Abstract
Cell cycle regulatory pathways in breast cancer are incompletely described. Here, we report an important role in estrogen receptor α (ERα)–positive breast cancer cells for the protein kinase C1 (PKC1)–interacting protein RBCK1 in supporting cell cycle progression by driving transcription of ERα and cyclin B1. RBCK1-depleted cells exhibited increased accumulation in G2-M phase of the cell cycle, decreased proliferation, and reduced mRNA levels for ERα and its target genes cyclin D1 and c-myc. Chromatin immunoprecipitation revealed that ERα transcription is associated with RBCK1 recruitment to the ERα promoter, suggesting that transcriptional regulation is one mechanism by which RBCK1 affects ERα mRNA levels. G2-M phase arrest was mediated independently from reduced ERα levels, instead associated with transcriptional inhibition of the key G2-M regulator cyclin B1. In breast tumor samples, there was a positive correlation between levels of RBCK1, ERα, and cyclin B1 mRNA levels. Our findings suggest that RBCK1 regulates cell cycle progression and proliferation of ERα-positive breast cancer cells by supporting transcription of ERα and cyclin B1. Cancer Res; 70(3); 1265–74
Introduction
The cell cycle consists of several controlled phases that lead to DNA replication and cell division. The progression of these phases is strictly controlled by the timely expression and activation of activating and inhibitory proteins, such as different cyclins and their associated cyclin-dependent kinases (CDK; ref. 1). Alterations in the expression of cell cycle regulatory proteins are commonly found in many cancers (2).
Estrogens promote mammary epithelial cell proliferation and growth of estrogen-dependent breast cancer cells by stimulating the expression of genes encoding cell cycle regulatory proteins (3). The biological effects of estrogens are mediated through binding to two estrogen receptor (ER) isoforms, ERα and ERβ, which are members of the nuclear receptor superfamily of transcription factors (4, 5). Estrogenic control of the cell cycle in breast cancer cells has been well described for the G1-S phase transition (6). ERα increases the expression of cyclin D1 and c-myc in the presence of estradiol. Both of these factors promote cell cycle progression by decreasing the association between the cyclin E/CDK2 complex and the two CDK inhibitors p21Cip1/WAF1 and p27Kip1, an essential step for successful G1-S phase transition (7, 8). More than 90% of breast tumors have an altered expression of one or several regulatory proteins in the G1-S phase transition (9, 10).
Altered expression of cell cycle regulators in breast cancer has been best described and studied for the G1-S phase transition. However, key regulators of the G2-M phase, such as cyclin B1, Polo-like kinase 1 (Plk-1; ref. 11), and the cdc25B (12) phosphatase, are all overexpressed in breast cancer cells (13–15).
RING finger–containing E3 ubiquitin ligases have emerged as important proteins in breast cancer, in many cases displaying altered expression levels (16), and are therefore interesting diagnostic markers and drug targets (17). The Oncomine database3
identifies studies in which the E3 ubiquitin ligase RBCC protein interacting with protein kinase C1 (RBCK1; Genbank accession NP_112506; ref. 18) has an elevated mRNA expression in breast cancer compared with normal breast tissue (P < 0.001; ref. 19) and elevated mRNA levels in ERα-positive human breast cancer compared with ERα-negative human breast cancer tissue (P < 0.001; ref. 20).The human RBCK1 gene product has a predicted length of 510 amino acids. The NH2-terminal portion (1–250) of this protein contains an ubiquitin-like domain and a RanBP2-type zinc finger domain. Residues 282 to 332 comprise a C3HC4-type RING zinc finger domain, a domain that has been implicated in protein-protein interactions and DNA binding (Supplementary Fig. S1A). RBCK1 possess an autoubiquitination activity (18), a characteristic of E3 ubiquitin ligases, through its RING-between rings-RING (RBR; ref. 21) domain in the COOH-terminal portion of the protein (282–493). Bioinformatics further reveals that the human RBCK1 protein is homologous to E3 ligases involved in cell cycle regulation such as CUL9, RNF144b/p53RFP, and RNF14/ARA54 (Supplementary Fig. S1B; refs. 22–24).
RBCK1 was originally identified as a protein kinase C (PKC) subtypes η (25), β (26), and ζ (27) interacting protein. In rat cardiac cells, RBCK1 plays a role in PKCβ-dependent cell growth (26). Interestingly, the PKC family of isozymes has higher activity in breast cancer compared with normal breast tissue (28). All the PKC subtypes interacting with RBCK1 are important for progression into the S phase in breast cancer cells and for estrogen signaling (29–32).
Based on the role of RBCK1-interacting proteins in breast cancer and estrogen signaling, exhibiting homology to cell cycle regulatory proteins together with overexpression in ERα-positive breast cancer, we hypothesized that RBCK1 might have a regulatory function in cell cycle progression in breast cancer cells.
Materials and Methods
Cell culture
Original MCF-7 human epithelial breast cancer cells developed at the Michigan Cancer Foundation (33) were kindly provided by Dr. Robert P.C. Shiu (University of Manitoba, Winnipeg, Manitoba, Canada). Cells were cultured in DMEM (Invitrogen) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin (Invitrogen). Cells were incubated at 37°C in 5% CO2. A 10 μmol/L stock solution of 17β-estradiol (E2; Sigma) was prepared in ethanol. Experiments evaluating E2 treatment, FCS with dextran-coated charcoal, and DMEM without phenol red were used. Ethanol was used as vehicle control.
Small interfering RNA
The individual siRNA constructs are as follows: RBCK1, Individual Stealth Select siRNA referred to in this study as siRBCK1 oligo 1 and 2, respectively; ERα, Individual Stealth Select siRNA, referred to in this study as siERα oligo 1 and 2, respectively; and control siRNA, Individual Stealth Select siRNA called siControl (Invitrogen). siRNA transfections were carried out using a final concentration of 50 nmol/L oligo (at 40–60% confluence) by using INTERFERin transfection reagent (Polyplus Transfection) according to the manufacturer's recommendation.
Western blot analysis
Cells were seeded into six-well plates 24 h or into 10-cm plates 48 h before transfection, and then lysed 72 h posttransfection in Laemmli sample loading buffer. For immunoblotting, the primary antibodies used were as follows: monoclonal mouse β-actin antibody (Sigma Aldrich), polyclonal goat RBCK1, polyclonal rabbit cyclin B1, and ERα antibody (Santa Cruz Biotechnology).
RNA isolation and quantitative real-time PCR
MCF-7 cells were seeded into six-well plates 24 h before siRNA transfection. Total RNA was extracted using the RNeasy kit (Qiagen). Two micrograms of total RNA were reverse transcribed into cDNA using SuperScript III reverse transcriptase (Invitrogen) with random hexamers. Reverse transcription-PCR (RT-PCR) measurements were performed in triplicate in a 7500 ABI Real-time PCR thermocycler (Applied Biosystems) using the SYBR Green reagent (Applied Biosystems). cDNA was used to produce standard curves as 2-fold serial dilutions; Ct values were converted to nanograms; and RPLP0 and GAPDH were used as reference genes. Primer sequences are given in Supplementary Table S1. For the analysis of ERα-specific transcripts, primer sequences have been published before (34).
Chromatin immunoprecipitation
MCF-7 cells were seeded in 150-mm dishes and were transfected with siRNA against RBCK1 or siControl after 24 h. Chromatin was prepared 72 h posttransfection as previously described (35). Chromatin fractions were immunoprecipitated with 1 μg of the indicated antibodies, and the immune complexes were recovered using protein A/G-Sepharose (50% slurry; Pharmacia) and processed as described by Matthews and colleagues (35). The antibodies used were RBCK1 (Santa Cruz Biotechnology) and mouse anti-human IgG (Santa Cruz Biotechnology). Immunoprecipitated DNA was quantified by RT-PCR. Input DNA was used to produce standard curves as 2-fold serial dilutions and chromatin immunoprecipitation (ChIP) data were converted to percentages of total input. Primer sequences are given in Supplementary Table S1.
Quantification of cell viability
MCF-7 cells were seeded into 96-well plates 24 h before siRNA transfections. Metabolic activity was measured with WST-1 cell proliferation reagent (Roche), and the number of viable/proliferating cells was quantified at 450 nm using a plate reader (Powerwave X Select, Bio-Tek Instruments, Inc., KC4).
Fluorescent-activated cell sorting
The experimental outlines are described in detail in the supplementary materials. Cells were harvested and fixed in 70% ethanol for 30 min on ice followed by staining with propidium iodide. Staining was measured using flow cytometry by the fluorescence intensity (FL-1, 530 nm) of 10,000 cells, data acquisition was done on a BD LSR II flow cytometer (BD Biosciences), and data were analyzed using the software supplied by the manufacturer.
Human breast tumor samples
The 13 breast tumor samples in this study have been previously described (36). The ethical committee of the Karolinska Institute approved the studies.
Statistics
All experiments were carried out in triplicates. Student's t test, Pearson correlation coefficient, or multiple regression analysis (ANOVA) was used for group comparisons. A P value of <0.05 was considered significant.
Results
RNAi depletion of RBCK1 reduces cell proliferation by inducing G0-G1 and G2-M cell cycle arrest
RBCK1 expression in the human breast cancer cell line MCF-7 was shown by immunoblotting and RT-PCR (Fig. 1A). To determine the role of RBCK1 in cell proliferation and control of the cell cycle, we used siRNAs to decrease RBCK1 levels in MCF-7 cells as shown in Fig. 1A. Cells transfected with either of two RBCK1 siRNAs grew markedly slower than cells transfected with siRNA control (data not shown). Furthermore, we observed that RBCK1-depleted cells displayed a time-dependent decrease in proliferation compared with cells transfected with siRNA control (Fig. 1B).
Next, we determined the effect of RBCK1 depletion on E2-dependent cell cycle progression. The estrogen-induced entry into the S phase was completely abolished in the RBCK1-depleted cell population (Supplementary Fig. S2A and B; Fig. 1C). Additionally, we observed a significantly higher proportion of cells in the G2-M phase after RBCK1 depletion in both the E2- and vehicle-treated cell populations (Supplementary Fig. S2B; Fig. 1C). This was associated with a decrease in the number of cells in the G0-G1 phase (Supplementary Fig. S2B; Fig. 1C). These data suggest that RBCK1 depletion desensitizes the cells to estrogen, decreasing the E2-dependent entry into the S phase, and additionally causes an E2-independent arrest in the G2-M phase. Experiments performed in T-47D breast cancer cells showed similar results (data not shown).
Reduction of RBCK1 levels reduces the levels of ERα, potentially by reduced recruitment of RBCK1 to the ERα promoter
As RBCK1 depletion reduced E2-stimulated cell proliferation, we investigated the effects of RBCK1 depletion on ERα levels. RBCK1-depleted cells showed a decrease in ERα protein levels both in vehicle- and E2-treated cells (Supplementary Fig. S2A; Fig. 2A), suggesting that the observed inhibition of E2-dependent entry into the S phase after RBCK1 depletion is due to reduced ERα levels.
We further investigated the effects of RBCK1 depletion on the mRNA levels of ERα and ERα target gene mRNA levels. RBCK1-depleted cells had significantly decreased ERα mRNA levels for both vehicle- and E2-treated cells (Fig. 2B). Importantly, RBCK1 depletion reduced the expression of the ERα-responsive genes pS2 (37), cyclin D1, and c-myc, showing a decrease in ERα signaling (Fig. 2B). Decreased ERα signaling was also observed in T-47D cells under similar conditions (data not shown).
Seven promoters have been identified for the ERα gene (38). We found ERα transcripts derived from promoters A, B, and E2 (referred to as ERα mRNA A, B, and E2) to be expressed in MCF7 cells, with ERα transcripts derived from promoter B showing the highest expression as previously described (39). Transfection with siRNA-targeting RBCK1 resulted in significantly decreased transcripts derived from promoters A and B, whereas transcripts derived from promoter E2 was not changed (Fig. 2C, left). To assay the mechanism behind the promoter-specific downregulation of ERα, we investigated if endogenous RBCK1 associates with the ERα promoters. We performed ChIP assays to detect the recruitment of RBCK1 to promoters A, B, and E2, respectively. RBCK1 was recruited to the ERα promoter B and the recruitment was reduced after RBCK1 depletion (Fig. 2C, right). However, we could not observe the recruitment of RBCK1 to promoters A and E2 (Fig. 2C, right).
The determination of RBCK1 and ERα mRNA levels in ductal breast cancers showed a strong positive correlation between RBCK1 and ERα mRNA levels (r2 = 0.31; P < 0.05; Fig. 2D). Taken together, our data suggest that RBCK1 recruitment to the ERα promoter regulates ERα expression, and thereby estrogen signaling, in breast cancer cells.
ERα depletion does not affect the progression of the G2-M phase
To directly determine the importance of reduced ERα expression and signaling, which was observed upon RBCK1 depletion, for the S phase and G2-M phase in MCF-7 cells, we performed a siRNA-mediated knockdown of ERα. ERα mRNA and protein levels were reduced 72 hours posttransfection (Fig. 3A and B). As for RBCK1-depleted cells, the estrogen-induced progression into the S phase was abolished in ERα-depleted cells (Supplementary Fig. S3A and B; Fig. 3C), with a corresponding increase in the number of cells in the G0-G1 phase (Fig. 3C). These results are in agreement with previous studies about the role of ERα in cell cycle progression (6). Additionally, we observed no significant difference in the number of cells in the G2-M phase (Fig. 3C). ERα protein levels were reduced in vehicle- and E2-treated cells (Supplementary Fig. S3A; Fig. 3D) to similar levels as in RBCK1-depleted cells (Fig. 1D). These data are consistent with that the inhibitory effect of RBCK1 knockdown on the G1-S phase progression in MCF-7 cells is related to the inhibition of ERα signaling while the RBCK1-induced G2-M cell cycle arrest occurs in an ERα-independent manner.
RBCK1 depletion affects the expression of genes crucial for mitotic entry
The frequency of double nuclei in RBCK1-depleted cells was the same as in control cells (data not shown). This indicates that the arrest in the G2-M phase due to RBCK1 depletion occurs in the G2 or in the very early M phase (40). To pursue a potential mechanism for the G2-M arrest upon RBCK1 depletion, we analyzed the mRNA levels of genes crucial for the transition from the G2 to the M phase: cyclin B1, Plk-1, cdc25B, and cyclin B2. After 72 hours of knockdown of RBCK1, we found cyclin B1, Plk-1, and cdc25B, but not cyclin B2, mRNA levels to be decreased both in vehicle- and E2-treated cells (Fig. 4). Thus, decreased levels of cyclin B1, Plk-1, and cdc25B are associated with the observed G2-M arrest upon RBCK1 depletion.
Downregulation of cyclin B1 precedes G2-M arrest in RBCK1-depleted cells
Because the expression of several genes important for G2-M progression was affected after RBCK1 depletion, we conducted a time course study to investigate the series of events from RBCK1 depletion to G2-M arrest. MCF-7 cells were synchronized in minimal medium for 24 hours and were released by changing to a regular medium, at which point the cells were transfected with RBCK1 siRNA or control siRNA (Supplementary Fig. S4). The cells were analyzed at regular intervals for the distribution of cells in the different cell cycle phases and for the expression of selected mRNAs. We observed an increase in the proportion of cells in the G2-M phase already after 48 hours in the RBCK1 siRNA–transfected cell populations, increasing further at 72 hours (Fig. 5A).
RBCK1 mRNA levels started to decrease 24 hours posttransfection of RBCK1 siRNA relative to control (Fig. 5B). Notably, the reduction of cyclin B1 mRNA occurred at 36 hours in the RBCK1-depleted cells (Fig. 5B), whereas the accumulation of cells in the G2-M phase was first observed at 48 hours. Thus, the downregulation of cyclin B1 precedes the G2-M arrest. The levels of Plk-1 and cdc25B mRNAs were not decreased until at 72 hours in RBCK1-depleted cells (Fig. 5B). Altogether, our results suggest that downregulation of cyclin B1 could be the target responsible for the G2-M arrest induced by RBCK1 silencing. Immunoblot analysis confirmed decreased cyclin B1 protein levels in RBCK1-depleted cells compared with control cells (Fig. 5C). To initiate studies to reveal a potential mechanism for the reduction of cyclin B1 mRNA, we investigated RBCK1 recruitment to the cyclin B1 promoter by performing ChIP assays. However, no recruitment of RBCK1 could be observed, suggesting that RBCK1 affects cyclin B1 mRNA expression indirectly (data not shown).
Furthermore, we assayed mRNA levels of RBCK1 and cyclin B1 in tumor samples. We observed a positive correlation (r2 = 0.34, P < 0.05) between RBCK1 and cyclin B1 mRNA expression in the tumors (Fig. 5D), consistent with the observations in MCF-7 cells.
Discussion
This study identifies a role for RBCK1 in the regulation of cell cycle progression and proliferation in breast cancer cells. Three observations led us to investigate the role of RBCK1 in breast cancer cells. First, RBCK1 had originally been described as a protein associated with PKCs (25–27), proteins known to be important for cell cycle progression through the G1-S phase in breast cancer cells (29–32); second, RBCK1 mRNA is elevated in breast cancer samples (19, 20); and third, RBCK1 is a member of the RBR family of proteins (21), which includes several proteins involved in cell cycle progression (22–24, 41).
RBCK1 depletion led to a decrease in MCF-7 cell proliferation (Fig. 1B). Subsequent fluorescence-activated cell sorting (FACS) analysis of RBCK1-depleted cells revealed a marked increase in the number of cells in the G2-M phase (Fig. 1C). Additionally, E2-dependent progression into the S phase, which requires ERα, was significantly reduced in RBCK1-depleted cells (Fig. 1C). ChIPs revealed that, in MCF-7 cells, RBCK1 is recruited to the major ERα promoter (Fig. 2C). RBCK1 depletion resulted in reduced recruitment to the promoter and reduced ERα expression (Fig. 2A and C), suggesting that RBCK1 might act as a novel transcription factor or transcriptional cofactor for regulation of ERα. Although RBCK1 depletion also reduced ERα transcripts derived from promoter A, we could not observe any recruitment of RBCK1 to promoter A. However, because the distance between ERα promoters A and B is short, it is possible that RBCK1 binding to promoter B also regulates transcription from promoter A.
RBCK1 has previously been shown to shuttle between the nucleus and cytoplasm (42) and has been suggested to be a potential transcriptional activator because it can bind DNA through its RING finger (43), supporting the possibility that RBCK1 might act as a transcription factor. The other RBR family proteins that we found to be homologous to RBCK1 are also E3 ligases with transcriptional activity; RNF14/ARA54 regulates cell cycle progression through affecting cyclin D1 expression in colon cancer cells (24), and RNF31/ZIBRA has been reported to repress transcription through direct binding to promoters (44).
As inhibition of ERα reproduces the effects observed with inhibition of RBCK1 in the S phase but not the G2-M phase (Fig. 3C), we hypothesize that RBCK1 inhibition affects cell proliferation by two distinct pathways: one that acts through ERα and one that acts independently of ERα (Fig. 6). The E2-dependent regulation of ERα target genes decreased upon RBCK1 depletion, including cyclin D1 and c-myc (Fig. 2B), which are known to have key roles in the G1-S transition (6).
The RBCK1 siRNA–induced increase of cells in the G2-M phase could be due to downregulation of cyclin B1 (Fig. 4) because downregulation of this protein precedes G2-M arrest (Fig. 5A). The activation of the CDK1/cyclin B1 complex is a critical step in the transition from the G2 to the M phase (1), which is regulated by the Plk-1 kinase and the cdc25B phosphatase (11, 45). However, Plk-1 and cdc25B seemed to be downregulated as a consequence of the G2-M arrest (Fig. 5B). Interestingly, RBCK1 does not seem to regulate cyclin B1 expression through recruitment to the cyclin B1 promoter, as seems to be the case for regulation of ERα.
During the final preparation of this article, a global study by Mullenders and colleagues (46) identified that RBCK1 depletion might lead to failed induction of p21Cip1/WAF1 expression when inducing cell cycle arrest in the human fibroblast cell line BJtsLT. However, we found that RBCK1 depletion increased p21Cip1/WAF1 expression in MCF-7 cells (data not shown); this is expected because c-myc is a transcriptional repressor of p21Cip1/WAF1 (47) and RBCK1 depletion leads to decreased levels of c-myc (Fig. 2B). It is unclear why RBCK1 affects p21 differently in these two studies, but it might reflect the use of different cell lines as model systems.
The RBCK1 homologues and the E3 ubiquitin ligases CUL19, RNF144b, and RNF14 regulate cell cycle progression (22–24), and further investigation will reveal whether RBCK1 is a de facto E3 ubiquitin ligase and whether this activity is required for cell cycle regulation.
PKC isoforms β and η have been shown to interact with RBCK1 and the β isoform phosphorylates RBCK1, thereby preventing its self-ubiquitination (18) leading to increased levels of RBCK1. Previous work in breast cancer cells has shown that active PKC β and η, as well as estrogen through ERα, increase the expression of cyclin D1 (30, 32). One possible mechanism of PKC regulation of cyclin D1 expression is PKC-mediated phosphorylation of RBCK1, which leads to its stabilization, thereby increasing the levels of ERα and cyclin D1. Further work will be needed to elucidate RBCK1 function in the context of PKC signaling in ERα-dependent breast cancer cells.
In agreement with the effect of RBCK1 depletion on ERα and cyclin B1 levels in breast cancer cell lines, we found that RBCK1 mRNA correlates with ERα (Fig. 2D) and cyclin B1 (Fig. 5D) expression in breast cancer tumors. To further extend the analysis of breast cancer samples, we analyzed publicly available microarray data to investigate correlations between RBCK1 and ERα expression. Gene expression–profiling studies performed on breast cancer tissues were selected from Oncomine. Multiple regression analysis revealed a positive and significant correlation between RBCK1 and ERα mRNA expression for the investigated studies (Supplementary Table S2).
Therapy using antiestrogens is the mainline treatment of ERα-positive breast cancer and reduces the death rate in patients by 30% (48). Interestingly, cyclin B1 levels tend to be increased in breast tumors (13) and have been suggested as a prognostic factor for breast cancer (49). It is thus of interest to find upstream regulators of ERα and cyclin B1 expression; indeed, transcriptional regulation by RBCK1 might provide a target for affecting the expression of these genes.
Disclosure of Potential Conflicts of Interest
J-Å. Gustafsson is shareholder of KaroBio AB and consultant of KaroBio AB and Bionovo. The other authors disclosed no potential conflicts of interest.
Acknowledgments
We thank Nina Heldring for the critical reading of the manuscript and Eric W-F. Lam from Charing Cross Hospital, London, for providing the breast tumor tissue samples.
Grant Support: The Swedish Cancer Fund (J-Å. Gustafsson and K.D. Wright). N. Gustafsson was supported in part by a Ph.D. fellowship, KID-medel, from the Karolinska Institute.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.