Background
Autophagy is a process of cellular protein degradation through the autophagosomic-lysosomal pathway, which plays an important role in cell differentiation and maintenance of cellular homeostasis. However, it is usually defective in tumor cells [
1,
2].
Beclin 1, the mammalian orthologue of the yeast
Atg6/Vps30 gene, is the first identified tumor suppressor gene in human to mediate autophagy [
3,
4]. It was originally isolated by a yeast-two-hybrid screen and its protein was identified as an interacting partner of Bcl-2, an important anti-apoptosis protein [
5]. Beclin 1 has a regulatory role in the process of vesicle nucleation of autophagy [
5,
6]. Previous studies demonstrated that over-expression of
beclin 1 induced apoptosis via activation of caspase-9 in gastric cancer cells [
7], while partial silencing of
beclin 1 aggravated apoptosis in hepatic cancer cells [
8]. The different effects of
beclin 1 on cell death and cell survival in different cells depend on the cellular context.
Beclin 1 was mapped to a tumor susceptibility locus approximately 150 kb centromeric to
BRCA1 on human chromosome 17q21 [
9]. Allelic loss of chromosome 17q21 is often found in human prostate, breast and ovarian cancer [
10‐
13].
Beclin 1 encodes an evolutionarily conserved 60 kDa coiled coil protein that is widely expressed in human normal adult tissues [
9]. It has been reported that reduced levels of
beclin 1 expression and mono-allelic deletion were observed in human breast cancer cell lines and tissues [
9]. Whether there are other mechanisms for the loss of
beclin 1 expression in breast cancer remains to be determined.
DNA methylation is the major epigenetic modification that involves alterations of chromatin structure. There are increasing evidences that aberrant methylation of CpG islands in 5' regulatory region of tumor suppressor gene leads to transcriptional silencing in cancer [
14‐
16]. The human
beclin 1 gene contains a 1.5 kb CpG island from the promoter to part of the intron 2, suggesting that DNA methylation may be responsible for down-regulation of
beclin 1 expression in cancer. In addition, the promoter-associated CpG island of
beclin 1 contains E2F target site and four putative consensus Sp1 binding sites [
17]. In the present study, we detected the mRNA and protein expression levels of beclin 1 and explored the possible effects of DNA methylation and LOH on decreased gene expression in breast cancer tissues. The results here provided some new insights into the regulation of
beclin 1 in breast cancer.
Methods
Tissue samples
20 pairs of tumors and adjacent normal tissues from newly diagnosed patients with sporadic breast invasive ductal cancer (IDCs) were collected from the First Affiliated Hospital of China Medical University after the approval of Institutional Review Board and patients' informed consents. All the patients were females without family hereditary breast cancer. Radio and chemo therapy was not applied to the patients before surgical operation. The patients were consecutive cases. The median age of these patients was 48.5 year-old (range, 40 - 74). The clinicopathologic parameters, including patient's age, tumor size, tumor grade, lymph node status, estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor-type 2 (HER2) immunoreactivity were obtained from clinical records. The tissues were obtained after surgical resection and subsequently microdissected with the assistance of pathologists. The corresponding adjacent normal tissues were derived from sites adjacent at least 1 cm away from the tumors. Tissues for immunohistochemistry use were fixed in 10% buffered formalin, embedded in paraffin, and sectioned with a microtome. Sections were stained with hematoxylin and eosin for histological examination by at least two pathologists. Tissue fragments were immediately frozen and stored in liquid nitrogen till used.
RNA extraction and quantitative reverse transcription PCR
Total RNA from tissues was extracted using TRIZOL reagent (Invitrogen) according to the manufacture's protocol. The first strand cDNA was obtained from total RNA (0.5 μg) and oligo (dT) using the Reverse Transcription System (Promega). For real-time quantitative RT-PCR, gene specific primers and TaqMan fluorescent hybridization probes were used. β-actin was used to normalize the quantity of specific mRNA. The sequences of
beclin1 and β-actin primers are listed in Table
1. The PCR product was 85 bp for
beclin1 and 295 bp for β-actin. Their authenticity was confirmed by DNA sequencing. The amplification efficiency determined for both target and housekeeping genes was equal. Relative expression levels were calculated by the 2
-ΔΔCt method [
18]. Each assay was done in triplicate.
Table 1
Nucleotide sequences of primers used
Bl-F | TGCAACCTTCCACATCT | RT-PCR |
Bl-R | TTCCACGGGAACACTG | RT-PCR |
β-actin-F | TCACCCACACTGTGCCCATCTACGA | RT-PCR |
β-actin-R | CAGCGGAACCGCTCATTGCCAATGG | RT-PCR |
Bl-QF | TCTGCCTTCCTCTGTAG | Q-PCR |
Bl-QR | TTCCACGGGAACACTG | Q-PCR |
18S-QF | ACATCCAAGGAAGGCAGCAG | Q-PCR |
18S-QR | TTCGTCACTACCTCCCCGG | Q-PCR |
Bl-Taqman | FAM-CACAGTGGACAGTTTGGCACAATCA-TAMRA | Q-PCR |
β-actin-Taqman | FAM-CAGCCGTGGCCATCTCTTGCTCGAA-TAMRA | Q-PCR |
18S-Taqman | FAM-CGCGCAAATTACCCACTCCCGA-TAMRA | Q-PCR |
Bl-bs1F | GTTTTTTAAAGTGTTGGAATTATAAG | Bisulfite first PCR |
Bl-bs1R | AACTCCTAATCCACAAACTCACAA | Bisulfite first PCR |
Bl-bs1F' | TTGTTGTTGTTTTGAGATGGAGTT | Bisulfite nested PCR |
Bl-bs1R' | AAAAATATAAAAACCAAAACC | Bisulfite nested PCR |
Bl-bs2F | GGGTTTGTGAGTTTGTGGATTAG | Bisulfite first PCR |
Bl-bs2R | AAAAAAAACTCCAATAAAAACC | Bisulfite first PCR |
Bl-bs2F' | AGTTTGTGGATTAGGAGTTTTTGTT | Bisulfite nested PCR |
Bl-bs2R' | TAAAAATTCCCAAACTCCCTTCTA | Bisulfite nested PCR |
Bl-bs3F | ATTTTAGAAGGGAGTTTGGGAATT | Bisulfite first PCR |
Bl-bs3R | TTAAACCCTTCCATCCCTAAAAC | Bisulfite first PCR |
Bl-bs3F' | TTTTGGGTTTTAAATTGTTTTTGTT | Bisulfite semi-nested PCR |
Bl-bs3R' | TTAAACCCTTCCATCCCTAAAAC | Bisulfite semi-nested PCR |
Bl-bs4F | ATATTGTGGATTTTTGAGAGTTTTT | Bisulfite first PCR |
Bl-bs4R | AAATCTTTCTTTTACTACTAAAAACTCTCT | Bisulfite first PCR |
Bl-bs4F' | TTGTAATTTTAGTATTTTGGGAGAT | Bisulfite nested PCR |
Bl-bs4R | CTCTATTACCCAAACTAAAATACAATAATA | Bisulfite nested PCR |
D17S579F | AGTCCTGTAGACAAAACCTG | Microsatellite analysis |
D17S579R | CAGTTTCATACCAAGTTCCT | Microsatellite analysis |
Genomic DNA extraction and quantitative PCR
Genomic DNA of tumors and matched adjacent normal tissues was extracted with the Dneasy Tissue Kit (Qiagen). A Taqman-based quantitative PCR method using PRISM 7300 Real-Time PCR Thermal Cycler (Applied Biosystem) was applied to detect the relative copy number of beclin 1. DNA content per haploid genome was normalized to that of 18S rDNA and calculated by the 2
-ΔΔCt method [
18]. Table
1 showed the sequences of the specific primer pairs and the Taqman probes. PCR products were sequenced to confirm their authenticity. Each assay was done in triplicate. A dilution series was performed to quantify primer efficiency.
Mutational and microsatellite analysis
To search for mutations of the
beclin 1 gene by polymerase chain reaction - single strand conformation polymorphism (PCR-SSCP) analysis, each of the 12 exons was amplified using its specific primers. The sequences of the primers and PCR conditions were as described by Aita et al [
9]. Denatured PCR products were loaded onto non-denaturing 6% polyacrylamide gels, electrophoresed at 4°C, stained with silver staining, and visualized.
To analyze LOH of the
beclin 1 gene, a microsatellite marker on 17q21, D17S579, was selected through UniSTS database of National Center for Biotechnology Information (NCBI). Primer sequences are shown in Table
1. We defined LOH as a completely absent or significantly decreased signal intensity of one allele.
Sodium bisulfite modification and sequencing
The methylation status of the
Beclin 1 CpG island was determined by the bisulfite sequencing method [
19]. 0.5~1 μg genomic DNA was digested with BamHI that does not cut within the sequence of the CpG island. DNA was denatured by 0.3 M NaOH at 42°C for 30 min. Then 3 M sodium bisulfite (Sigma) and 10 mmol/L hydroquinone were added, which was further incubated at 50°C for 16 h. Modified DNA was purified using Wizard DNA purification resin (Promega), re-treated with 0.3 M NaOH at 37°C for 15 min, precipitated with ethanol, and resuspended in 50 μL water. Semi-nested or nested PCR was used to amplify these regions. Primer sequences are detailed in Table
1. Five clones of each PCR sample were picked up and sequenced.
Immunohistochemistry
Protein expression of beclin 1 was performed on breast tissues fixed by formalin and embedded by paraffin. Slides of 4 μm sections were deparaffinized with xylene and antigen retrieval was accomplished by heat. The sections were then incubated in 3% hydrogen peroxide at room temperature for 5 min to block endogenous peroxidase activity. Slides were then incubated with rabbit anti-beclin 1 polyclonal antibody (Cell Signaling) at 1:300 dilutions at 4°C overnight. Then rinse the slides three times in PBS for 5 m each and incubate in biotin-labeled rabbit anti-rabbit secondary antibodies for 1 h at room temperature. After washing three times with PBS, the staining was performed using 3, 3'-diaminobenzidine. Sections were counterstained with hematoxylin. Staining was considered positive when cytoplasmic staining was observed in at least 30% of the neoplastic cells. And staining intensity was the strength of the signal and was evaluated on a scale of 0 to 2: 0, negative; 1+, weak staining; 2+, moderate/strong staining. We also compared IHC data between tumors and corresponding adjacent normal tissues by percentage of positive cells and intensity of staining to estimate the changes of beclin 1 expression.
P53 protein expression was performed using anti-p53 monoclonal antibody (Ab-8, Lab Vision Corporation), which recognized both wild type and mutant forms of the p53 protein. BRCA1 and BRCA2 protein expression were assessed using an anti-BRCA1 monoclonal antibody (Ab-1, Oncogene Research Products) and an anti-BRCA2 polyclonal antibody (Ab-2, Oncogene Research Products). Samples were considered positive when 20%, 10%, and 10% of the cells were stained with p53, BRCA1 and BRCA2, respectively.
Data analysis
Data were expressed as mean ± standard deviation. The ratio of beclin 1 mRNA expression in the tumor to that in the corresponding adjacent normal tissue in each case was transformed using a common logarithm. For statistical comparisons of these log-transformed data between two groups, a t-test was used. All statistical analyses were performed using SPSS software, version 10.0 (SPSS Japan Inc., Tokyo, Japan). For all of the statistical tests, a two-sided p-value of less than 0.05 was considered statistically significant.
Discussion
It has been reported that tumor suppressor genes with high frequencies of LOH in human chromosomal region 17q21 and epigenetic silencing, are important in hereditary and sporadic breast cancer tumorigenesis [
25,
26]. One of these genes is
BRCA1, whose decreased expression was often observed in breast cancer with epigenetic silencing and mono-allelic deletion of the
BRCA1 gene [
27‐
29]. In the present study, we showed that the mRNA and protein expression of
beclin 1, another 17q21 gene, was frequently down-regulated in breast tumors. This down-regulation in some tumors was due to allele loss in gene copy number and in some others was due to DNA methylation. There was no significant association between
beclin 1 mRNA expression and clinocopathologic parameters.
P53, BRCA1 and BRCA2 are clearly involved in the development of both sporadic and hereditary breast cancers [
30‐
34]. Mutations of these genes are the most common genetic variations and are widely distributed in breast cancer cells. Tumors with missense mutations displayed positive nuclear immunoreactivity of p53 for the effect of mutation on the prolongation of p53 half-life through increased protein stability [
35]. Cancers with wild-type P53, including breast cancer [
35], are generally negative for p53 immunoreactivity due to the rapid degradation of normal p53 protein. Meanwhile, loss or reduction of BRCA1 and BRCA2 expression has been exhibited in sporadic breast cancers [
36]. Immunohistochemistry of p53 and BRCA1/2 proteins in tumor cells may be useful as an additional method in exploring the relationship between the beclin 1 expression levels and the status of these important tumor suppressor genes in breast cancer. Therefore, we examined their expression levels by immunohistochemcial staining in 20 breast tumors. By IHC staining of p53 protein, we found 2 cases (T11 and T17) exhibited more than 50% positive nuclear staining cells of all tumor cells. IHC over-expression of p53 might reflect that genetic alterations also occurred in these tumors. We further analyzed the correlation between
beclin 1 expression and p53 over-expression and no obvious association was obtained. It is known that the normal BRCA1 protein plays an important role in repairing breaks in DNA [
37]. However, when
BRCA1 exhibited loss of expression due to genetic or epigenetic alterations, abnormal repair function of BRCA1 may lead to DNA replication errors and cancerous growth [
38,
39]. Liang
et al [
3] reported that
beclin 1 played a negative regulatory role in mammary cell growth and tumorigenesis using gene-transfer techniques. Our data showed that
beclin 1 expression was significant higher in the BRCA1 positive tumors than in the negative ones, suggesting
beclin 1 expression may be related to cell growth in breast cancer.
Beclin 1 is mapped to a region approximately 150 kb to
BRCA1 on chromosome 17q21, which is usually deleted in breast, ovarian and prostate cancer [
8‐
11]. LOH at the
beclin 1 locus was observed in 9 out 22 (41%) of the breast cancer cell lines [
9]. In the present study, all examined tumors exhibited a LOH rate of at least 45% at the
beclin 1 locus and no mutation was found in the coding regions of
beclin 1. These results confirmed that frequent allelic losses of
beclin 1 were part of reasons for
beclin 1 down-regulation and mutations were unlikely the main regulatory mechanism for
beclin 1 inactivation in breast cancer. Meanwhile, allelic loss of
beclin 1 was only found in a portion of the breast tumors with down-regulated expression, indicating that mechanisms other than allelic deletion may be responsible for the decrased mRNA expression.
The CpG island in
beclin 1 spans over 1.5 kb from the promoter to the intron 2 [nucleotide (nt) -528 to 977] with the transcription start site defined as +1, so it is divided into four regions to be cloned. The first region contained part of the beclin 1 promoter from -528 to -65. The second region was located between -83 and 164 from the proximal promoter to part of the first intron. The third region laid between 137 and 400 from a portion of the first intron to part of the second exon. The fourth region was situated between 733 and 977 from a portion of the second intron. We found that certain CpG dinucleotides at the promoter and the intron 2 of the
beclin 1 gene are hypermethylated in breast tumors. As dense methylation of the 5' CpG islands was not detected in normal tissues, we speculated that this methylation patterns in tumor cells were aberrant and disease associated. As a whole, the regulation of
beclin 1 expression was a little complicated. In some cases, T16 and T20, the decreased expression was due to aberrant DNA methylation; While in T1 and T12, the decrease was from LOH; And in the other cases, T3 and T8, both aberrant DNA methylation and LOH devoted to the decreased expression of
beclin 1 (Table
8). For this reason, different mechanisms seemed to be involved in the regulation of
beclin 1 expression.
Table 8
Summary of LOH and aberrant DNA methylation in breast tumors showing down -- regulated expression of beclin 1
Some tumors (T10 and T19) with low expression of
beclin 1 did not have either LOH or methylation. Other identified mechanisms might contribute to the regulation of beclin 1 expression, such as bcl-2 overexpression [
40]. Recently, a novel beclin 1 binding protein, UVRAG, was found to positively regulate autophagy signaling pathway mediated by beclin 1 in colon cancer [
41]. Exploring different involving factors that could affect the regulatory expression of
beclin 1 might contribute to the pathogenesis of human cancers.
Aberrant methylation of 5' CpG islands associated with down-regulated mRNA expression of
beclin 1 indicated that methylation might be a new mechanism for loss of expression of
beclin 1, which has not been shown previously. Sequence analysis revealed that there were four putative consensus Sp1 binding sites at the promoter [
17] and one putative Sp1 binding site at the intron 2, which share high sequence homology to the consensus Sp1 motif. It has been reported that aberrant DNA hypermethylation in the promoter of a gene can silence its expression [
42], and that methylation in the introns can silence or enhance its expression [
43‐
45]. Whether DNA methylation in the promoter and the intron 2 of
beclin 1 affects the binding affinity of transcriptional factor needs to be further determined.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
LZD carried out molecular biological studies, participated data collection, analysis and the preparation of the manuscript. CB and JF collected tissue samples and participated in the clinical part of the study and immunohistochemistry. WYQ, XYJ and LXJ supervised the project and overviewed the analysis of the data and the manuscript. All authors have read and approved the present manuscript.