Background
Breast cancer is the most common female cancer and is the second leading cause of cancer-related deaths among females worldwide [
1]. Estrogen receptor (ER)-positive breast cancers account for 60-70% of all breast cancers, but the remaining 30-40% of breast cancers are ER-negative breast tumors, which do not express ER, a protein with both prognostic and predictive value [
2,
3]. Unfortunately, ER-negative breast cancers are resistant to endocrine therapy, which reduces recurrence and mortality rates whether chemotherapy is given or not [
3,
4]. Therefore, ER-negative breast cancers recur and metastasize more readily, and consequently, patients with this cancer type have a worse prognosis and shorter survival rates compared with those with ER-positive breast cancers. This underscores the importance of the identification of new prognostic markers and additional drug targets for this class of breast cancer.
Serine plays an essential role in the synthesis of biomolecules that support cell proliferation. Recent evidence implies that hyperactivation of serine contributes to tumorigenesis [
5]. Within cells, serine is synthesized through a three-step reaction. First, 3-phosphoglycerate is oxidized into phosphohydroxypyruvate (pPYR) by phosphoglycerate dehydrogenase (PHGDH). Successively, phosphohydroxypyruvate (pPYR) is catalyzed by phosphoserine aminotransferase (PSAT1) to produce phosphoserine (pSER), which is then dephosphorylated by 1-3-phosphoserine phosphatase (PSPH) to form serine. Two recent studies have reported that the gene encoding phosphoglycerate dehydrogenase (PHGDH) is amplified in a significant subset of human tumors, which supports the idea that metabolic reprogramming occurs as the result of genomic modifications of metabolic enzymes, which independently contribute to tumorigenesis [
6,
7]. PSAT1 was found to be up-regulated in colon cancer, esophageal squamous cell carcinoma (ESCC) and non-small cell lung cancer (NSCLC), and has been shown to enhance cell proliferation, metastasis and chemoresistance, which all contribute to a poor prognosis [
8‐
11]. However, the expression and underlying mechanism of PSAT1 in ER-negative breast cancer are not well understood. These observations have prompted us to speculate the role of PSAT1 in the initiation and development of ER-negative breast cancer.
Activating transcription factor 4 (ATF4) is a member of the cyclic adenosine monophosphate responsive element-binding (CREB) protein family, which has been reported to be a potent stress-response gene that is expressed in a wide variety of tumors [
12,
13]. ATF4 can protect tumor cells against stresses and a range of cancer therapeutic agents via the regulation of cellular adaptation to adverse circumstances [
14‐
19]. Previous studies have shown that ATF4 overexpression exists in many tumors, which suggests that it may play an important role in tumor formation, progression and metastasis [
17,
19‐
22].
In the current study, PSAT1 was significantly up-regulated in ER-negative breast cancer and was correlated with a poor patient prognosis. Moreover, PSAT1 was found to be regulated by ATF4, which then activated the GSK-3β/β-catenin pathway. This resulted in the enhancement of cyclin D1 expression and the promotion of cell proliferation.
Methods
Patients and tissue specimens
The archival material used in this study was obtained from the Department of Pathology at the Harbin Medical University Cancer Hospital, and included tissues from 297 patients with histologically confirmed ER-negative breast cancer (Additional file
1) and 112 matched normal tissue samples from patients who presented from 2006 to 2007. For the extraction of protein and RNA, fresh tissues from individuals with ER-negative breast cancer and normal controls were collected and stored at −80 °C immediately after resection [
23]. None of the patients received adjuvant chemotherapy, immunotherapy, or radiotherapy before surgery, and the patients with recurrent tumors, metastatic disease, bilateral tumors, or other previous tumors were excluded. Pathologists diagnostically examined tumors for confirmation of ER-negative breast cancer and benign breast diseases. After surgery, adjuvant systemic therapy was determined according to the National Comprehensive Cancer Network (NCCN) guidelines. This study was approved by the Ethical Committees of Harbin Medical University. Written informed consent was obtained from all subjects who participated in this study.
Cell culture
MDA-MB-468, MDA-MB-231, MDA-MB-453, BT-549, HCC70, Hs578T and MCF-7 cells were cultured in RPMI-1640 or DMEM (Gibco, Carlsbad, CA, USA). All media were supplemented with 10% fetal bovine serum (FBS). MCF-10A cells were cultured as described previously [
24]. All cells were incubated at 37 °C in humidified air containing 5% CO
2.
Plasmid, Lentivirus production and infection
Regarding the knock down of PSAT1, two human PSAT1 targeted RNAi (RNAi#1: TTCCAAGTTTGGTGTGATT; RNAi#2: ACTCAGTGTTGTTAGAGAT) sequences were obtained from GeneChem Co. Ltd. (Shanghai, China). As a control, scrambled versions of these sequences were used. The sequences shown above were inserted into the GV248 vector plasmid. For the overexpression of PSAT1, full-length human PSAT1 cDNA was cloned into the pLVX-puro vector. Lentiviral particles were constructed and packaged by Shanghai GeneChem Co. Ltd. Briefly, the cells were infected with lentivirus to generate stable cell lines. After 24 h, the cells were transferred to medium containing 4 μg/ml puromycin and were cultured for 3 days.
Interfering RNA and transfection
ATF4 siRNAs and scrambled negative control siRNA were purchased from Invitrogen (Invitrogen, CA, USA). The siRNAs were transfected into cells using Lipofectamine 3000 (Invitrogen, CA) according to the manufacturer’s protocol. The sequences are as follows: ATF4-RNAi#1: 5′-CUGCUUACGUUGCCAUGAUTTAUCAUGGCAACGUAAGCAGTT-3′; ATF4-RNAi#2: 5′-CCCUUCAGAUAAUGAUAGUTTACUAUCAUUAUCUGAAGGGTT-3′; Scrambled-siRNA:5′-UUCUCCGAACGUGUCACGUTTACGUGACACGUUCGGAGAATT-3′.
Cell proliferation assays
A cell proliferation assay was performed with a CCK-8 kit (Beyotime Institute of Biotechnology, Shanghai, China) according to the manufacturer’s instructions. Briefly, 2 × 103 cells were plated in each well of a 96-well plates and were cultured overnight. According to the instructions, Cell Counting Kit-8 (CCK-8) reagent was added at 24, 48, 72 or 96 h and incubated at 37 °C for 1 h. Each assay was independently repeated three times in triplicate.
Cells were plated into a 6-well plate and cultured in media containing 10% FBS for 14 days. Colonies were fixed in methanol for 30 min, and 500 μl 0.5% crystal violet was added (Sigma, St. Louis, MO, USA) to each well for 30 min for visualization and counting.
Migration and invasion assays
Cells in serum-free media were placed into the upper chamber of an insert for the migration assays (8-μm pore size, Millipore), while for the invasion assays, the cells were seeded on plates coated with Matrigel (Sigma-Aldrich, USA). Medium containing 10% FBS was added to the lower chamber. After incubation at 37 °C for 12 h(Migration) or 24 h(Invasion), non-invading cells that remained in the top chambers were removed with a cotton swab, and the cells that had migrated to the underside of the membrane were fixed in 100% methanol for 30 min, air-dried, stained with 0.5% crystal violet, imaged, and counted under a light microscope.
RNA preparation and qRT-PCR
Total RNA was extracted using TRIzol reagent according to the manufacturer’s protocol (Invitrogen, Beijing, China), and cDNA was synthesized using a PrimeScript RT reagent Kit with gDNA Eraser (Takara Bio, Otsu, Japan). mRNA expression was examined by real-time PCR using FastStart Universal SYBR Green Master (Roche, Mannheim, Germany) with gene-specific primers and an ABI 7500 Fast Real-time PCR Detection System (Applied Biosystems, Foster City, CA, USA). The results were normalized to the expression of β-actin. The sequences of the primers used were as follows: PSAT1-F: 5′-GTCCAGTGGAGCCCCAAAA-3′; PSAT1-R: 5′-TGCCTCCCACAGACCTATGC-3′; CCND1-F: 5′-GCTGCGAAGTGGAAACCATC-3′; CCND1-R: 5′-CCTCCTTCTGCACACATTTGAA-3′; β-actin-F: 5′-CAACCGCGAGAAGATGACC-3′; β-actin-R: 5′-ATCACGATGCCAGTGGTACG-3′.
Flow cytometry analysis
Cells were seeded in 6-well plates, and after 24 h, the cells were harvested and washed twice with cold PBS. For the cell cycle analysis, the cells were fixed in ice-cold 75% ethanol overnight at 4 °C. After fixation, the cells were washed and resuspended twice in PBS and were then incubated with propidium iodide (BD Bioscience, San Jose, CA, USA) and RNase for 30 min at room temperature. For the cell apoptosis analysis, the cells were stained with PE Annexin V and 7-AAD (BD Bioscience, San Jose, CA, USA) for 15 min at room temperature. The cells were then analyzed using a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA, USA).
Western blotting analysis
Cultured cells or frozen tissue samples were harvested and lysed in RIPA buffer consisting of a 1% protease inhibitor mixture. A western blotting assay was performed as previously described [
25]. The following antibodies were used: anti-PSAT1(Abcam, Cambridge, MA, USA, 1:1000), anti-Cyclin D1 (Abcam, Cambridge, MA, USA, 1:1000), anti-p-GSK-3β (Cell Signaling Technology, Beverly, MA, USA, 1:1000), anti-GSK-3β (Wanleibio, Shenyang, China,1:500), anti-β-catenin (Wanleibio, Shenyang, China, 1:500), and anti-ATF4 (Cell Signaling Technology, Beverly, MA, USA, 1:1000); anti-β-Tubulin (Santa Cruz Biotechnology, CA, USA, 1:1000) as an internal control.
Immunohistochemistry (IHC)
A tissue microarray (TMA) that included samples from 297 consecutive patients with histologically confirmed estrogen receptor-negative breast cancer and 112 controls was generated according to a previously described method [
26]. The tissue sections were dried at 70 °C for 3 h for deparaffinization and hydration. Subsequently, the sections were washed with phosphate-buffered saline (PBS; 3 × 3 min). The washed sections were treated with 3% H
2O
2 in the dark for 5 to 20 min. After washing in distilled water, the sections were again washed with PBS (3 × 5 min). Antigen retrieval was performed in citrate buffer (pH 6.0) at 100 °C for 10 min. Each section was incubated with the polyclonal primary rabbit antibody against PSAT1 at a 1:100 dilutions (Abcam, Cambridge, MA, USA) overnight at 4 °C. After washing with PBS (3 × 5 min), each section was further incubated with an anti-rabbit secondary antibody (1:200; Abcam, Cambridge, MA, USA) at room temperature for 30 min. After another wash in PBS (3 × 5 min), each section was immersed in 500 μl of diaminobenzidine (DAB) working solution at room temperature for 3 to 10 min. Finally, the slides were counterstained with hematoxylin and mounted in crystal mount medium. PSAT1 expression was analyzed and scored independently by two observers based on the intensity and the distribution of positively stained tumor cells, which were demarcated by yellow particles observed in the cytoplasm. The PSAT1 staining index was classified into four groups: level 0 (no staining), level 1 (0-20% of tumor cells stained), level 2 (20-50% of tumor cells stained) and level 3 (>50% of tumor cells stained). Overall expression was then graded as either negative expression (level 0) or positive expression (levels 1-3) [
8].
Animal experiments
Animal experiments were approved by the Medical Experimental Animal Care Commission of Harbin Medical University. BALB/C-nu/nu nude mice were obtained from Beijing Vital River Laboratory Animal Technology Company. Approximately 5 × 106 cells (HCC70-NC or HCC70-KD) or 8 × 106 cells (BT-549-Vector or BT-549-PSAT1) in 200 μl of serum-free medium were injected directly into the right dorsal flank per mouse. Tumor growth was measured with calipers every 3 days, and the tumor volumes were calculated using the formula: 1/2 (length × width2). Mice were euthanized and tumor weight was examined 27 days after the injections.
Chromatin immunoprecipitation assay
Chromatin immunoprecipitation (ChIP) assays were performed using the ChIP Assay Kit (Beyotime, Shanghai, China) according to the manufacturer’s protocol with slight modifications. Cells were cross-linked with 1% formaldehyde and terminated after 10 min by the addition of glycine at a final concentration of 0.125 M. DNA was immunoprecipitated from the sonicated cell lysates using an ATF4 antibody (Cell Signaling Technology, Beverly, MA, USA); IgG (BD Biosciences, San Diego, CA, USA) served as the negative control. The DNA was subjected to PCR to amplify the ATF4 binding sites. The amplified fragments were then analyzed on an agarose gel. Chromatin (10%) was used before immunoprecipitation as the input control. The primer sequence was as follow: 5′-GTTTGCATCCCTGCGTGT-3′ and 5′-CCGAGCTTCCTCACCAACT-3′.
Statistical analyses
Data analyses were performed using Graph Pad (GraphPad Prism, La Jolla, CA, USA), Excel (Microsoft Corp, Redmond, WA, USA) and SPSS 20.0 (SPSS, Chicago, IL, USA. The Chi-square test was used to assess correlations between PSAT1 expression and the clinicopathological features of ER-negative breast cancer patients. Survival curves were generated using the Kaplan-Meier method and the log-rank test. Student’s t-test was used to determine significant differences between two experimental conditions. Data from The Cancer Genome Atlas for breast invasive carcinoma (TCGA BRCA) were downloaded from the UCSC Xena Database (
https://tcga.xenahubs.net/download/TCGA.BRCA.sampleMap/HiSeqV2.gz; Full metadata) and were used to detect PSAT1 and ATF4 expression in various types of breast cancer. The level of significance was set at
P < 0.05.
Discussion
It is well known that cancer cells possess distinct metabolic characteristics that are distinguishable from those of nonmalignant cells. Recent evidence has shown that serine metabolic reprogramming is due to the corresponding genetic changes in metabolic enzymes and that these gene modifications independently contribute to tumorigenesis [
6,
7].
PSAT1 is the protein-coding gene of phosphoserine aminotransferase, which catalyzes serine biosynthesis. PSAT1 is overexpressed in colon cancers where it contributes to cell proliferation and chemoresistance, which result in a poor prognosis [
9]. Liu et al. [
8] have shown that PSAT1 expression was elevated in ESCC and that it was significantly associated with disease stage, lymph node metastasis, distant metastasis and poor outcome. A recent study has shown that the expression of PSAT1 was up-regulated in NSCLC, which was verified by an IHC assessment of 138 specimens and a qRT-PCR assay and its overexpression has also been associated with a poor prognosis of NSCLC. Martens et al. [
30,
31] showed that PSAT1 inactivation by promoter methylation and low mRNA levels were both associated with a good outcome after tamoxifen treatment in ER-positive breast cancer. Our current study found for the first time that the expression of PSAT1 was significantly up-regulated in ER-negative breast cancers compared with ER-positive breast cancers, which was supported by the TCGA dataset. We then confirmed this finding by IHC using a tissue microarray, qRT-PCR and western blotting. Statistical analysis of these results showed that PSAT1 up-regulation was correlated with tumor development and poor prognosis.
Previous studies have shown that PSAT1 plays a vital role in cell proliferation as it acts as an oncogene in colon cancer and NSCLC [
9,
11]. Possemato et al. [
6] have illustrated that, through the suppression rate of the serine product, the inhibition of PSAT1 significantly decreased the proliferation of ER-negative breast cancer cells (MDA-MB-468 and BT-20) but not ER-positive breast cancer cells (MCF7). In this study, we also identified the function of PSAT1 in ER-negative breast cancer cells by applying gain- and loss-of-function approaches. We found that PSAT1 regulates the expression of cyclin D1, which is an important regulator of G1 to S phase in a variety of cancers, including breast cancer, to promote cell cycle progression [
32‐
34]. Glycogen Synthase Kinase-3 (GSK-3), a serine/threonine protein kinase, was initially considered to be a key enzyme involved in glycogen metabolism [
35], but is now recognized as a regulator of diverse cellular functions [
36,
37]. Due to its kinase activity, GSK-3β is able to target cyclin D1 and β-catenin [
38] [
28] for ubiquitin-dependent proteasomal degradation. Our current study has shown that PSAT1 enhanced the stability of cyclin D1 via the induction of the phosphorylation of GSK-3β. GSK-3β was inactivated by phosphorylation, which resulted in its accumulation and the nuclear translocation of β-catenin [
39,
40]. Consistently, we found that PSAT1 promoted the stability of β-catenin and its translocation into the nucleus through an enhancement of the phosphorylation of GSK-3β. β-catenin signaling has often been demonstrated to up-regulate the transcription of the cyclin-D1 protein [
41]. It is worthy to note that our current study of PSAT1 focused on GSK-3β, through which PSAT1 eventually enhanced the proliferation and metastasis of tumor cells [
8,
11]. Our current study found that PSAT1 enhanced the migration and invasiveness of ER-negative cells but reduced apoptosis (Additional file
2: Figure S1A and B). Given that previous studies have shown that GSK-3β is a promising target for cancer treatment, further research on the mechanism of PSAT1 and GSK-3β in ER-negative breast cancer may provide more valuable insight into optimal treatments for this type of breast cancer.
Yan et al. [
42] reported that PSAT1 was a direct target of miR-340 and that its overexpression partially reversed miR-340-mediated inhibition of viability, invasion and EMT in ESCC cells. ATF4 transcriptionally activates serine biosynthetic genes in response to serine starvation in non-small cell lung cancer, and additionally, it has been shown to play a crucial role in the regulation of PSAT1 after OSN (Oct4, Sox2, and Nanog) was expressed in mouse embryonic stem cells [
43,
44]. In this study, we found that the knockdown of ATF4 led to the down-regulation of PSAT1, and ChIP confirmed that ATF4 was bound to the ATF4-binding consensus sequences on the PSAT1 promoter in ER-negative breast cancer cells.
Conclusions
To conclude, for the first time, our study demonstrated that PSAT1 was significantly up-regulated in ER-negative breast cancer. Consequently, this up-regulation was able to enhance the proliferation of ER-negative breast cancer cells in vitro via the GSK3β/β-catenin/cyclin D1 pathway and was able to promote tumor development in vivo. In addition, further investigation showed that PSAT1 was activated directly by ATF4 in ER-negative breast cancer. These results indicate that, as an oncogene, PSAT1 plays a vital role in the development of ER-negative breast cancer.
Acknowledgements
The authors thank personnel at Harbin Medical University Cancer Hospital and Harbin Medical University for their generous help.