Downregulation of RBBP6 variant 1 during arsenic trioxide-mediated cell cycle arrest and curcumin-induced apoptosis in MCF-7 breast cancer cells

Aim: To determine the expression patterns of the RBBP6 spliced variants during arsenic trioxide-mediated cell cycle arrest and curcumin-induced apoptosis in MCF-7 cells. Materials & methods: As2O3 and curcumin were used to study cytotoxicity, cell cycle arrest, apoptosis and the expression of RBBP6 variants. The MUSE Cell Analyser was used to analyze cell cycle arrest, apoptosis and multicaspase activity while apoptosis was further confirmed using microscopy. Semi-quantitative RT-PCR was employed to quantitate the expression of the RBBP6 variants. Results: This study showed that the MCF-7 cells expressed RBBP6 variant 1 but lacked both variant 2 and variant 3. Both As2O3 and curcumin significantly downregulated RBBP6 variant 1 (p < 0.001). Conclusion: RBBP6 variants are promising therapeutic targets.

the Muse™ Cell Cycle Kit following the manufacturer's instructions (Merck Millipore, Darmstadt, Germany) and analyzed using the Muse R Cell Cycle Analyzer (Merck Millipore).

Multi-caspase assay
The MCF-7 cells were cultured in 6-well plates, overnight; starved for 12 h and treated for 24 h with curcumin, which also served as a positive control and arsenic trioxide concentrations (11 and 32 μM). After treatment with the arsenic trioxide and the positive control (curcumin), the percentages of cells with activated caspases were quantified using the Muse™ Multi-Caspase kit following the manufacturer's instructions (Merck Millipore) and analyzed using the Muse Cell Analyzer.

Apoptosis assay
The MCF-7 cells were cultured in 6-well plates and allowed to settle overnight; starved for 12 h and then treated for 24 h with curcumin, which also served as a positive control and arsenic trioxide concentrations (11 and 32 μM). After treatment with arsenic trioxide and curcumin, the percentages of apoptotic cells were quantified using the Annexin V/FITC kit following the manufacturer's instructions (Merck Millipore) and analyzed using the Muse Cell Analyzer.

Semi-quantitative reverse transcription polymerase chain reaction
The MCF-7 cells were plated at a density of 4 × 10 3 cells per T-25 culture flask in complete medium for 24 h and starved for further 12 h. The cells were then treated with arsenic trioxide, cobalt chloride and curcumin and incubated for 24 h. The untreated control and the treated samples were washed with 1× PBS, twice. Total RNA from all the cell groups was extracted using the TRIzol reagent (Thermo Fisher Scientific, MA, USA) and reverse transcription (RT) was performed using the AMV reverse transcription system manufacturer's instructions (Promega, WI, USA). cDNA was amplified using both the GAPDH primers (forward primer: AGCTGAACGGGAAGCTCACT; reverse primer: TGCTGTAGCCAAATTCGTTG) and RBBP6 primer sets specific to different variants: (RBBP6 variant 3 -forward primer: GGATAATATGTGGCATCACTTG; reverse primer: TCCCTGTATGA-CACTGTGTTG and RBBP6 variant 1 and 2 -forward primer: GTATAGTGTCCCTCCTCCAGG; reverse primer: GTAATTGCGGCTCTTGCCTCT). The PCR reactions were prepared using a 2× PCR Master Mix (Takara Bio Inc., Kusatsu Japan) according to the manufacturer's instructions. The reactions were subjected to 30 cycles comprising the three PCR steps (denaturation, annealing and extension) in the T100 Thermal Cycler (Bio-Rad, CA, USA). The products were electrophoresed on 1% agarose gels using a 100 bp DNA molecular weight marker (BioLabs, MA, USA) to confirm the sizes of the PCR products.

Immunocytochemistry assay
To evaluate the effect of arsenic trioxide on the expression and localization of Bax protein, immunofluorescence staining was performed. Breast cancer cells (MCF-7) were seeded on cover slips in 6-well plates at 1 × 10 5 cells/well and exposed to various concentrations of As 2 O 3 (11 and 32 μM), cobalt chlorite (100 μM) and curcumin (100 μM) for 24 h. After the 24-h incubation, the cells were washed twice with 1× PBS and fixed in 4% paraformaldehyde for 15 min at room temperature. After fixing, the cells were permeabilized with 0.25% Triton™ X-100 for 10 min at room temperature. Antibody-nonspecific binding was blocked using 0.5% BSA for 1 h at room temperature. Following blocking, the cells were incubated with anti-Bax primary antibody (Lot# E0707; Santa Cruz Biotechnology, TX, USA). A 1:50 dilution of the primary antibody was prepared in 1X PBS-BSA (0.5%) and incubated for one hr at room temperature. The cells were washed with sterile 1× PBS, then labeled with Flour R 488 goat anti-mouse IgG secondary antibody (Lot# 1810918; Thermo Fisher Scientific, MA, USA). The cells were counter-stained with DAPI (5 μg/ml) for 10 min and examined under the Nikon Eclipse TS100F Ti-E inverted microscope (Nikon Instruments, Shinagawa Japan).

Statistical analysis
The results of the different experiments (cell viability assay, cell cycle phases analysis and apoptosis analysis) performed in triplicates were expressed as the mean ± standard deviation (SD) using Instat Version 3 statistical software. Data were analyzed to get statistical significant differences by comparing two datasets (the untreated samples were compared with the treated samples) and the differences were considered significant when *p was ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.0001. Results were obtained from three independent experiments and were presented as ± standard error of the mean and the differences were considered significant when **p ≤ 0.01 and ***p ≤ 0.001.

Results
In vitro inhibition of MCF-7 cell growth by arsenic trioxide The effect of arsenic trioxide on the growth and viability of MCF-7 cells was assessed using MTT assay. As shown in Figure 1A, the arsenic trioxide significantly inhibited the growth and viability of MCF-7 cells in a dose-dependent   Figure 1B). After 24-h incubation, curcumin, which was also used as a positive control for apoptosis assay, reduced the growth of MCF-7 cells between 85 and 100 μM concentration ( Figure 1C). These results demonstrated that arsenic trioxide was potent in inhibiting the proliferation and inducing death of MCF-7 cells in vitro.

Morphological changes of MCF-7 cells due to cell cycle arrest & apoptosis
The treatment of the MCF-7 cells with arsenic trioxide, cobalt chloride and curcumin reduced the MCF-7 cell growth and adherence compared with the untreated control cells. Using DAPI staining and the inverted phasecontrast microscopy, 11 and 32 μM arsenic trioxide, 100 μM curcumin and 100 μM cobalt chloride affected the morphology of the MCF-7 cells exposed for 24 h. Many cells, consequently, shrunk, which is typical of cells undergoing apoptosis and evidently, losing their epithelial morphology (Figure 2A-F). DAPI staining revealed that treatment with arsenic trioxide (11 and 32 μM) and the positive controls (cobalt chloride and curcumin) induced typical apoptotic morphology. These included nuclear condensation, lost microvillus and apoptotic body formation ( Figure 3A  Arsenic trioxide induces caspase-dependent mode of death in MCF-7 cells Caspase activation correlates with the onset of apoptosis and caspase inhibition attenuates apoptosis. Therefore, the involvement of caspases in arsenic trioxide-induced cell death in MCF-7 cells was investigated ( Figures 6 & 7). As shown in Figure 6, the percentage of MCF-7 cells undergoing caspase-dependent mode of death after treatment with 11 μM of arsenic trioxide was found to be 51% ( Figure 6B). When the concentration of arsenic trioxide was increased to 32 μM, the percentage of cells undergoing caspase-dependent death also increased to 74% ( Figure 6C) compared with the untreated control ( Figure 6A). The same trend was observed in the positive control (100 μM of curcumin) that showed 67% ( Figure 6D).

Arsenic trioxide induces apoptosis in MCF-7 cells
To further confirm that arsenic trioxide treatment induced apoptosis in MCF-7 cells, the MUSE R Annexin V analysis was performed. Upon observation under the microscope, MCF-7 cells demonstrated apoptotic changes after treatment with 11 and 32 μM of arsenic trioxide but untreated cells maintained their epithelial morphology and remained attached on the culture flasks. The Muse Annexin V analysis showed that arsenic trioxide remarkably induced apoptosis in MCF-7 cells. As shown in Figures 8 & 9, percentage of cells undergoing apoptosis after treatment with 11 μM of arsenic trioxide was 44% ( Figure 8B) and when the concentration of arsenic trioxide was increased to 32 μM, the apoptosis percentage increased to 60% ( Figure 8C) when compared with the untreated control ( Figure 8A), the same trend was observed in the positive control (100 μM of curcumin) that induced 61% apoptosis ( Figure 8D) in MCF-7 cells.
Arsenic trioxide regulates the expression of RBBP6 transcripts Messenger RNA (mRNA) levels of RBBP6 transcripts in MCF-7 cells treated with arsenic trioxide (11 and 32 μM), curcumin (100 μM) and cobalt chloride (100 μM) for 24 h were analyzed using the conventional-PCR to determine Results were obtained from three independent experiments and were presented as ± SEM and the differences were considered significant when *p was ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.0001.  whether the RBBP6 can be implicated in the observed arsenic trioxide-induced cell cycle arrest and apoptosis. The RBBP6 variant 1 was found to be highly upregulated in the untreated MCF-7 cells when compared with the treated MCF-7 cells. Arsenic trioxide-, cobalt chloride-induced cell cycle arrest and arsenic trioxide-and curcumin-induced apoptosis downregulated the expression of RBBP6 variant 1. However, these compounds did not induce detectable levels of variant 2 (lanes 2-5 in Figure 10A) nor RBBP6 variant 4, which remains elusive and undetectable. GAPDH was used as a control to ascertain that equal amounts of the cDNAs from untreated and treated samples were used. Figure 10C (lanes 1-3) showed that normal cells (Hek 293 cells) express both variants 1 and 2 with the latter downregulated in breast cancer MCF-7 cells. Treatment of breast cancer cells with both arsenic trioxide and curcumin resulted in downregulation of the RBBP6 variant 1 in breast cancer cells as shown in Figure 10A (lanes 2, 3, 5), respectively. On the contrary, for the first time, we show that MCF-7 cells either do not express RBBP6 variant 2 or these cells express this variant at undetectable levels. Our results quantified from three independent experiments using Quantity One R 1D analysis software ( Figure 11) showed that variant 1 is highly expressed in MCF-7 cells and is downregulated by apoptosis inducers, arsenic trioxide and curcumin.
The mRNA levels of RBBP6 variant 3 were also determined using conventional-PCR. Figure 12 shows that the RBBP6 variant 3 is expressed by normal embryonic kidney cells compared with the breast cancer cells. The blank controls ( Figure 12 Figure 12A, lane 2). These results were quantified from three independent experiments using Quantity One R 1D analysis software ( Figure 13). Our results showed that both variants 2 and 3 were undetectable in cancer cells, especially breast cancer MCF-7 cells and cervical cancer Caski cells. This suggests that both variants 2 and 3 may be crucial in maintaining cell homeostasis, which is lost during carcinogenesis while the expression of variant 1 may favor the carcinogenesis process.
Arsenic trioxide regulates the expression of Bax protein in MCF-7 cells In this section, the effect of arsenic trioxide on Bax expression was investigated. Bax protein is lowly expressed in MCF-7 breast cancer cells ( Figure 14C & D) because it is involved in the induction of apoptosis, which is a process that is inhibited in breast cancer. The treatment of MCF-7 breast cancer cells with arsenic trioxide showed increased expression of Bax protein ( Figure 14E-H   in MCF-7 breast cancer cells, which suggest that RBBP6 variant 1 downregulation may be involved in the intrinsic apoptotic pathway.

Discussion
This study showed that RBBP6 variant 1 may be involved in breast cancer development and its expression in breast cancer cells may promote cell survival. It is not surprising that upon apoptosis induction, this variant is downregulated. On the other hand, the smaller RBBP6 variant 3 might be involved in the regulation of cell cycle  arrest, especially G2/M cell cycle arrest as previously shown in kidney embryonic cells [4], which was undetectable in breast cancer cells. Downregulation of cell cycle regulatory biomolecules favors the carcinogenesis process, hence downregulation of RBBP6 variant 3 in breast cancer is not surprising. It should be stressed that apoptosis is a well-defined and probably the most frequent form of programmed cell death, but nonapoptotic types of cell death might also be of biological significance [12]. Results were obtained from three independent experiments and were presented as ± standard error of the mean and the differences were considered significant when ***p < 0.001). The density was measured using Quanty-One software. Breast cancer cells are resistant to apoptosis, consequently, growing uncontrollably. In this study, we have shown that breast cancer cells may have lost the expression of RBBP6 variants 2 and 3 in favor of the carcinogenesis process. RBBP6 variant 4 has not been demonstrated but it is likely that it is one of the anticancer variants, which may be critical for maintaining cell homeostasis. RBBP6 has a wide range of functions and these include a role in cell cycle regulation, apoptosis, protein stability and mRNA processing [8,13,14]. RBBP6 has been shown to induce Results were obtained from three independent experiments and were presented as ± standard error of the mean and the differences were considered significant when **p < 0.001. The density was measured using Quanty-One software. cell cycle progression by ubiquitinating p53 through Murine Double Minute 2 (MDM2), therefore, promoting carcinogenesis [3]. RBBP6 variant 1 has ubiquitin ligase activity and this activity leads to enhanced degradation of p53, the cell guardian, which is crucial for antitumor formation. Furthermore, RBBP6 variant 3 is involved in G2/M arrest, but its absence triggers cell cycle progression and high proliferation rates in normal kidney cells [4]. Consequently, it would be advantageous for cancer cells not to express this variant to support their rapid growth. Some lines of evidence showed that the enhanced expression of different RBBP6 variants correlate with poor clinical prognosis in colon, prostate and esophageal cancer [15][16][17]. Breast cancer treatment remains a challenge and therefore, . Results were obtained from three independent experiments and were presented as ± standard error of the mean and the differences were considered significant when *p was ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.0001. The fluorescence intensity was measured using the Image J software (https://imagej.nih.gov/ij/docs/index.html).
there is a need for more specific and effective therapeutic tools. RBBP6 is a promising therapeutic target and there are a lot of promising drug development targets, such as arsenic trioxide (As 2 O 3 ). As 2 O 3 is a potent US FDA approved drug confirmed to treat acute promyelocytic leukemia in patients that relapsed after chemotherapy. Arsenic trioxide exhibits therapeutic effects in a variety of human cancers such as human hepatocellular carcinoma, gastric cancer and cervical cancer [18][19][20]. Arsenic trioxide hinders cancer development and progression through targeting cellular pathways, leading to inhibition of cell proliferation and invasion, and promoting apoptosis. Little is known about the expression and regulation of the human RBBP6 splice variants by arsenic trioxide during cell cycle progression and breast cancer development. Therefore, in this study, arsenic trioxide was chosen due to its antitumor effect in a variety of human cancers [18][19][20]. The effect of arsenic trioxide on MCF-7 cells was compared with the positive controls (curcumin and cobalt chloride) because these cytotoxic agents have also been shown to induce cell cycle arrest and apoptosis in cancer cells [10,11]. As previously shown, both the positive controls (curcumin and cobalt chloride) consistently induced apoptosis and cell cycle arrest, respectively in MCF-7 breast cancer cells.
This study has shown that arsenic trioxide and curcumin-induced cell cycle arrest and apoptosis in breast cancer MCF-7 cells, respectively (Figure 1-9). These findings support the results of previous studies showing similar effects on gynecological cancers and other solid tumors [9][10][11]21]. These findings suggest that As 2 O 3 , cobalt chloride and curcumin should additionally be investigated as potential novel chemotherapeutic agents for the adjuvant treatment of malignant human tumors. It was evident that the untreated control cells retained normal morphology and attached firmly to the culture plates with random orientation (Figure 2), while cells treated with arsenic trioxide, cobalt chloride and curcumin showed remarkable cellular effect, which included decrease in cell numbers, rounding effects, decrease in cell size, detachment from the substratum. Furthermore, DAPI staining ( Figure 3B & E) confirmed that arsenic trioxide induced cell cycle arrest and apoptosis that was shown by loss of cell microvilli, nuclear condensation and formation of apoptotic bodies.
Quantitative analysis of apoptotic cells and cell cycle arrest (Figures 4-9) using the MUSE Cell Analyzer demonstrated that As 2 O 3 induced evident caspase-dependent apoptosis but not necrosis in MCF-7 cells. It was also observed that arsenic trioxide induced G2/M arrest in MCF-7 cells ( Figure 4B). The G2/M arrest did not only inhibit the proliferation of cells but also triggered several apoptosis features that were evident after DAPI staining. These results suggest that As 2 O 3 -induced growth inhibition mainly depended on the induction of caspasedependent apoptosis and cell cycle arrest in MCF-7 cells. These results further showed that caspases are involved in arsenic trioxide induced apoptosis (Figures 6 & 7). They support a study which was done in MCF-7 breast cancer cells to show that As 2 O 3 exposure significantly increases the level of caspase-3 [22]. The upregulation of caspase 3 by arsenic trioxide was also shown in HT-29 colon cancer cells, suggesting that arsenic trioxide induces caspase-3-dependent pathway [23]. Even though caspase-3 has been previously reported to be inactivated in MCF-7 cells [24], this study has shown that caspase-dependent apoptosis is induced by arsenic trioxide and curcumin and RBBP6 variant 1 may be critical in the process. Other effector caspases may be involved in the induced apoptosis in MCF-7 cells. There are contradictory reports on the expression of caspase-3 in MCF-7 but it is possible that the mutated form responds to different apoptosis inducers [25].
We further showed that arsenic trioxide and curcumin regulate the expression of RBBP6 variants, especially the two big transcripts; variants 1 and 2 (Figures 10 & 11). Figure 10A (lane 1) demonstrated that breast cancer cells express RBBP6 variant 1 and lack the expression of variant 2. Treatment of these cells with arsenic trioxide, cobalt chloride and curcumin diminished the expression of RBBP6 variant 1. This suggested that this variant may promote breast cancer development while diminished expression of variant 2 may also support carcinogenesis. We previously suggested that RBBP6 variants and isoforms may have opposing cellular functions [4]. In this study, we showed that RBBP6 variant 3 is expressed in noncancerous cells, Hek 293 s but undetectable in breast MCF-7 cancer cells (Figures 12 & 13). We also showed that noncancer cells, at least, in Hek 293, express both variants 1 and 2. Arsenic trioxide did not only regulate the expression of RBBP6 transcripts but also regulated the expression of apoptosis-related protein, Bax ( Figure 14). These results showed that As 2 O 3 is effective against MCF-7 cells and also regulates the expression of RBBP6 variants, especially variant 1, which makes both RBBP6 variant 1 and arsenic trioxide an important therapeutic target.

Conclusion
This study showed that there are RBBP6 variants that are procarcinogenic and those that are anticarcinogenic. Different RBBP6 variants could be targeted for cancer therapeutic development. In conjunction with RBBP6 expression, arsenic trioxide, cobalt chloride and curcumin should be further explored as cancer drugs.

Future perspective
Little is known about the expression and regulation of the human RBBP6-spliced variants by arsenic trioxide, cobalt chloride and curcumin during cell cycle progression and breast cancer development. Taken together our findings indicate that arsenic trioxide, cobalt chloride and curcumin have substantial bioactivity against MCF-7 cells and regulate the expression of RBBP6 variants. This will allow scientists to target RBBP6 and these cytotoxic agents for future drug development. Further investigations using animal models and other breast cancer types are needed to provide new insights into the potential application of these cytotoxic agents in the treatment of breast cancer and regulation of RBBP6.

Summary points
• Arsenic trioxide, curcumin and cobalt chloride decreased the viability of the MCF-7 breast cancer cells in a dose-dependent manner.
• Arsenic trioxide and cobalt chloride induced G2/M cell cycle arrest of the MCF-7 cells.
• RBBP6 variant 1 may support the carcinogenesis process in breast cancer MCF-7 cells.
• RBBP6 variant 3 is against the carcinogenesis process and it is downregulated in MCF-7 cells.
• Arsenic trioxide upregulated the expression of Bax protein in MCF-7 breast cancer cells.
• Therefore, our study suggests that there are RBBP6 variants that are procarcinogenic and there are those that are anticarcinogenic. It also suggests that arsenic trioxide regulates both RBBP6 transcripts and Bax protein.
Author contributions L Makgoo and K Laka carried out the experiments and analyzed the data. Z Mbita supervised the work and the analysis. All the authors wrote the manuscript.