliferation in human breast cancer.
Breast cancer is frequent among women in developed countries and is a major cause of cancer death (1). The formation of breast cancer involves changes of expression of many genes. For example, in 20-30% of human breast carcinomas, epidermal growth factor receptor-2 (HER2) is overexpressed; this cancer subtype can be effectively treated with targeted therapeutics (2). Identification of overexpressed genes in a subset of cases might, therefore, contribute to the discovery of candidate target molecules for breast cancer prevention and therapy.
Mammary carcinomas in rat, either spontaneously arising or experimentally-induced by carcinogens, have served as models of human breast cancer because of similarities in histopathology and hormone dependence (3). We hypothesised that common alterations of gene expression in rat models of mammary cancer play a vital role in the formation of rat mammary cancer as well as human breast cancer. We herein re-analyzed our previous microarray data on rat mammary cancer (4), which included spontaneous carcinomas and those induced by radiation, 1-methyl-1-nitrosourea (MNU), and 2-amino-1-methyl-6-phenylimidazo [4,5-b]pyridine (PhIP). We found that NOTCH-regulated ankyrin repeat protein (NRARP; Nrarp for the rat gene) was commonly overexpressed in the mammary carcinomas of these four different aetiologies. NRARP has been reported to regulate the NOTCH and wingless-type mouse mammary tumor virus integration site (WNT) signalling pathways (5-8), two pathways that are vital to breast cancer cell proliferation (9, 10). To clarify the function of NRARP overexpression in human breast cancer, we analyzed published microarray data on gene expression of primary human breast carcinomas with distinct NRARP levels. We further investigated the effect of silencing of NRARP on expansion of two human breast cancer cell lines.
Materials and Methods
Microarray data analysis of rat carcinomas. Microarray data were analysed using the GeneSpring GX 11.5.1 software (Agilent Technologies, Santa Clara, CA, USA). We used gene expression profiles of rat mammary carcinomas from our previous study (4) [Gene Expression Omnibus database, http://www.ncbi.nlm.nih.gov/geo/; accession no., GSE44786). These mammary carcinomas had been obtained from female Sprague-Dawley rats that were either left untreated, irradiated with γ-rays (1 Gy), injected with MNU (40 mg/kg, once), or administered PhIP (40 mg/kg/day for 10 days by gavage); gene expression data had been obtained for spontaneous (n=3), radiation-induced (n=4), MNU-induced (n=9), and PhIP-induced (n=4) adenocarcinomas as well as normal mammary tissue (n=3) using GeneChip Rat Genome 230 2.0 microarrays (Affymetrix, Santa Clara, CA, USA) carrying 31,024 probe sets. In the present analysis, gene data were excluded if their raw value was less than the lower quartile in a microarray. Fold change of expression values compared to normal tissue was calculated when data were present for all samples. Genes showing more than 5-fold up-regulation in all four models were selected using one-way analysis of variance with the Benjamini-Hochberg's false discovery rate (q) set at 0.05.
Microarray data analysis of human carcinomas. Microarray data on four cohorts of patients with breast cancer [expO (11), IPC (12), Rotterdam (13) and Boston (14); accession no., GSE2109, GSE21653, GSE12276, and GSE19615] on GeneChip Human Genome U133 Plus 2.0 microarrays (Affymetrix), containing 54,675 probe sets, were obtained from the above mentioned database. Samples having a very small or large NRARP value (probe set ID, 226499_at; lowest or highest 10% of the samples) were selected from each of the four data series. Genes whose expression correlated with NRARP expression were identified based on Pearson's correlation coefficient (r) in each data series. Gene lists were imported into the Database for Annotation, Visualization and Integrated Discovery (DAVID) tool (v6.7) to identify relevant pathways in the Kyoto Encyclopedia of Genes and Genomes (KEGG) (15, 16).
Cell culture. Human MCF7 and T47D breast cell lines and the untransformed mammary epithelial cell line MCF10A were obtained from the American Type Culture Collection (Manassas, VA, USA). MCF7 and T47D cells were maintained in RPMI-1640 (Sigma-Aldrich, St. Louis, MO, USA) containing 10% (v/v) heat-inactivated fetal bovine serum (Biowest, Nuaillé, France), 100 U/ml penicillin, and 100 μg/ml streptomycin (Sigma-Aldrich). MCF10A cells were maintained in a 1:1 mixture of Dulbecco's modified Eagle's medium and Ham's F12 medium supplemented with 5% (v/v) equine serum, 10 μg/ml bovine insulin (Sigma-Aldrich), 20 ng/mL recombinant human EGF (BD Biosciences, Franklin Lakes, NJ, USA), 100 ng/ml cholera toxin (Cosmo Bio, Tokyo, Japan), 0.5 μg/ml hydrocortisone (Sigma-Aldrich), 100 U/ml penicillin, and 100 μg/ml streptomycin. Sub-confluent MCF7 and T47D cells were detached from the culture dish by trypsinization and collected by centrifugation (300× g, 5 min). Cells were then suspended in the growth medium and strained sequentially through 70-μm and 40-μm sieves. To evaluate gene expression, cells were seeded at 1.7×105 and 2.5×105 cells/well, respectively, in 6-well plates and analyzed on day 4, before cells became confluent. To evaluate the effect of siRNAs on cell growth, MCF7 and T47D cells were seeded at 6×103 and 1×104 cells/well, respectively, in 96-well plates.
Knockdown experiments. A mixture of four short interfering RNA (siRNA) clones directed against human NRARP mRNA (siGENOME SMARTpool, cat. No. M-032409) and a control siRNA mixture (siGENOME Non-Targeting siRNA Pool #2, cat. No. D-001206-14) were obtained from Thermo Fisher Scientific (Waltham, MA, USA). Another pair of NRARP-directed and control siRNA clones was purchased from Life Technologies, Inc. (Silencer Select siRNA, Pre-designed, NRARP, cat. no. 4392420-s54365 and Negative Control #1 siRNA, cat. no. 4390843; Carlsbad, CA, USA). A mixture of siRNA (20 nM) and transfection reagent (Lipofectamine RNAiMAX, 0.07%; Life Technologies) in Opti-MEM medium (Life Technologies) was incubated for 15 min at room temperature and then added to freshly-seeded cells in the growth medium to give a final siRNA concentration of 3.3 nM. Every 24 h, a batch of cells was washed with phosphate-buffered saline, trypsinized, stained with trypan blue to assess viability (which was consistently 80-85%), and counted with a haemocytometer. Gene-knockdown efficiency was checked by one-step quantitative reverse transcription-polymerase chain reaction (RT-PCR) (CellAmp Direct RNA Prep kit and One Step SYBR PrimeScript RT-PCR kit II; Takara Bio, Otsu, Japan).
Quantitative RT-PCR. Total RNA was isolated with a Maxwell 16 Instrument (Promega, Madison, WI, USA) or AllPrep DNA/RNA Mini kit (Qiagen, Valencia, CA, USA). cDNA was synthesized using a Moloney murine leukemia virus reverse transcriptase (Toyobo, Osaka, Japan) and random hexamers. Quantitative PCR was performed on an Mx3000P real-time PCR system (Agilent Technologies) with a commercial mixture of Taq DNA polymerase and fluorescent dye (SYBR Premix Ex Taq; Takara Bio), and primers listed in Table I. The PCR program consisted of denaturation at 95°C for 10 s and 45 amplification cycles of denaturation at 95°C for 5 s and annealing/elongation at 60°C for 20 s. A 1:10,000 dilution of an amplified PCR product served as a standard for each gene. Limiting dilutions of the standard were used to construct a standard curve, which served to calculate the efficiency of PCR (which ranged from 83-94%) and to determine expression values. Expression values were further standardised between samples using the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH). When PCR efficiency was not significantly different between the compared genes, relative expression as a percentage of GAPDH was used as a measure of relative expression. The significance of differences between groups was assessed on logarithmic-transformed data.
Results
Nrarp is overexpressed in different rat models of mammary cancer. To search for genes commonly overexpressed in multiple breast cancer models, we analyzed our previously published microarray data on four models of rat mammary carcinomas (4), including spontaneous, radiogenic, and chemically induced (MNU or PhIP) cancer. We found 12 transcripts having gene annotations and showing robust overexpression in all models (q<0.05, and ≥5-fold vs. normal mammary tissue; Table II). Many of these genes (e.g. Areg, Ccnd1, Dsp, Id4, Mmp3, Prlr, and Scd) have well-documented roles in cancer (17-23). The functional role of Nrarp in cancer cells is unknown. Nrarp is positively-regulated by the Notch signaling pathway (24) and, conversely, plays a pivotal role in regulating Notch and Wnt signaling pathways during embryonic and organ development of various animal species (5-8). Indeed, Notch1 and Notch4 were overexpressed (q<0.05) by ~2-fold in all examined rat models of mammary cancer, whereas Notch2 and Notch3 were not (data not shown).
Association between NRARP expression and genes involved in cell proliferation in human breast cancer. Next we examined the consequences of NRARP overexpression on gene expression in human breast cancer. We first obtained gene expression profiles for 938 human breast carcinomas from four large-scale studies (11-14). Breast carcinomas having the lowest and highest 10% NRARP expression in a cohort were chosen for analysis. There was a 3- to 5-fold difference in NRARP expression between the medians of the two groups, with considerable variation within each group (Figure 1A). We then identified a set of genes whose expression levels correlated with NRARP expression (|r| ≥0.35; Figure 1B and C). Of these genes, the expression of 216 correlated positively with NRARP expression in at least three out of the four cohorts (Figure 1B, underlined). Several KEGG pathways (16) were associated with these genes, including ‘cell cycle’, ‘DNA replication’, ‘mismatch repair’, ‘Notch signalling pathway’, ‘pyrimidine metabolism’ and ‘spliceosome’ (cut-off, p<0.05 and q<0.3, Figure 1D). Most of these genes are involved in cell proliferation; the identification of the Notch signaling pathway is notable as it positively regulates NRARP (24). Sixty-six genes showed a consistent negative correlation with NRARP expression in at least three out of the four cohorts (Figure 1C, underlined). No KEGG pathway, however, was found to be associated with these genes. Thus, our findings suggest an association between high NRARP expression and up-regulation of cell proliferation in primary human breast cancer.
NRARP silencing reduces growth and gene expression of breast cancer cells. We then examined whether NRARP expression contributes to the growth of human breast cancer cells. First, we evaluated NRARP expression in human breast cell lines (MCF7 and T47D) by quantitative RT-PCR. Compared with an untransformed mammary epithelial cell line (MCF10A), NRARP expression was highly elevated in MCF7 and T47D cells (Figure 2A). As NRARP expression is up-regulated by Notch proteins (24), we also examined the expression of four Notch genes (NOTCH1, NOTCH2, NOTCH3, NOTCH4) in these cell lines. NOTCH1 expression was comparable between lines. In contrast, expression of NOTCH2 and NOTCH3 was elevated in the cancer cell lines (Figure 2A) compared to untransformed cells. NOTCH4 expression was too low to be quantified in all cell lines examined. The association between NRARP and NOTCH3 expression was concordant with our observation in primary breast cancer (Figure 1D), in which expression of NOTCH1 and NOTCH3 was positively correlated with NRARP expression.
We then knocked-down NRARP in these breast cancer cells to analyze the function of overexpressed NRARP. MCF7 and T47D cells were seeded such that in the absence of siRNA, they reached subconfluency on day 4. We silenced NRARP by transfecting the cells with a mixture of four siRNA clones targeting NRARP upon seeding and measured the number of cells on days 1 to 6 after transfection. The siRNA reduced NRARP expression by ~70% in MCF7 and T47D cells compared with cells transfected with a control siRNA cocktail (Figure 2B and D). Concomitantly, the number of cells decreased to ~50% after day 5 of transfection compared with cells treated with control siRNA (Figure 2C and E). Another siRNA clone targeting NRARP showed a similar tendency (data not shown).
We reasoned that the observed decrease in cell number should be preceded by changes in the expression of cell cycle-related genes. We therefore used quantitative RT-PCR to assess the expression of genes involved in cell-cycle regulation such as cyclins D1 (CCND1) and E1 (CCNE1), the retinoblastoma protein (RB1), cyclin-dependent kinases 2, 4 and 6 (CDK2, CDK4, CDK6), and transcription factors Dp-1, Dp-2 (TFDP1, TFDP2) on day 4 after transfection of MCF7 cells. Silencing of NRARP significantly reduced the expression of CCND1, CCNE1, CDK2, TFDP1, TFDP2, and RB1, whereas expression of CDK4 and CDK6 remained constant (Figure 3). Given that these affected genes govern the G1/S transition, we concluded that NRARP overexpression stimulates cell proliferation possibly by promoting this cell-cycle transition.
Discussion
Our findings indicate that NRARP (Nrarp in rat) is highly expressed in four models of rat mammary cancer and a subset of human primary breast cancer; concomitantly, certain NOTCH genes are up-regulated. An association was suggested between NRARP expression and pathways involved in cell proliferation in human primary breast cancer. Moreover, silencing of NRARP in human breast cancer cell lines led to reduced cell growth, which was preceded by down-regulation of a set of genes that govern the G1/S transition. Thus, NRARP overexpression may potentiate cell proliferation in human breast cancer.
NRARP was originally identified as a downstream target of NOTCH signalling (24), but to date, only a few reports have been published regarding NRARP function. Mammalian NRARP encodes a 114-residue protein that has two ankyrin-repeat motifs. NRARP of some vertebrate species reportedly blocks ubiquitination and subsequent degradation of the WNT pathway transcription factor LEF1, leading to activation of the WNT signal (6, 7). The WNT pathway is activated in human breast cancer by various mechanisms and positively regulates the expression of cyclin D1, a key regulator of G1/S transition (9). Thus, it is reasonable to speculate that NRARP overexpression may activate the WNT pathway in a subset of human breast cancer. The mechanism by which NRARP regulates the level of other transcripts (e.g. RB1, TFDP1 and TFDP2) is unclear, indicating the need for further study. It is also important to confirm that expression of NRARP protein corresponds with high expression of its transcript. We were unable to detect the protein by western blotting using both the antibodies that are commercially available (data not shown).
NRARP expression itself, on the other hand, is positively-regulated directly by the NOTCH pathway (24). As previously observed (25), the NOTCH pathway is often deregulated in human breast cancer. Our study confirmed deregulated expression of NOTCH genes in four rat mammary cancer models and two human breast cancer cell lines (Figure 2A). However, its overexpression showed interspecies differences (i.e. Notch1 and Notch4 overexpression in rats vs. NOTCH2 and NOTCH3 overexpression in humans). We also observed a positive correlation between NRARP expression and some NOTCH signalling pathway genes in human breast cancer (Figure 1D). These findings strongly support the idea that deregulation of NOTCH at least partially contributes to NRARP overexpression. Interestingly, NRARP is involved in suppression of NOTCH-dependent gene expression, indicating that it may be a key component in a negative feedback loop for the NOTCH signalling pathway (5-7). In light of the evidence that NRARP activates the WNT pathway in some vertebrate cells, it is tempting to hypothesise that NRARP plays a role in aberrant ‘rewiring’ of NOTCH and WNT pathways, in which NOTCH signals incoming from neighbouring cells lead to expression of WNT target genes in breast cancer cells.
Acknowledgements
We are grateful to Dr. Hiroyuki Nogawa (Chiba University) for careful reading and improving the manuscript. We thank the International Genomics Consortium and the Expression Project for Oncology for their efforts. This research was supported in part by MEXT KAKENHI (grant numbers 20710049 and 23710074). TO received a scholarship from the Japan Students' Services Organization.
Footnotes
-
↵* These Authors contributed equally to this work.
- Received February 17, 2014.
- Revision received March 9, 2014.
- Accepted March 10, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved