Elevated expression of RNA methyltransferase BCDIN3D predicts poor prognosis in breast cancer

Background BCDIN3D is a member of the Bin3 methyl-transferase family that targets the 5' mono-phosphate of nucleic acids. Although BCDIN3D has been shown to increasetumorigenic phenotypes and invasiveness in MDA-MB-231 cells, its the clinical implications in breast cancer remain unclear. Methods We screened for BCDIN3D using tissue microarrays constructed from 250 patients who were histologically confirmed to have invasive ductal breast carcinoma at the Fudan University Shanghai Cancer Center. Results The survival analysis by Kaplan-Meier and Cox regression showed that BCDIN3D expression level served as a prognostic factor for disease-free survival (P = 0.042). The prognostic value of BCDIN3D was most significant in triple-negative breast cancer (TNBC) patients (P = 0.007). Conclusions BCDIN3D might serve as an important prognostic factor for TNBC patients.


bAckground
Breast cancer is the most common cancerin women, and accountfor approximately 23% of allcancer cases and approximately 14% of cancer deaths [1]. It is a heterogeneous disease embracing several different phenotypes [2], including luminal A, luminal B, human epidermal growth factor receptor 2 (HER2)-enriched, and triple-negative breast cancer (TNBC) [3].Breast cancer results from the accumulation of genetic and epigenetic alterations [4,5]. Epigenetic alterations, as defined by modifications of DNA, histones and coding/ noncoding RNAs [6,7], also contribute to various phases of neoplastic development including initiation, promotion, invasion, metastases.
RNA modification is a kind of epigenetic regulation of gene expression, analogous to DNA methylation and histone modification. [8]. Methylation is a ubiquitous modification that affects several residues/ sites in molecules [9]. Methyl-transfer reactions to RNA nucleotides are catalyzed by a variety of RNA-MTases that include more than 60 members with hundreds of homologs, and which have so far been divided into four super-families [10]. BCDIN3D, a member of the Bin3 methyltransferase family, share homology within their putative S-adenosyl Methionine (SAM) binding motif from S. pombe to human. SAM (S-adenosylmethionine) is well known as the methyl donor for methyl-transferases that modify DNA, RNA, histones [11].
BCDIN3D is a methyltransferase that targets the 5'mono-phosphate of nucleic acids.As depletion of BCDIN3D in the MDA-MB-231 cells abolishes anchorage-independent growth and decreases invasiveness in MDA-MB-231 cells, but not growth ormigration [12], its prognostic value may be of interestHerein, we investigated BCDIN3D expression and its association with tumor progression and clinical outcome in a cohort of 250 patients who had undergone surgery for breast cancer in eastern Chinese women population.

expression patterns of bcdIn3d in breast cancer patients
In this cohort of 250 patients, TMAswere immunostained for BCDIN3D (representative images, Figure 1). The specificity of the antibody against BCDIN3D was confirmed by Western blot of human cell lysis. Positive staining of BCDIN3D was detected in 49.6% of tumors according to the scoring criterion described above (n = 124; 49.6% positive, 50.4% negative; Figure 1a-1d). We further analyzed relationships between clinicopathologic features and the expression level of BCDIN3D. The percentages of positive staining of protein were consistent across all subsets of patients (Table 2).

BCDIN3D was identified as a significant prognostic factor in breast cancer patients, especially in triple-negative breast cancers (tnbc)
Both univariate and adjusted multivariate survival analyses showed significant differences in DFS between the BCDIN3D positive and negative groups. BCDIN3D positive group having a significantly higher incidence of disease eventsin both univariate analysis (HR = 1.754; 95% CI: 1.012-3.039; P = 0.045) and multivariate analysis. (HR = 1.904; 95% CI: 1.081-3.354; P = 0.026) ( Table 3). The BCDIN3D positive group also showed worse DFS in the Kaplan-Meier survival analysis (P = 0.042; Figure  2a). Thus, these results indicate that BCDIN3D expression isdirect associated with breast cancer recurrence. Then we analyzed the relationship between BCDIN3D expression and survival according to the different breast cancer subtypes. We found the prognostic value of BCDIN3D for DFS was most significant among patients with TNBC (P = 0.007; Figure 2b). In the TNBCsubset, patients with positive BCDIN3D staining showed a higher likelihood of occurence of disease events (HR = 3.584; 95% CI: 1.319-9.737; P = 0.012) in univariate analysis and remain the same trend in multivariate analysis. (HR = 3.719; 95% CI: 1.345-10.283; P = 0.011)( Table 4).

dIscussIon
Breast Cancer, is now known to involve epigenetic abnormalities along with genetic alterations [13]. Epigenetic modifications are early events in breast carcinogenesis and could be usefulfor early detection, prognosis, and targeted therapy of breast cancer [14]. Post-transcriptional gene regulation by noncoding RNA commonly referred as microRNAs is a kind of epigenetic regulation [15,16]. It has become one of the most dynamic and fast-growing fields in science. Despite the rapid advance of RNA research, the enzymes that post-transcriptionally modify RNA have been less investigated [17].
A post-transcriptional modification enzyme BCDIN3D O-methylates the 5′ terminal monophosphate group of the precursor of some miRNA, which increasedtumorigenic phenotypes and cells invasiveness [12]. To our knowledge, no one else has explored its clinical significance. Our study is the first to evaluate the prognostic value of BCDIN3D in breast cancer patients. Our finding that patients with BCDIN3D positive tumors had worse DFS than the BCDIN3D negative group supports previous biological function study results from other researchers. Interestingly, this association was found to be most significant among patients with TNBC. TNBC breast tumors lack ER, PR, and HER-2 expression and occupy 15-20% of all breast cancers. It is generally more aggressive, has higher rates of relapse and decreased overall survival [18]. As few biomarkers are widely considered to be predictive for TNBC prognosis, and thus predictive factors for TNBC are urgently need. Our findings suggest that BCDIN3D might serve as an important prognostic factor for TNBC patients.
Our study had several limitations, including the small data set, few cases of stage I patients, lack of validation in an independent series of cases, and the composition of the study cohort which did not exactly represent that of the breast cancer population. Howeverour study was based on the use of TMAs, which can guarantee the consistency and coherence of these factors.

conclusIons
In conclusion, our results associate, for the first time, higher BCDIN3D levels with worse DFS, especially Oncotarget 53897 www.impactjournals.com/oncotarget in triple-negative breast cancer population, which suggests its potential use as a predictive biomarker. Furthermore, as more future basic and clinical studies uncover the underlying mechanism in this pattern, BCDIN3D could emerge as a desperately needed therapeutic targetin triplenegative breast cancer.

ethics statement
This study was approved by the Ethics Committee of FDUSCC, and each participant signed an informed consent document.

Tissue microarrays (TMAs)
TMAs were constructed from formalin-fixed, paraffin-embedded samples of carcinomas obtained from the250 breast cancer patients. A hematoxylin-and eosinstained section of each tumor block was used to mark representative tumor regions. Tissue cylinders with a diameter of 2 mm were punched from the above regions and transferred to recipient array blocks using a Tissue Micro Arrayer (Beecher Instruments, Sliver Springs, MD, USA). TMAs werecomposed of duplicate cores from different areas of the same tumor to compare staining patterns.

IHC experimental procedures
The tissue micro arrays were subjected to immunohistochemical staining for BCDIN3D, using a 2-step protocol (GTVisionTMIII). The primary antibodies used were monoclonal BCDIN3D antibody (Santa Cruz Biotechnology, CA, USA sc-390348). The specificity of the BCDIN3D antibody was validated by western blot ( Figure S1). The TMAs were deparaffinized with xylene, and rehydrated with an ethanol gradientThe sections were then rinsed with phosphate buffer solution     Oncotarget 53901 www.impactjournals.com/oncotarget (PBS) for immunohistochemical staining. For antigen retrieval the sections were immersed in 0.01 M tris sodium citrate pH 6.0 and boiled at 121 °C for 5 min followed by 2 min simmering. All slides were incubated with nonspecific staining blocking agent for 20 minutes to quench endogenous peroxidase activity and then with anti-BCDIN3D (1:200) at 4 °C overnight. Primary antibodies were detected by HRP-conjugated secondary antibodies followed by colorimetric detection with 3, 3-diaminobenzidine (DAB). The TMAs were then counterstained with Gill hematoxylin and dehydrated in an ascending ethanol series before clearing with xylene and mounting under a coverslip.

Evaluation of immunohistochemical variables
TMAs were stained and scored semi-quantitatively. The score used in all subsequent analysis was the average across the available cores. Staining was graded for intensity of staining (0, no staining; 1+, faint/equivocal; 2+, moderate; 3+, strong) and percentage of cells stained (0, no staining; 1+, < 10% of cells stained; 2+, 10% -50% of cells; and 3+, >50% of cells stained). For this study, SI ≥3 was defined as positive staining, and SI ≤2 was defined as negative staining. Scoring was reviewed in parallel by two experienced breast disease pathologistswho were blinded to all clinical data.

statistical analyses
Disease-free survival(DFS) was defined as the time between the date of the primary surgery to the date of relapse/breast cancer specific death or September 2013. The first recurrence of disease at a local, regional, or distant site; diagnosis of contralateral breast cancer; and the breast cancer specific death were considered DFS events. Patients with study end date and loss of followup were considered to be censored. Correlations between clinical-pathological parameters and BCDIN3D were tested using the Chi-squared test. Survival outcomes were estimated using the Kaplan-Meier method and were compared between the groups using log-rank statistics. Univariate and multivariate analyseswere carried out using the Cox risk proportion model. Statistics was analyzed using SPSS (version 13.0; SPSS Company). All P values are two-sided; P value less than 0.05 was considered significant. All analyses were based on the observed data with the assumption that missing data were completely at random.