SBF-1 exerts strong anticervical cancer effect through inducing endoplasmic reticulum stress-associated cell death via targeting sarco/endoplasmic reticulum Ca2+-ATPase 2

Cervical cancer is one of the most common carcinomas in the genital system. In the present study, we report that SBF-1, a synthetic steroidal glycoside, has a strong antigrowth activity against human cervical cancer cells in vitro and in vivo. SBF-1 suppressed the growth, migration and colony formation of HeLa cells. In addition, severe endoplasmic reticulum (ER) stress was triggered by SBF-1, and 4-phenyl-butyric acid, a chemical chaperone, partially reversed SBF-1-induced cell death. To uncover the target protein of SBF-1, the compound was labeled with biotin. The biotin-labeled SBF-1 bound to sarco/ER Ca2+-ATPase 2 (SERCA2) and colocalized with SERCA2 in HeLa cells. Moreover, SBF-1 inhibited SERCA activity, depleted ER Ca2+ and increased cytosolic Ca2+ levels. 1,2-Bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid, a chelator of Ca2+, partially blocked SBF-1-induced ER stress and growth inhibition. Importantly, knockdown of SERCA2 increased the sensitivity of HeLa cells to SBF-1-induced ER stress and cell death, whereas overexpression of SERCA2 decreased this sensitivity. Furthermore, SBF-1 induced growth suppression and apoptosis in HeLa xenografts, which is closely related to the induction of ER stress and inhibition of SERCA activity. Finally, SERCA2 expression was elevated in human cervical cancer tissues (n=299) and lymph node metastasis (n=8), as compared with normal cervix tissues (n=23), with a positive correlation with clinical stages. In all, these results suggest that SBF-1 disrupts Ca2+ homeostasis and causes ER stress-associated cell death through directly binding to SERCA2 and inhibiting SERCA activity. Our findings also indicate that SERCA2 is a potential therapeutic target for human cervical cancer.

The endoplasmic reticulum (ER) is a membranous system that is essential for the function and survival of mammalian cells. 1 To achieve an optimum state for survival, a cell may undergo various pathways in the ER, such as ones that regulate protein folding, posttranslational modifications, lipid and steroid synthesis, gene expression, cellular metabolism and calcium signaling. When the pathways are disturbed, ER functions become overwhelmed and the accumulation of misfolded proteins within the ER lumen ultimately leads to ER stress and initiates the unfolded protein response (UPR) to restore ER proteostasis. 2,3 The UPR is generally considered to be the transcriptional induction of molecular chaperones in response to ER stress. 4 Three ER stress arms have been identified so far: protein kinase RNA-like ER kinase (PERK), inositolrequiring enzyme-1 (IRE1) and activating transcription factor-6 (ATF6). 3,5,6 These three ER stress arms can act as coregulators of most targets to ensure the redundancy and robustness of this adaptive response. 3,7 If the response fails or prolongs, apoptotic cell death ensues. 3,8 Calcium homeostasis is involved in a multitude of signaling in ER, in which calcium is actively accumulated by sarco/ER Ca 2+ -ATPase (SERCA) transport ATPases. 9,10 As SERCAdependent calcium transport is the only calcium uptake mechanism in this organelle, the regulation of SERCA function by the cell constitutes a key mechanism to adjust calcium homeostasis in the ER depending on the cell type and its state of differentiation. 10 The SERCA pump is encoded by a family of three genes, SERCA1, 2 and 3, which are highly conserved but localized on different chromosomes. 11 At present, more than 10 different SERCA isoforms have been discovered in this family. These isoforms exhibit both tissue and developmental specificity, suggesting that they contribute to unique physiological properties of the tissue in which they are expressed. 11,12 SERCA expression levels can undergo significant changes during cell differentiation or tumorigenesis, leading to modified ER calcium storage. 10 For example, expression levels of SERCA3, a lower calcium affinity calcium pump, are highly variable. In several cell systems, SERCA3 expression is selectively induced during differentiation, whereas during tumorigenesis SERCA3 expression is decreased. 10,[13][14][15][16][17][18] Diverse levels of SERCA2 are also correlated with tumorigenesis. [19][20][21] Cervical cancer is the third most common carcinoma in women worldwide. [22][23][24] The American Cancer Society estimates that 12 360 new cases and 4020 deaths of cervical cancer are projected to occur in the United States in 2014. 22 Despite more and more efforts are being made to improve therapy, a significant proportion of women still die from recurrence and chemoresistance, 24 which is a hallmark of cancer cells. 25,26 Searching for an efficient agent is still a requirement for the therapy of the cancer. Among such efforts, induction of ER stress by targeting SERCA might be a potential therapeutic strategy to treat apoptosis-resistant cancer cells. 27,28 Our previous study identified that SBF-1, a synthetic steroidal glycoside, had a very strong antitumor activity in various cancer types. 29,30 Herein, we report a new property of SBF-1 for characterizing its anticancer activity as a SERCA inhibitor that directly binds to SERCA2. In human cervical cancer cells, ER Ca 2+ homeostasis is disrupted and ER stress-mediated cell death is induced by SBF-1 both in vitro and in vivo.
SBF-1 binds to SERCA2 and increases the intracellular Ca 2+ levels. To find out binding proteins of SBF-1, SBF-1 was labeled with biotin (Supplementary Figure S2a). The biotin conjugate of SBF-1 (biotin-SBF-1) still showed a strong antigrowth activity (IC 50 , 436.63 ± 48.79 nM) despite an obvious decrease as compared with SBF-1 (IC 50 , 45.66 ± 6.27 nM; Supplementary Figure S2b). Biotin-SBF-1 was then incubated with HeLa whole-cell lysates and streptavidin-conjugated sepharose beads in the presence or absence of 10-to 20-fold excess of SBF-1. The proteins bound to the beads were separated with SDS-PAGE and the bands between 100 and 130 kDa were cut and analyzed with liquid chromatography-mass spectrometry (LC/MS). Sarco/ ER Ca 2+ -ATPase 2 (SERCA2), the most abundant SERCA isoform in HeLa cells (Supplementary Figure S3a), was identified to be a binding protein of SBF-1 (Figures 2a and b), and biotin-SBF-1 colocalized with SERCA2 in HeLa cells (Figure 2c). Furthermore, SERCA activity of HeLa cells was significantly suppressed by both 10 and 100 nM SBF-1 ( Figure 2d) and the protein level of SERCA2 was compensatorily increased (Supplementary Figures S3b and c), whereas the mRNA level of SERCA2 was not changed (Supplementary Figure S3d). Moreover, ER Ca 2+ was depleted ( Figure 2e) and intracellular Ca 2+ levels were significantly increased by exposure to 100 nM SBF-1 in both a concentration-and time-dependent manner (Figure 3a and Supplementary Figure S4). BAPTA (1,2-Bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid), a cell-permeant Ca 2+ chelator widely used in Ca 2+ -related study, 27,32 partially reversed SBF-1-induced proliferation suppression ( Figure 3b) and markedly attenuated ER stress induced by SBF-1 (Figures 3c and d) in HeLa cells.
The SERCA2 level controls the sensitivity of HeLa cells to SBF-1. To further demonstrate the role of SERCA2 in the antitumor effects of SBF-1, we stably knocked down SERCA2 in HeLa cells (Supplementary Figures S5a and b), which had no influences on the growth of HeLa cells under normal culture conditions (Supplementary Figure S5c). However, the SERCA2 silence greatly enhanced SBF-1-induced repression of cell growth ( Figure 4a) and migration ( Figure 4b). Then, we confirmed that the protein levels of ATF6α (p50) and phospho-eIF2α Ser51 (Figure 4c), and the mRNA levels of CHOP and GRP78 (Supplementary Figure S6) in HeLa cells with stable SRECA2 knockdown were increased more significantly after exposure to SBF-1, compared with cells with stable NC lentivirus infection. In addition, SERCA2b overexpression had no influences on the growth of HeLa cells under normal culture conditions (Supplementary Figure S7), but partially reduced SBF-1-induced proliferation suppression ( Figure 4d). The increase in protein levels of CHOP by SBF-1 was almost completely blocked in HeLa cells transfected with hSERCA2b as compared with cells transfected with mock vector (Figure 4e). The above results indicate that SBF-1 suppresses the HeLa cell growth and migration depending on the activity and level of SERCA2.
SBF-1 inhibits tumor growth at a very low dose in HeLa xenografts with decreased SERCA activity and increased ER stress and apoptosis. To evaluate the antitumor effects of SBF-1 in vivo, we daily intraperitoneally injected vehicle (0.1% dimethyl sulfoxide (DMSO) in phosphate-buffered saline (PBS)) or 5 μg/kg SBF-1 to tumor-bearing nude mice. Changes in tumor volumes and body weight were measured and recorded every 2 days. The growth of HeLa xenografts was significantly inhibited after 8 days of administration of SBF-1, in comparison with vehicle-treated group (Figure 5a). The tumor mass and the percentage of tumor weight in body weight was significantly decreased by SBF-1 treatment (Figure 5b), whereas body weight was not influenced (Supplementary Figure S8). Furthermore, the expression of   SERCA2 positively correlates with the malignant progress of human cervical cancer. To evaluate the relationship between SERCA2 and malignance of cervical cancer, the protein level of SERCA2 was detected by immunohistochemistry in tissue arrays of human cervical cancer. Modest SERCA2 expression was detected in normal cervix tissues (n = 9) and cancer adjacent normal cervix tissue (NAT, n = 14). Comparably, a significantly elevated SERCA2 expression was detected in malignant tumor tissues (carcinoma, n = 299) and lymph node metastasis (metastasis, These results indicate that SERCA2 has a positive correlation with the malignance of human cervical cancer and could be a therapeutic target for cervical cancer therapy.

Discussion
Cervical cancer is the most common life-threatening disease among women worldwide, and the therapeutic effects are bleak because of recurrence and chemoresistance. 24 Thus, it is significant to search for effective targets for the treatment of cervical cancer. In the present study, we found that low concentrations of SBF-1 (10 and 100 nM) significantly inhibited the growth, migration and colony formation of HeLa  1a-d). ER stress is a well-known reason for cell death. 33 SBF-1 induced ER stress and UPR responses in a concentration-and time-dependent manner, and the chemical chaperone PBA partially reversed the antigrowth activity of SBF-1 (Figures 1e-i). Compared with slowly dividing cells, rapidly dividing cancer cells are more sensitive to ER stress inducers, because of the high levels of protein folding. 2 Compared with naive T lymphocytes and ECV304 cells (slowly dividing cells), Jurkat and HeLa cells (rapidly dividing cancer cells) were more sensitive to SBF-1 (Figure 1a and Supplementary Figure S11), suggesting that SBF-1 antigrowth effects is not due to general cytotoxicity. High concentrations of Ca 2+ in ER lumen are essential for ER folding capacity. 34 SERCA is a pump to transport Ca 2+ from the cytoplasm to ER lumen, 11,12 and inhibition of SERCA activity impairs ER Ca 2+ homeostasis and causes severe ER stress. 34-37 SERCA2, the major SERCA isoform of HeLa cells, was found to bind with SBF-1, and SERCA activity was suppressed by SBF-1 (Figures 2a-d). Accordingly, SBF-1 depleted ER Ca 2+ and increased intracellular Ca 2+ levels, and Ca 2+ chelator BAPTA partially reversed SBF-1-induced ER stress and growth suppression (Figures 2e and 3). Interestingly, the baseline Ca 2+ levels in SBF-1-treated cells were obviously higher than those in TG-treated cells (Figure 2e), suggesting the differences between SBF-1 and TG. It is well known that TG is a specific SERCA inhibitor, which depletes ER Ca 2+ quickly. 27 After incubation with TG for 48 h, ER Ca 2+ was almost completely depleted and exhausted. However, there was still abundant cytosolic Ca 2+ , and a small quantity of ER Ca 2+ in SBF-treated cells. As shown in Figure 2c, the colocalization rate of SBF-1 and SERCA2 was~60%, indicating that SBF-1 also bound to other proteins, which was supported by the competitive binding assay (Figure 2a). LC/MS identified many other proteins, including plasma membrane calcium-transporting ATPase 1 (PMCA1, data not shown). Moreover, SBF-1 started to increase the baseline cytosolic Ca 2+ levels after 3-h incubation (Supplementary Figure S4) and did not induce cytosolic Ca 2+ exhaustion, which was quite different from TG. Combined with these data, we supposed that SBF-1 might inhibit the activity of SERCA and Ca 2+ pumps in the plasma membrane (such as PMCA1), resulting in Ca 2+ release from ER and accumulation in the cytoplasm, which might explain the strong antigrowth effects of SBF-1.
It has been reported that impairment of SERCA2 levels increase the sensitivity of cells to ER stress. 19,34,38,39 Coincidentally, knockdown of SERCA2 increased the sensitivity of HeLa cells to SBF-1, whereas overexpression of SERCA2 decreased the sensitivity of HeLa cells to SBF-1 (Figure 4 and Supplementary Figure S7). Furthermore, in vivo experiments indicated that a very low dose of SBF-1 (5 μg/kg) markedly inhibited the growth of HeLa xenografts, with repression of SERCA activity and induction of abundant apoptosis and severe ER stress ( Figure 5 and Supplementary Figure S9).
Aberrant SERCA expression has been reported to be relevant with the susceptibility and progression of colon The protein levels of phosphorylated eIF2α Ser51 , total eIF2α and CHOP were detected by immunoblotting (left panel), and the relative band intensity was analyzed with Image J software and normalized with α-tubulin (middle and right panels). α-Tubulin was performed as a loading control. Images shown were representative of more than three independent experiments. Data were means ± S.D. of more than three independent experiments. Scale bars, 100 μm. *Po0.05, **Po0.01 versus 0 nM SBF-1 (a) or vehicle (b-d) SBF-1 induces ER stress-associated cell death via targeting SERCA2 W Li et al cancer, lung cancer and liposarcoma. [13][14][15][16][17][18][19][20][21] However, the correlation of SERCA2 with the malignance of cervical cancer is not reported before. The following experiments indicated that elevated SERCA2 expression was detected in malignant cervical carcinomas and lymph metastasis, with a positive correlation with the clinical stages of malignant cervical carcinomas (Figures 6a-c).
In all, as illustrated in Figure 6d, when cells are exposed to SBF-1, SERCA activity is suppressed, inducing depletion of ER Ca 2+ and increase of cytosolic Ca 2+ , which disturbs ER folding capacity and increases unfolded and misfolded proteins, activating the signaling pathways of UPR response and causing ER stress-associated cell death. This study suggests that SERCA2 could be a therapeutic target in human cervical cancer and SBF-1 might be a novel SERCA inhibitor to induce cell death.

Materials and Methods
Materials. SBF-1 is a synthetic steroidal glycoside. 30    MTT assay. A total of 2 × 10 3 cells were seeded into 96-well plates, and incubated with various concentrations of SBF-1 at 37°C for indicated time periods. Four hours before measurement, 20 μl per well MTT solutions (4 mg/ml in PBS) were added into the 96-well plates. After incubation at 37°C for 4 h, the supernatants were discarded and 200 μl per well DMSO was added. Then, the plates were measured under an FL × 800 Fluorescence Microplate Reader (BioTek, Winooski, VT, USA) at 570 nm. The optical density values of the wells without SBF-1 were considered as 100% cell growth rate.
Migration and colony formation assay. A total of 2 × 10 5 cells per well were seeded into 6-well plates and incubated with DMSO or various concentrations of SBF-1 for 48 h. For migration assay, the cells were washed once with serum-free DMEM and diluted to a concentration of 1 × 10 5 per ml, and 0.1 ml cell suspension was added to the inner room of transwell (Costar) pretreated with 10 μg/ml fibronectin. Next, 0.6 ml DMEM supplanted with 20% FBS was added into the wells of 24-well plates and the cells were incubated at 37°C for another 24 h. Then, the cells were fixed with 4% paraformaldehyde (PFA) at room temperature for 30 min and stained with crystal violet (Sigma; 1% in distilled water) for 10 min. After washing two times with distilled water, migrated cells were counted with an optical microscope (Olympus, Shinjuku, Tokyo, Japan; five fields per transwell). For colony formation assay, 1 × 10 3 cells per well were seeded into 6-well plates and incubated at 37°C for another 2 weeks. Then, the cells were fixed with 4% paraformaldehyde at room temperature for 30 min and stained with crystal violet for 10 min (1% in distilled water). After washing two times with distilled water, photos were taken with a digital camera (Olympus), and the number of colonies, the cell number of which exceeded 80, were counted with an optical microscope (Olympus).
Immunoblotting. In brief, cells were collected and washed with ice-cold PBS two times before being lysed in radio immunoprecipitation assay (RIPA) lysis buffer (Beyotime, Haimen, China; 50 mM Tris (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mM phenylmethylsulfonylfluoride, 0.15 U/ml aprotinin and 1 mg/ml pepstatin). Whole-cell lysates were collected and proteins were resolved by SDS-PAGE and were then electrotransferred onto polyvinylidene fluoride membranes. Then, the membranes were blocked in 5% bovine serum albumin (Sigma) at room temperature for 1 h, and the blots were incubated with primary antibody at 4°C overnight and with secondary antibody at room temperature for 1 h. After extensive washing, the blots were developed with a chemiluminescence assay system (Cell Signaling Technology) and exposed to films (Kodak, Rochester, NY, USA) for appropriate time periods. The densitometry of immunoblots was quantified with Image J software (NIH, Bethesda, MD, USA) and normalized with loading controls.  When cells are exposed to SBF-1, SERCA activity is suppressed, inducing depletion of ER Ca 2+ and increase of cytosolic Ca 2+ , which disturbs ER folding capacity and increases unfolded and misfolded proteins, activating the signaling pathways of UPR response and causing ER stress-associated cell death. ATF6N, N terminal of ATF6α Competitive binding assay. HeLa whole-cell lysates were, respectively, incubated with 10 μM biotin, 10 μM biotin-SBF-1, 10 μM biotin-SBF-1 plus 100 μM SBF-1 or 10 μM biotin-SBF-1 plus 200 μM SBF-1 and 50 μl streptavidin-conjugated sepharose beads (GE Healthcare) at 4°C overnight for 12 h. Then, the mixture was centrifuged at 4000 × g for 1 min to obtain the precipitation. After washing five times with RIPA lysis buffer, the beads were boiled in 2 × loading buffer (100 mM Tris-HCl (pH 6.8), 4% SDS, 1% bromphenol blue, 20% glycerol and 2% β-mercaptoethanol).
After centrifugation at 4000 × g for 2 min, the supernatant were collected and separated with SDS-PAGE.
Silver staining and LC/MS. SDS-PAGE gels were fixed (40% ethanol, 10% acetic acid in distilled water) at room temperature (RT) overnight. Then, the gels were washed once with distilled water and sensitized (150 ml ethanol, 34 g anhydrous sodium acetate, 1.57 g anhydrous sodium thiosulfate and metered to 500 ml with distilled water) RT for 30 min. After washing two times, the gels were silver-stained (0.125 g silver nitrate and metered to 50 ml) at RT for 20 min. Next, the gels were washed once with distilled water and colorized with formaldehyde (20 μl in 50 ml distilled water) for appropriate time periods. Finally, the reaction was stopped by adding ethylene diamine tetraacetic acid (0.1% in distilled water). Interested bands were cut and analyzed with LC/MS, which was performed by Shanghai Institute of Biochemistry Proteomics Center (Shanghai, China).
Immunofluorescence. Cells incubated with 0.1% DMSO or 10 μM biotin-SBF-1 for 1 h were fixed in 4% paraformaldehyde (pH 7.4) for 30 min. Then, the cells were softly washed two times with PBS, and incubated with 0.5% Triton X-100 for 30 min. Thereafter, cells were softly washed two times with PBS and blocked with 3% bovine serum albumin for 1 h. Cells were incubated with primary antibodies (anti-SERCA2 and anti-biotin antibodies, respectively) at 4°C overnight. After washing two times, the cells were incubated with secondary antibodies for 1 h and the nucleus was stained with 4, 6-diamidino-2-phenylindole (DAPI) for 1 min. The fluorescent signals were detected with a mercury lamp (Olympus U-RFL-T) and analyzed by Image Pro Plus 6.0 software (Media Cybernetics Inc., Rockville, MD, USA). Plasmids and lentivirus. Construction of pcDNA3.1(+)-hSERCA2b was performed as before. 19 Construction of sh-SERCA2 lentivirus was performed by Neuron Biotechnology (Shanghai, China). The vector of the shRNA was pLKD-CMV-G&PR-U6 and the map was shown in Supplementary Figure S1; the sequence was 5′-CAAAGUUCCUGCUGAUAUA-3′. Control lentivirus particles were provided by Neuron Biotechnology and the sequence was 5′-UUCUC CGAACGUGUCACGU-3′. For transient transfection, cells were seeded into six-well plates and allowed to grow to 40% confluence. Then, 2 μg pcDNA3.1(+) or pcDNA3.1(+)-hSERCA2b together with 2 μl Lipofectamine 2000 (Life Technologies) were added into the cells, respectively. Twenty-four hours after transfection, the cells were incubated with DMSO or various concentrations of SBF-1 for another 48 h. For lentivirus infection, 1 × 10 5 cells in a well were seed into 12-well plates. Twelve hours later, 1 × 10 6 plaque-forming units of control lentivirus particles or sh-SERCA2 lentivirus particles were added to the cells, respectively. Three days after infection, the infected cells were incubated with 8 μg/ml puromycin (P8833; Sigma) for another 2 weeks. Then, the cells were harvested and the protein levels of SERCA2 were detected with immunoblotting. Cells stably infected with control or sh-SERCA2 lentivirus were incubated with DMSO at various concentrations of SBF-1 for 48 h. Then, the cells were collected for the following experiments.

Detection of cytosolic
In vivo tumor xenografts model. One million HeLa cells (in 0.1 ml PBS) were subcutaneously injected into the right flanks of nude mice. Seven days later, all the mice formed visible tumors. Then, the mice were distributed into two groups according to tumor volumes. Vehicle (0.1% DMSO in PBS, n = 5) and 5 μg/kg SBF-1 (n = 6) were daily intraperitoneally injected to the tumor-bearing mice for 12 days. Body weight and tumor volumes were measured and recorded every 2 days. Long diameter (L) and short diameter (S) of a tumor were measured with a vernier caliper and the tumor volume was calculated as follows: L × S 2 /2. Twelve days after injection, the mice were killed and tumors were separated. Tumor weight was measured and tumor sections were fixed in formalin. The rest of the sections were frozen in liquid nitrogen and stored at − 80°C.
Immunohistochemistry and TUNEL staining. Paraffin sections of xenografts were cut into 5-μm-thick sections. Then, the sections were infused in xylene (10 min per time for three times), followed by infusion in 90% ethanol (3 min), 75% ethanol (3 min), 50% ethanol (3 min) and ddH 2 O (3 min). For immunohistochemistry, the sections were boiled in antigen restoration solution (P0081; Beyotime) for 3 min and cooled naturally. Cooled sections were incubated with 3% H 2 O 2 at RT for 10 min and blotted with 3% horse serum for 1 h. Next, the sections were incubated with primary antibodies at 4°C overnight. The following steps were in accordance to the protocols provided by the manufacturer of GTVision III Detection System/Mo&Rb (GK500705; Gene Tech Company Limited, Shanghai, China). The nucleus was stained with hematoxylin. TUNEL staining was performed according to the instructions provided by the manufacturer of DeadEnd Fluorometric TUNEL System (TB235; Promega, Madison, WI, USA). The nucleus was stained with DAPI (Beyotime) for 1 min and the percentage of TUNEL-positive cells was calculated with a mercury lamp (Olympus U-RFL-T).
Statistical analysis. Data are expressed as means ± S.D. One-way ANOVA was used to evaluate the differences among more than three groups. The Student's t-test was used to evaluate the difference between two groups. Po0.05 was considered to be significant.

Conflict of Interest
The authors declare no conflict of interest.