Shikonin reduces tamoxifen resistance through long non-coding RNA uc.57

Tamoxifen resistance is a serious problem in the endocrine therapy of breast cancer. Long non-coding RNAs play important roles in tumor development. In this study, we revealed the involvement of lncRNA uc.57 and its downstream gene BCL11A in TAM resistance. Tamoxifen-resistant MCF-7R cells showed lower expression of uc.57 and higher expression of BCL11A mRNA and protein than the parental MCF-7 cells. Moreover, levels of uc.57 mRNA were lower and BCL11A mRNA were higher in breast cancer tissues than in precancerous breast tissues. Shikonin treatment reduced tamoxifen resistance in MCF-7R cells both in vitro and in vivo, targeting uc.57/BCL11A. Fluorescence in situ hybridization and RNA immunoprecipitation analyses showed that uc.57 binds to BCL11A. Uc.57 overexpression downregulated BCL11A and reduced tamoxifen resistance in MCF-7R cells both in vitro and in vivo. BCL11A knockdown also reduced tamoxifen resistance by inhibiting PI3K/AKT and MAPK signaling pathways. It thus appears shikonin reduces tamoxifen resistance of MCF-7R breast cancer cells by inducing uc.57, which downregulates BCL11A to inhibit PI3K/AKT and MAPK signaling pathways.


INTRODUCTION
Breast cancer is the most common malignancy and second leading cause of cancer-related deaths in women worldwide [1]. Tamoxifen (TAM) has been extensively used for endocrine therapy for over four decades to improve the survival of hormone receptor-positive breast cancer patients [2]. However, TAM resistance results in breast cancer recurrence and mortality in a large number of cases [3].
Shikonin (SK) is an active ingredient isolated from the Chinese herb, Lithospermum erythrorhizon, which has been used for thousands of years in traditional Chinese medicine. It exerts anti-inflammatory [4], wound healing [5], and anti-cancer [6,7] effects. SK inhibits estrogen-dependent tumor cell growth and promotes the anti-estrogen effect of endocrine therapy in breast cancer [8,9]. It modulates mitogen-activated protein kinase (MAPK) and phosphoinositide 3-kinase (PI3K) pathways to suppress growth and survival of the malignant tumor cells [7,10]. MAPK and PI3K pathways are critical players in TAM resistance [11][12][13] and their inhibition promotes TAM sensitivity [14][15][16]. This suggests that SK would be therapeutically beneficial to reduce TAM resistance in breast cancer cells.
Long non-coding RNAs (lncRNAs) are non-protein encoding transcripts that are longer than 200 nucleotides that are involved in tumorigenesis by post-transcritpional regulation of oncogenes and tumor suppressors [17]. LncRNAs are involved in TAM resistance via numerous mechanisms in addition to estrogen receptor (ER) signaling [11,18,19]. Many lncRNAs are conserved in humans [20,21] and have potential clinical applications in anti-tumor therapy [22].
In this study, several ultra-conserved lncRNAs were selected and screened in a TAM-resistant breast cancer cell line and its parental cell line. Then, after proving SK could reduce TAM resistance, the relationship between SK-induced TAM sensitivity and lncRNAs was studied in vitro and in vivo.
We analyzed the UCSC genome database and identified BCL11A as the uc.57 target gene on human chromosome 2 ( Figure 1C). BCL11A expression was higher in breast cancer tissues than in precancerous tissues ( Figure 1D). BCL11A mRNA and protein levels were detected in MCF-7R and MCF-7 cells. BCL11A was highly expressed in MCF-7R cells than in MCF-7 cells, which was inversely correlated with uc.57 expression ( Figure 1E-1F). These data indicated that uc.57 and BCL11A were associated with TAM resistance.

Shikonin reduces TAM resistance in breast cancer cells in vitro and in vivo
The IC 50 for 4-OH-TAM in MCF-7 was 0.14 µM at 24 h (data not shown), but had no effect in MCF-7R cells. In MCF-7R cells, IC 50 for SK was 2.55 µM at 24 h (data not shown), which was slightly lower than the dose selected for this study.
MCF-7 and MCF-7R cells were separately divided into the control (DMSO), TAM (0.1 µM 4-OH-TAM), SK (2 µM SK) and TAM+SK (0.1 µM 4-OH-TAM and 2 µM SK) treatment groups. After 24 h, cell viability was tested by CCK8 assay. The results revealed that in MCF-7 cells, both SK and TAM treatment can inhibit the cell growth and the co-treatment of SK and TAM was more effective than the single treatment of SK. In MCF-7R cells, TAM+SK treatment was most effective than SK treatment alone, whereas TAM treatment had no effect on viability of MCF-7R cells (Figure 2A). What's more, TAM treatment inhibited approximately 35% cell growth of MCF-7 cell while SK+TAM treatment inhibited approximately 69% MCF-7R cell growth and 70% MCF-7 cell growth, which was about twice as much as the effect of TAM treatment on MCF-7 cells. We also treated MCF-7R cells with different doses of SK (0, 0.5, 1, 1.5, 2, 2.5, 3 µM) with or without TAM treatment (0.1 µM 4-OH-TAM) ( Figure 2B). A lower dose of SK (0, 0.5, 1 µM) with or without TAM had no inhibitory effect on MCF-7R cells. A higher dose of SK (1.5, 2, 2.5, 3 µM) became effective to MCF-7R cells while a combination of such dose of SK and TAM inhibited the cell growth even more.
In the in vivo human breast tissue-derived SCID mice model ( Figure 2C), we transplanted MCF-7R-lv-NC cells (negative control cells expressing GFP) and treated with control, TAM, SK or TAM+SK for 5 weeks. The tumor size of the TAM group was similar to control group, whereas tumor size was reduced in the SK group. Moreover, tumor volume was significantly reduced in TAM+SK group than in SK group alone ( Figure 2D, 2E). Since TAM treatment alone had no effect on MCF-7R cells, these data suggested that SK reduced TAM resistance of MCF-7R cells in the TAM+SK group.

SK decreases TAM resistance by inhibiting PI3K/AKT and MAPK pathways through uc.57/ BCL11A
We observed a dose-dependent increase in uc.57 levels in MCF-7R cells treated with 0-3 µM SK ( Figure 3A). Then, we determined expression of uc.57 and BCL11A in control, TAM, SK, and TAM+SK treated MCF-7R cells. While uc.57 levels in TAM and control groups were low, its expression increased in SK and TAM+SK treatment groups ( Figure 3B). Conversely, BCL11A mRNA levels were high in control and TAM alone treated MCF-7R cells, but decreased in the SK and TAM+SK treatment groups ( Figure 3C). Western blot analysis demonstrated that BCL11A protein expression was similar to its mRNA profile in the 4 treatment groups ( Figure 3D). In addition, both SK and TAM+SK treatments downregulated PI3K/AKT and MAPK (MEK/ERK) signaling pathways ( Figure 3D). These data suggested that SK inhibited PI3K/AKT and MAPK signaling pathways that promote TAM resistance by downregulating BCL11A, which was a result of uc.57 upregulation.

Uc.57 negatively regulates BCL11A
We performed the FISH assay of uc.57 and BCL11A RNAs in MCF-7R cells and demonstrated that they co-localize with each other, mostly in the cytoplasm ( Figure 4A). RIP assay demonstrated physical interaction between uc.57 and BCL11A in MCF-7R cells ( Figure  4B). We then compared BCL11A mRNA and protein levels in uc.57 overexpressing MCF-7R-lv-uc.57 cells ( Figure 4C) and MCF-7R-lv-NC control cells. We observed that overexpression of uc.57 downregulated BCL11A mRNA and protein levels ( Figure 4D, 4E). These data suggested that uc.57 negatively regulates BCL11A.

DISCUSSION
LncRNAs play an important role in the resistance of breast cancer to TAM [11,19]. In this study, we analyzed eight ultra-conserved lncRNAs (uc. 26, uc.41, uc.44, uc.48, uc.51, uc.57, uc.64, uc.250) in the TAMresistant breast cancer cell line, MCF-7R and its parental cell line, MCF-7. Uc.57 levels were lower in MCF-7R cells than in MCF-7 cells. Moreover, uc.57 expression was lower in breast cancer tissues than in precancerous tissues. These results suggested that uc.57 was associated with breast cancer and TAM resistance. To explore the relationship between uc.57 and TAM resistance further, we searched the UCSC genome database and identified BCL11A, which is located on human chromosome 2, as the target gene of uc.57. BCL11A was overexpressed in breast cancer tissues than in precancerous tissues. This was consistent with a previous report that BCL11A promoted triple-negative breast cancer [23]. We observed higher BCL11A mRNA and protein levels in MCF-7R cells than in MCF-7 cells. This suggested that  SK inhibits estrogen-dependent cell growth and enhances the anti-estrogen activity during endocrine therapy through ER signaling [8]. SK also enhances the therapeutic efficacy of taxol in breast cancer by inhibiting AKT and ERK [24]. PI3K/AKT and MAPK/ ERK signaling pathways are important mechanisms of TAM resistance [11][12][13]. Therefore, we postulated that SK may be beneficial to overcome TAM resistance. The results of the CCK8 assay showed that in MCF-7 cells, both TAM and SK could inhibit the growth of MCF-7R cells and co-treatment of them can be more effective. It is consist with reported literature [8]. In MCF-7R cells, TAM could not inhibit cell growth, whereas SK could. What's more, the SK and TAM co-treatment exerted the most significant inhibitory effect on MCF-7R proliferation. According to our results, SK+TAM treatment had similar inhibitory effect on MCF-7R and MCF-7 cell growth, which was about twice as much as the inhibitory effect of TAM treatment on MCF-7 cell growth. Moreover, in the human breast tissue-derived mouse models, we found that SK alone and TAM+SK treatments resulted in decreased tumor growth whereas TAM treatment resulted in similar tumor size comparable to the control group. Since single TAM treatment could not inhibit MCF-7R cell growth, the distinctive inhibitory effect of TAM+SK suggested that SK could not only directly inhibit the growth of TAM-resistant breast cancer cells but also decrease TAM resistance in TAM-resistant breast cancer cells. In our following study, we found that the efficacy of SK to overcome TAM resistance in MCF-7R cells was dose-dependent, only if SK was above the threshold concentration. The direct inhibitory effect of SK on MCF-7R cells and the reducing effect of SK on TAM exist at the same time. When SK does not inhibit cell growth, it would not decrease TAM resistance either. It suggests that both two effects work through similar mechanisms. In addition to inhibiting cell growth, SK can induce apoptosis and inhibit metastasis of tumor cells and PI3K/AKT and MAPK/ERK pathways are major mechanisms of such effects [7,25]. Since TAM resistance is strongly connected with these pathways, we postulated that their inhibition by SK played a significant role in overcoming TAM resistance.
SK treatment resulted in a dose-dependent increase in uc.57 levels in MCF-7R cells. Uc.57 was overexpressed and BCL11A was downregulated in MCF-7R cells treated with TAM+SK or SK alone, whereas control and TAM treated MCF-7R cells showed high BCL11A and low uc.57 expression. Moreover, TAM+SK treated cells showed highest uc.57 expression and lowest BCL11A expression among all groups. In addition to siRNA technology, some diazobenzene-related compounds and methylation inhibitors can modulate lncRNAs in their particular ways [26]. However, natural products and their derivatives related with lncRNAs have not yet been extensively studied. As one of the derivatives, SK was reported to inhibit inflammatory response in rheumatoid arthritis synovial fibroblasts by targeting lncRNA-NR024118 [27], but its association with lncRNAs has not been reported in cancer research. On the other hand, although lncRNAs related to TAM resistance have been previously studied [18,19], TAM sensitizing agents that target lncRNAs in cancer cells have not been reported. The findings of our study suggest that SK targets uc.57/BCL11A to overcome TAM resistance in breast cancer cells. The PI3K/AKT and MAPK signaling pathways are related to TAM resistance as well as SK function [7,11]. Both, SK and TAM+SK treatments inhibited the PI3K/AKT and MAPK signaling pathways in addition to suppressing BCL11A expression and the effect of TAM+SK treatment is more significant. These results implicated BCL11A in the SKmediated attenuation of TAM resistance in MCF-7R cells. Since lncRNAs are associated with cancer growth and progression by cancer pathways [28], activity of PI3K/ AKT and MAPK signaling pathways in TAM resistance is mediated by uc.57/BCL11A. BCL11A has a critical role in breast cancer stem and progenitor cells [23]. BCL11A knockdown inhibited PI3K/AKT and MAPK pathways in MCF-7R-si-BCL11A cells, thereby sensitizing them to TAM treatment. The activation of PI3K and MAPK pathways directly induces TAM resistance while they are mediated by upstream regulation element, such as IGFR [11,13]. The data in our study suggested that BCL11A is positively correlated with TAM resistance, and PI3K, MAPK pathways are downstream mechanisms of it.
Our study has several limitations. First, our study lacked clinical evidence because TAM-resistant specimens were not readily available. Second, the detailed mechanism of the regulation of BCL11A by uc.57 and the downstream pathways needs to be explored further. Third, we did not determine if SK decreased TAM resistance in uc.57 KO cells because uc.57 levels were very low. Finally, clinical application of SK requires more preclinical data.
In conclusion, our study demonstrates that SK treatment can decrease the resistance of breast cancer cells to TAM by upregulating uc.57. Uc.57 binds to BCL11A mRNA and suppresses its expression, which results in suppression of TAM-resistant PI3K/AKT and MAPK signaling pathways. Therefore, uc.57 and BCL11A are potential therapeutic biomarkers in TAM-resistant breast cancer patients.

Patient tissue samples
Thirty pairs of breast cancer specimens were acquired from patients that had breast cancer surgery at the First Affiliated Hospital of Nanjing Medical University, China. All tissues were frozen in liquid nitrogen immediately after surgical removal and stored at −80°C. The study was approved by the ethics and research committee of the First Affiliated Hospital of Nanjing Medical University. Informed consent was obtained from all patients included in the study.

Cell culture and chemicals
The human breast cancer cell line MCF-7 was obtained from the American Tissue Culture Collection (Manassas, VA, USA). The TAM-resistant human breast cancer cell line MCF-7R was kindly provided by Dr.

CCK-8 cell viability assay
Cell viability was tested by Cell Counting Kit-8 (Beyotime, Shanghai, China) assay in accordance with the manufacturer's instructions. 5 × 10 3 cells were seeded into each well of a 96-well plate and incubated for 24 h followed by various chemical treatments. After 24 h, 10 μl of CCK8 solution was added to each well and incubated for 2 h. Then, the plates were read in an automated microplate reader (Tecan, Grodig, Austria) at a wavelength of 450 nm (OD value). Each experiment was performed four times.
GAPDH was used as internal control. The relative expression of uc.57 and BCL11A was calculated by 2 −ΔΔCT (ΔCt = Ct value of uc.57 or BCL11A minus Ct value of GAPDH; ΔΔCt = ΔCt of the treatment group minus ΔCt of the treatment group). The experiments were performed in triplicate.

RNA immunoprecipitation (RIP) assay
RIP assay was performed according to previously published protocol [29] using Mgna RIP Kit (Milipore, Temecula, CA, USA) . Briefly, MCF-7R cells were washed with cold PBS, scraped from plates, and lysed in RIP lysis buffer. The beads and antibody complexes (with IgG or BCL11A antibody) were prepared overnight. The RNAs were immunoprecipitated with antibodies, extracted with Trizol and analyzed by qRT-PCR. Relative amounts of immunoprecipitated RNAs were normalized to the input and plotted as fold enrichment versus the IgG control.

Western blotting
Total protein lysates were prepared from the various groups of cells and their concentrations were measured with the BCA protein assay kit (ThermoScientific, Tewksbury, MA, USA). Equal amounts of protein samples (60 μg) were separated by 10% SDS-PAGE and transferred onto PVDF membranes (Bio-Rad, CA, USA) for 1 hour. The membranes were blocked in 5% skimmed milk in 1× TBST for 1 h and then incubated overnight with the following primary antibodies according to manufacturer's instructions: anti-PI3K, anti-AKT, anti-phospho-AKT, anti-ERK, antiphospho-ERK, anti-MEK and anti-phospho-MEK antibodies were all obtained from Cell Signaling Technology (Danvers, MA, USA); anti-GAPDH antibody was from Bioworld (Nanjing, China), and anti-BCL11A antibody was from Abcam (Cambridge, USA). Then, the membranes were incubated for 2 h with horseradish peroxidase conjugated anti-rabbit and anti-mouse secondary antibodies (Bioworld, Nanjing, China). The blots were developed with the SuperSignal West Femto Maximum sensitivity substrate kit (Thermo Scientific) and the chemiluminiscent signals were detected by the FluorChem E System (ProteinSimple, San Jose, CA, USA). The density of each band was quantified using Image Pro Plus 6 software. The phosphorylated protein expression was normalized to the corresponding total protein levels. The expression of other target proteins was normalized to GAPDH levels.

Cell transfection
Since MCF-7R cells showed low expression of uc.57 and high expression of BCL11A, uc.57 was overexpressed, whereas BCL11A was knocked down in this study. The lncRNA uc.57 was overexpressed in MCF-7R cells with commercial lentiviral constructs (Genepharma, Shanghai, China) to establish a new stable cell line, MCF-7R-lv-uc.57. The MCF-7R cells transfected with LV3 empty construct, MC7-7R-lv-NC was used as a negative control. The MCF-7R cells that were 40% to 50% confluent were infected with lentiviruses with the NC and uc.57 vectors at a multiplicity of 8. Next, 5 μg/ml polybrene (Genepharma, Shanghai, China) was added to the cells to increase transfection efficiency. Stable cell lines were selected in medium containing 3 μg/ml puromycin (Sigma, USA) for 1 week and analyzed for uc.57 expression. To generate BCL11A knockdown MCF-7R cells, siRNA targeting BCL11A (Genepharma, Shanghai, China) was transfected in accordance with the manufacturer's instructions. MCF-7R cells transfected with control scramble siRNA was used as control.

Tumor xenograft and tumorigenicity assay
Since MCF-7 cells are estrogen dependent, tumors were established in a novel SCID mice model with human mammary transplant tissue according to previously established protocol [30]. Five-to-seven-week-old female SCID mice (C B-17IcrCrl-scid-bgBR) were purchased from the Model Animal Research Center of Nanjing University (Nanjing, Jiangsu Province, China). Normal fresh discarded human breast tissues were obtained from elective reduction mammoplasty surgery. All protocols were conducted in accordance with the ethical guidelines of the Declaration of Helsinki and were approved by the ethics and research committee of the First Affiliated Hospital of Nanjing Medical University. Three pieces of 4mm 3 fresh breast tissue were transplanted under the skin on the left mid-dorsal flank of each SCID mouse. After one week, the implanted breast tissues were inoculated with GFP expressing MCF-7R-lv-uc.57 or MCF-7R-lv-NC cells (5 × 10 5 in 0.2 ml PBS).
Twenty SCID mice were injected with MCF-7Rlv-NC cells into human breast tissue transplants. A week later, they were randomly and equally divided into the control, TAM, SK, and TAM+SK groups and injected every alternate day with either DMSO, 20 μg/ml 4-OH-TAM, 200 μg/ml SK or 200 μg/ml SK plus 20 μg/ml 4-OH-TAM (1 ml each), respectively.
To explore the function of uc.57 in TAM resistance, we transplanted five mice with MCF-7R-con cells and another five mice with MCF-7R-lv-uc.57 cells. Both groups were treated every day with 1ml 20 μg/ml 4-OH-TAM. All the mice were sacrificed and observed grossly by whole body imaging (Illumatool 9900, Lightools Research, Encinitas; CA, USA) to determine the nature of tumors formed. Tumors were measured every week by vernier calipers, and the mice were euthanized after seven weeks. The volume of the implanted tumors was calculated by using the formula, volume = (width 2 × length)/2. The initial volume of human breast tissue was subtracted from the final volume. The animal experiments were approved by the NJMU Institutional Animal Care and Use Committee.

Statistical analysis
IBM SPSS Statistics v19 software was used for all statistical analyses. Student's t-test was used to determine significant differences between two groups. Unpaired t-test was performed to analyze qRT-PCR data for breast cancer tissues and para-cancerous tissues. A P < 0.05 was considered statistically significant.