The epigenetically downregulated factor CYGB suppresses breast cancer through inhibition of glucose metabolism

Human breast cancer cell lines (MB231, MCF7, T47D, MB468, SK-BR-3, BT549) and immortalized human mammary epithelial cell lines MCF10A and HMEC were purchased from ATCC (American Type Culture Collection). The cells were maintained in RPMI 1640 (Gibco) supplemented with 10% fetal bovine serum (FBS; Invitrogen), 100 U/mL penicillin, and 100 μg/mL streptomycin at 37 °C/5% CO₂. Stable overexpression of CYGB and GLUT1 was generated by transfection of pcDNA3.1(+)-CYGB and pReceiver-GLUT1 plasmids into MCF7 and MB231 cells using Lipofectamine 2000 (Invitrogen). Transfected cells were then screened using neomycin and puromycin for pcDNA3.1(+)-CYGB and pReceiver-GLUT1 respectively. Empty control plasmids were used to generate control lines. The stably transfected cells were pooled and used. All tissue samples were collected at the First Affiliated Hospital of Chongqing Medical University [19].

Genomic DNA was extracted from tissue and cell samples using the DNAzol® Reagent (Invitrogen) and the QIAamp® DNA Mini Kit (Qiagen). Total RNA was isolated from tissue and cell samples using TRIzol® (Invitrogen). The concentration of RNA samples was determined by spectrophotometry using NanoDrop™ 2000 (Thermo Scientific). DNA and RNA integrity were determined using gel electrophoresis. Samples were stored at − 80 °C until utilized.

BT549, MB231 and MCF7 cells were treated with 5-aza-2′- deoxycytidine (Aza; Sigma-Aldrich) and trichostatin A (TSA; Sigma-Aldrich) as previously described [20]. Cells were collected for DNA and RNA analyses following treatment.

Reverse transcription was performed using the Promega GoScript™ reverse transcriptase (Promega). 2 μL of cDNA (1200 ng/μL) were added to a 10 μL RT-PCR reaction mixture. GAPDH was used as the control. The primer sequences and specific conditions were listed in Table 1. PCR was performed using Go-Taq (Promega). Gel electrophoresis (120 V, 25 min) was performed using 2% agarose gels. Results were obtained using a BioRad Gel Doc XR+ system. Real-time PCR was performed using Maxima SYBR® Green/ROX qPCR Master Mix (Promega) according to the manufacturer’s protocol. RNA was isolated from primary breast tissues (17 paired cases of tumors and adjacent samples). The primers are listed in Table 1. Samples were amplified for 40 cycles using the 7500 Real-Time PCR System (Applied Biosystems). β-actin was used as the reference control.

MSP was conducted for 40 amplification cycles using AmpliTaq®-Gold DNA polymerase (Applied Biosystems), with annealing temperatures at 60 °C and 58 °C for methylated and unmethylated samples, respectively (Table 1). Amplicons were analyzed as previously described [21].

Cells were seeded in 96-well plates. After 24, 48, and 72 h, CellTiter 96® AQueous One Solution Reagent (MTS; Promega) was added according to the manufacturer’s protocol. Readings were taken after 1.5 h of incubation at 37 °C using the Infinite 200 Pro microplate reader (Tecan). Experiments were repeated three times independently.

Stably transfected MCF7 and MB231 cells were seeded in six-well plates at 500 cells/well. Cells were cultured for 14 days, then fixed with 4% paraformaldehyde and stained with crystal violet staining. Visible colonies were counted. Assay was repeated three times independently.

Stably transfected MCF7 and MB231 cells were plated in 6-well plates at 37 °C with 0.35% top agarose containing 1 × 10³ cells and 1.2% bottom agarose, with corresponding RPMI 1640 medium containing 10% fetal bovine serum in all agarose. Cell colonies were photographed after 3 weeks of incubation. Each experiment was repeated three times independently.

MCF7 and MB231 cells were transfected with pcDNA3.1(+)-CYGB or pcDNA3.1(+) plasmid. 48 h post-transfection, cells were collected using 0.1% trypsin, then washed with PBS and fixed in cold 70% ethanol. Fixed cells were stained with propidium iodide (PI, 50 mg/mL) at 4 °C for 30 min and analyzed with FACS Calibur™ (BD Biosciences). Data was analyzed using the CellQuest™ software (BD Biosciences). The experiment was repeated three times independently.

Annexin V- fluorescein isothiocyanate (FITC; BD Biosciences) and PI staining were performed according to the manufacturer’s protocol. Double-stained cells were analyzed using FACS Calibur™ (BD Biosciences). Data was analyzed using the CellQuest™ software (BD Biosciences).

Transfected MCF7 and MB231 cells were plated in six-well plates at 1 × 10⁵ cells/well. After 24 h, the cells were washed three times with phosphate-buffered saline (PBS) and then stained with AO/EB for 5 min. Cells were visualized using a fluorescence microscope (CTR4000B, Leica). The percentages of apoptotic cells were calculated using the formula: percentage of apoptotic cells (%) = (amount of apoptotic cells/total cells examined) × 100%.

Stably transfected MCF7 and MB231 cells were used for transwell cell migration and Matrigel™ invasion assays. Assays were carried out according to previously described protocol [22, 23].

Western blotting was done as previously described [19]. The following primary antibodies were used in this study: CYGB (Santa Cruz Biotechnology, sc-365,246), GAPDH (Cell Signaling Technology, #2118), p53 (Santa Cruz Biotechnology, sc-126), p21 (Cell Signaling Technology, #2947), β-actin (Abcam, ab8226), GLUT1 (Santa Cruz Biotechnology, sc-377,228), HXK2 (Santa Cruz Biotechnology, sc-374,091), TIGAR (Santa Cruz Biotechnology, sc-166,290).

The ATP assay was performed to quantify ATP levels according to manufacturer’s instruction (#S0026, Beyotime, Shanghai, China). Briefly, 3 × 10⁴ cells were incubated with 200 μL lysis buffer reagent, then centrifuged at 12000 g × 5 min. 20 μL supernatant was mixed with 100 μL ATP working solution and incubated at room temperature for 5 min. Luminescence signal was measured with luminometer. All conditions were assayed in triplicates.

The level of cellular lactic acid was measured using a lactic acid assay kit (A019–2, Nanjing Jiancheng Bioengineering Institute, China). Briefly, 4 × 10³ cells /well were plated in a 96-well plate and cultured overnight. Then culture medium was removed and the cells were incubated in 150 μL of lysis buffer for 1 h at 37 °C. The plate was centrifuged at 400 g for 5 min, 20 μL supernatant was collected. The cell lysates were incubated with 100 μL of detection buffer at room temperature in dark. The absorbance was measured with a microplate reader at 530 nm. Lactic acid was calculated according to manufacturer’s protocol.

Animal experiments were approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University. MB231 cells stably transfected with pcDNA3.1-CYGB or empty vector were injected subcutaneously (5 × 10⁶) into the mammary pads of 4-week-old female BALB/c nude mice [4] to generate xenograft models. Xenograft size was assessed twice a week by caliper measurement and calculated using the following equation: Volume = 0.5 × length×width². The mice were sacrificed by day 20 of injection and the xenografts were removed and weighed. 4 μm-thick paraffin sections were prepared from mice #2 for immunohistochemistry (IHC) staining. IHC staining was conducted as previously described [22] using the same antibodies as Western blot.

Protein sample preparation, iTRAQ labeling, LC-TRIPLE TOF protein identification and bioinformatics analysis are performed as previously described [24]. Briefly, cells were lysed and sonicated for protein collection. Protein concentration was determined using the Bradford method. 8-plex iTRAQ labeling was performed following the manufacturer’s protocol (AB Scienx). Labeled peptides were fractionated, followed by reverse-phase LC-TRIPLE TOF analysis.

Detailed information is presented in Additional file 1: Methods.

Statistical analyses were performed with SPSS software for Windows, version 16 (IBM). Student’s t-test, the χ2 test, and Fisher’s exact test were used to evaluate assay results and compare methylation status. For all tests, a p-value of less than 0.05 was considered statistically significant.

To investigate the role of CYGB in breast cancer, we first examined its expression in paired tumor and tumor adjacent tissues from 17 patients. Compared with that in breast tumor adjacent tissues, CYGB expression was significantly lower in breast tumor samples (Fig. 1a). The results were validated with The Cancer Genome Atlas (TCGA) dataset with 489 cases, which showed a significant decrease of CYGB expression in breast cancer that includes both invasive ductal breast carcinomas (IDBC) and invasive lobular breast carcinomas (ILBC) (Additional file 2: Figure S1A). Importantly, the higher CYGB expression was association with better patient survival, which was detected in the TCGA dataset with 1764 cases (Fig. 1b). A similar CYGB expression-patient survival association was also seen in an independent dataset from Kaplan-Meier Plotter with 2778 cases (data not shown). These results are consistent with literature and strongly suggest that CYGB is a potential tumor suppressor in breast cancer [13].

Previous studies have found that there are highly concentrated CpG sites susceptible to DNA methylation in the promoter region of human CYGB, and CYGB promoter methylation was found in several types of malignancies [12, 13, 25]. To investigate whether promoter hypermethylation was responsible for decreased CYGB expression in breast cancer, DNA samples from 195 breast tumors and 16 surgical margin tissue samples were subjected to MSP assay. Hypermethylation was detected in 170 (87%) breast tumor tissue samples, while only in 1 (6.7%) normal breast tissue sample (Table 2, Additional file 2: Figure S1B). Higher promoter methylation levels associated with lower CYGB expression was seen in the 36 breast tumor-normal tissue pairs from TCGA dataset (Fig. 1c). A similar trend of promoter methylation associated with reduced CYGB expression was also detected in breast cancer cell lines (Fig. 1d). In addition, CYGB expression was restored or strongly increased in BT549, MB231, and MCF7 cells by demethylation treatment with Aza and TSA (Fig. 1e). Taken together, these results suggest that CYGB is a potential tumor suppressor that is suppressed by promoter hypermethylation in breast cancer.

To study the cellular function of CYGB in breast cancer, MCF7 and MB231 cells were stably transfected with pcDNA3.1-CYGB or control plasmids (Fig. 2a, b) and the effect of CYGB expression restoration on cell proliferation and survival was assessed. In these cells, CYGB protein expression was increased by 25 and 7.6 fold in MCF7 and MB 231, respectively, which is comparable of the decrease fold in tumor tissues (Additional file 3: Figure S2A). Ectopic CYGB expression inhibited proliferation in both cell lines, which was detected by MTS (Fig. 2c), colony formation in soft agar (Fig. 2d), and monolayer colony formation assays (Fig. 2e, f). However, suppression of CYGB had no effect on proliferation in HMEC cells (Additional file 3: Figure S2B). Cell cycle arrest at the G0/G1 checkpoint in CYGB-overexpressing cells was detected by propidium iodide (PI) staining followed by flow cytometry assay (Fig. 3a). Higher spontaneous apoptosis rates in CYGB-overexpressing cells were detected by AO/EB staining (Fig. 3b, Additional file 4: Figure S3). Additionally, cell migration and invasion abilities were also suppressed by CYGB, which were detected by transwell assay (Fig. 3c, d). These results suggest that CYGB suppresses the malignant properties of breast cancer cells through cell cycle arrest and apoptosis induction.

To further explore the molecular mechanism by which CYGB suppresses breast cancer, we used the proteomics approach of 8-plex isobaric tags for relative and absolute quantitation (iTRAQ) to examine altered pathways in MCF7 cells with stable CYGB transfection. The results showed expression of 371 proteins was significantly changed after ectopic CYGB expression. The main altered pathways were related to glucose metabolism (metabolic, pyruvate, propanoate, and carbon metabolism, gluconeogenesis and glycolysis pathways, Fig. 4a, b), suggesting that CYGB plays a major role in glucose metabolism in breast cancer cells. To verify this hypothesis, GC-MS was used to detect metabolic alteration in CYGB overexpression cells. PCA, PLS-DA, and OPLS-DA evaluations showed that ectopic CYGB expression significantly increased and decreased levels of 21 and 16 metabolites, respectively. Again, the glucose metabolism pathways in CYGB-transfected cells were altered (Fig. 4c, d), which was shown as decreased intracellular glucose and its metabolites such as DHAP and fructose-6-phosphate (F-6-P; Fig. 4c, d). These results suggest that CYGB suppresses breast cancer through inhibiting glucose metabolism-related pathways.

A previous study showed that CYGB is able to stabilize p53 in the osteosarcoma cell line U2OS [16], and p53 is well known in regulating glucose metabolism [26, 27]. Thus, we examined if CYGB modulates the glucose metabolic pathway involved p53.. Expression of p53 and downstream target p21 was used as the p53 pathway markers. Indeed, both p53 and p21 expression was elevated in ectopic CYGB expression cells. Meanwhile, the expression of the p53-inhibiting factor MDM2 was decreased (Fig. 5a). Interestingly, the expression of TIGAR and PGM1, glucose metabolism regulators downstream of p53, was increased by 10 and 26 folds, respectively (Fig. 5a), suggesting that p53 plays a role in CYGB-mediated glucose metabolism regulation. p53 protein was stabilized by CYGB, consistent with decreased MDM2 in MCF7 cells (Fig. 5c). Altogether, these results suggest CYGB suppresses glucose metabolism at least in part through p53 in breast cancer cells have wild-type (WT) p53. However, it was noted that expression of p21, TIGAR and PGM1 was also slightly increased by 1.5–2 folds in MB231 cells that have the loss-of-function p53 R280K mutation (Fig. 5b), implying that CYGB may also regulate glucose metabolism through an alternative p53-independent mechanism.

Based on that intracellular glucose was reduced and glycolysis was suppressed by CYGB restoration, we further investigated key factors in glucose metabolism, focusing on GLUT1 (the main glucose transporter) and HXK2 (key enzyme in the first step of glucose metabolism). GLUT1 mRNA and protein was significantly decreased by CYGB in MCF7 and MB231 cells (Fig. 6a, b). There was a decrease in HXK2 at the protein level while the mRNA level remained unaffected (Fig. 6a, b). Functional assays and gene expression analysis showed similar relationship of p53 and GLUT1 or HXK2: an inverse association of expression between CYGB and either GLUT1 or HXK2 was found in the TCGA breast cancer dataset (Accessed through www.​cbioportal.​org, Additional file 5: Figure S4). This notion was supported by that GLUT1 and HXK2 expression was reduced in CYGB-transfected p53-null cells (Additional file 6: Figure S5).

The inhibition of GLUT1 expression by CYGB was well correlated to reduced cellular glucose (Fig. 4d). The cellular ATP level was reduced in ectopic CYGB expression cells (Fig. 6c), consistent with suppressed glucose metabolism. Lactate was reduced, reflecting reduction of glycolysis in ectopic CYGB-expressing cells (Fig. 6d). Taken together, these results suggest CYGB suppresses GLUT1-mediated glucose intake and metabolism pathways mediated by HXK2.

We further validated the role of GLUT1 and HXK2 in CYGB-mediated tumor suppression. Ectopic expression of GLUT1 promoted proliferation while suppressed apoptosis in CYGB-overexpressing cells (Fig. 7a, b, c). Similar rescue effect was observed with HXK2 overexpression (Additional file 7: Figure S6A, B, C). These results support that CYGB inhibits breast tumor cell growth through suppression of glucose metabolism involving GLUT1 and HXK2.

To explore the functions of CYGB in vivo, the nude mice xenograft model was used. Tumors derived from CYGB-expressing cells grew slower, which was shown as smaller tumor volume and weight than that of tumors derived from the control vector-transfected cells (Fig. 8a, b, c). IHC staining of the CYGB-expressing tumors showed lower GLUT1 and HXK2 and higher TIGAR expression (Fig. 8d), substantiating that CYGB suppresses breast cancer through the regulation of these key glucose metabolism factors. Figure 9 summarized the conclusions indicated by in vitro and in vivo results in this study.