http://orcid.org/0000-0001-7091-1959
Figures
Abstract
Background
Long non-coding RNAs (lncRNAs) are characterized as having 200 nucleotides or more and not coding any protein, and several been identified as differentially expressed in several human malignancies, including breast cancer.
Methods
Here, we evaluated lncRNAs differentially expressed in triple-negative breast cancer (TNBC) from a cDNA microarray data set obtained in a previous study from our group. Using in silico analyses in combination with a review of the current literature, we identify three lncRNAs as potential prognostic factors for TNBC patients.
Citation: Rodrigues de Bastos D, Nagai MA (2020) In silico analyses identify lncRNAs: WDFY3-AS2, BDNF-AS and AFAP1-AS1 as potential prognostic factors for patients with triple-negative breast tumors. PLoS ONE 15(5): e0232284. https://doi.org/10.1371/journal.pone.0232284
Editor: Sushanta K. Banerjee, University of Kansas Medical Center and VA Medical Center, UNITED STATES
Received: December 5, 2019; Accepted: April 10, 2020; Published: May 13, 2020
Copyright: © 2020 Rodrigues de Bastos, Nagai. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) provided support for this study in the form of Research Grant Number 309524/2017-2, awarded to MAN. CAPES (Coordinação de Aprimoramento de Pessoal de Nível Superior) provided support for this study in the form of a PhD fellowship for DRB. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Breast cancer represents the second most common type of cancer worldwide and with high morbidity and mortality rates considering only females [1]. Breast cancer shows high heterogeneity, which impacts the clinical course of the disease [2]. Differences in the expression profile among patients contribute to the heterogeneity that gives the tumor tissue differences in malignant behavior and impact on the prognosis and response to usual treatments [3]. Based on the expression of molecular biomarkers, breast cancer is classified into four main subtypes: Luminal A, Luminal B, HER2-positive, and triple-negative (TNBC). The Luminal A and B breast cancer subtypes express hormone receptors, estrogen receptor (ER) and progesterone receptor (PR). The HER2-positive subtype is negative for the hormonal receptors and shows high expression levels of the human epidermal growth factor receptor 2. The triple-negative subtype (TNBC) consists of breast tumors negative for the expression of hormone receptors and the HER2 oncoprotein [4–6].
TNBC can be classified into at least six distinct subtypes with differences in clinical behavior and treatment response [7, 8]. TNBC subtype comprises between 15 and 20% of all breast tumors and is generally considered to have a poorer prognosis as patients do not respond to endocrine therapy and target-directed therapies [9, 10]. TNBCs are commonly observed in younger and obese women, being more prevalent in premenopausal African American women [11]. Germline mutations of BRCA1 and BRCA2 are found in about 20% of patients with TNBC [12]. Also, tumors of the TNBC subtype show a high frequency of P53 and Rb1 mutations [13].
ncRNAs comprise a broad class of RNAs that can be divided into two subclasses based on their size: long non-coding RNAs and small non-coding RNAs [14]. Long non-coding RNAs (lncRNAs) are characterized by having 200 nucleotides or more and not coding for any protein [15]. lncRNAs have been identified as differentially expressed in several human malignancies [16–18], including breast cancer [19, 20]. There are several mechanisms of action attributed to these lncRNA molecules, highlighting their participation in protein recruitment, scaffold, endogenous competition for microRNAs (ceRNA) and enhancer of other genes [21, 22]. LncRNAs play an important role in cellular homeostasis, including participation in multiple pathological processes and tumorigenesis of various tissues [23, 24]. However, to date, the function of most lncRNAs has not been thoroughly characterized [25], requiring in silico and experimental studies to further investigate their role in cell and tumor biology.
Many lncRNA molecules have been associated with the tumorigenic process, and their differential expression may represent a promising category of potential new biomarkers. In cancer cells, the lncRNA HOTAIR (HOX transcript antisense intergenic RNA), was associated with the polycomb repressing complex 2 (PRC2) and the histone demethylation enzyme lysine-specific demethylase 1 (LSD1) resulting in epigenetic regulations that lead to the tumor growth and metastasis [26]. The overexpression of HOTAIR was observed in TNBC patients and was associated with increased cell proliferation [27] and worst prognosis [28, 29]. Besides, HOTAIR has been detected in the blood of patients, and high circulating levels were correlated with worse prognosis and less response to neoadjuvant chemotherapy [30]. Furthermore, other circulating lncRNAs, such as MALAT1, GAS5, H19, and MEG3, has been associated with survival and chemotherapy response [31–34]. LncRNAs have opened a new field of study for researchers around the world, and significant functions are attributed to these molecules, which may directly impact patient survival and therapy response [35]. In this sense, the identification of differentially expressed transcripts of lncRNAs in tumor tissues is necessary to detect new potential prognostic and predictive biomarkers in breast cancer.
The SPARC gene (secreted protein acidic and rich in cysteine, also named as osteonectin or basement-membrane protein 40) encodes for a 32 kDa matricellular glycoprotein, that has been involved in several biological processes, such as differentiation, proliferation, migration, and adhesion [36, 37]. Moreover, abnormal SPARC expression has been associated with tumor characteristics such as growth and metastasis in different cancer types [38]. Previous studies from our group evaluating the gene expression profile in breast epithelial cells with the difference in ERBB2 expression before and after treatment with Docetaxel identified differential expression of SPARC as a potential biomarker of chemosensitivity [39, 40]. Subsequently, we investigate the potential prognostic value of the SPARC protein using immunohistochemical analysis on tissue-microarrays. We found that low SPARC expression is associated with worse prognosis and more aggressive phenotypes of breast cancer, including TNBC [41]. More recently, we performed a study using cDNA microarrays to determine the expression profiling of a small subset of triple-negative breast tumors with differences in SPARC expression and clinical outcome, which lead us to identify several differentially expressed genes as potential new biomarkers candidate for TNBC [42].
In the present work, we sought to identify differentially expressed lncRNAs in TNBC with SPARC expression using a cDNA microarray data set obtained in a previous study from our group [42]. Using in silico analyses in combination with a review of the current literature, we examined the transcriptional profiles of lncRNAs between breast tumors and normal tissues and correlate gene expression levels to clinicopathological features and patient’s clinical outcome.
Material and methods
Data selection
A microarray data set, available on the Gene Expression Omnibus (GEO) online platform with access code GSE98931 [42] was reanalyzed by GeneSpring GX software (Agilent Technologies, Santa Clara, California, EUA) for identification of differentially expressed lncRNAs in SPARC positive vs. SPARC negative (fold change > = 1.5, p <0.05) after normalization and correction by Benjamini and Hochberg (Fig 1).
Selection and management of data from the microarray GSE98931. The data were initially processed in the GeneSpring software, resulting in two lists (up and down) containing coding and non-coding transcripts with a > = 1.5-fold change and a corrected p-value less than 0.05 were accepted as differentially expressed. The list was cleaned using R commands to exclude coding transcripts. The differentially expressed lncRNAs were evaluated using the KM Plotter online platform to investigate potential prognostic predictors associated with the basal subtype. We found 17 lncRNAs associated with worse RFS and 10 lncRNAs associated with low OS. *Number of common transcripts in both survival analyses.
Survival analyses
We analyzed the differentially expressed long noncoding RNAs and their association with overall survival (OS) and relapse free survival (RFS) in the KM Plotter online database (http://kmplot.com/analysis/) [43]. Validated probes were chosen according to the automatic best cut off selection criteria. For the survival curves, patients were stratified by the lncRNAs expression according to intrinsic subtype: all, basal, luminal A, luminal B and HER2 + [43].
TANRIC-TCGA data
Gene expression data of 942 breast cancer patients were downloaded using the TANRIC platform (https://www.tanric.org/) [44], of which 105 consisted of adjacent tissue samples and 837 were from tumor tissue. These data were cross-referenced with information available from TCGA (https://www.cancer.gov/about-nci/organization/ccg/research/structural-genomics/tcga) and UCSC Xena (xena.ucsc.) [45] to obtain clinicopathological data such as age, status survival, tumor stage, hormone receptors and human epidermal growth factor receptor type 2 (HER2). Cases of breast cancer or adjacent male breast tissue were excluded from the study [44].
cBioPortal data
Z-score expression data from 1.108 patients containing information about messenger RNAs (mRNAs) were downloaded on cBioPortal database (https://cbioportal.org). This information was cross-referenced with sample IDs from the TANRIC platform to pair samples for correlation analysis [46, 47].
Expression profile and gene networks
Where applicable, we investigate the expression profile through the online UALCAN database (ualcan.path.uab.edu) [48]. Correlation analysis was conducted based on information from lncRNAs and mRNAs according to literature notes, and mRNAs found were processed in the String database (string-db.org/) to identify biological networks [49].
Statistical analysis
The statistical analysis was performed using SPSS (Statistical Package for Social Sciences) version 25 and GraphPad Prism v. 7 (California, USA). The results considered statistically significant were that the p-value was less than 0.05 or according to the p-adjustment when appropriate. We used different statistical tests to evaluate normal distribution, association, correlation, and accomplish group comparisons, which includes Kolmogorov-Smirnov test, chi-square test or Fisher's exact test, correlation test, Mann-Whitney and Kruskal-Wallis, respectively.
Results
In the present study, we focused on the identification of long non-coding RNAs differentially expressed in TNBC. The analysis of the GSE98931 dataset revealed a total of 35 lncRNAs as differentially expressed in TNBC tumors (Fig 2, above the red line).
The red line represents the corrected p-value, which was set as cut-off (0.05). Positive values on the horizontal axis comprise up-regulated transcripts, and negative values on the same axis represent down-regulated lncRNAs, vertical lines at ~0.585 corresponding to 1.5 FC.
To access the potential prognostic value of the 35 lncRNAs identified as differentially expressed, we conducted survival analysis using the KM Plotter platform. Ten out of these lncRNAs (5 up and 5 down-regulated) were not found in the KM Plotter platform and could not be evaluated. For the other lncRNAs, we focused on those with significant associations with the survival rates of patients with breast tumors of the basal subtype, which best represents the triple-negative group of tumors. A total of 17 lncRNAs, 7 of which were up-regulated and 10 down-regulated, were associated with worse relapse-free survival in patients with the basal subtype (S1 Table).
Regarding overall survival, we identified 10 lncRNAs associated with worse prognosis in breast cancer patients with tumors of the basal subtype, of which 4 were up-regulated and 6 down-regulated in our cDNA microarray data (S2 Table). Up- and down-regulated lncRNAs identified as associated with poor prognosis in breast cancer patients with tumors of the basal subtype are summarized in Table 1. Table 1 also shows data reported in the literature classifying these lncRNAs as oncogene or tumor suppressor genes.
We further investigated the expression pattern of the 35 lncRNAs in TANRIC and TCGA databases. We downloaded the spreadsheet with lncRNA expression information on breast tumors directly from the TANRIC database, which contains a total of 12,727 annotated long noncoding RNAs. Twelve out of these 35 lncRNAs were excluded because they were not included in the TANRIC expression dataset; however, 18 lncRNA transcripts were found differentially expressed when comparing normal tissue and tumor tissue (p <0.05; S3 Table).
The clinicopathological features of the patients were downloaded from the TCGA and UCSC platform. Data from both databases were cross-referenced to obtain clinical information pertinent for subsequent analyses. Including data of the different subtypes of breast cancer according to the PAM50 classification (Basal, Luminal A, Luminal B, HER2+ and Normal-like).
The Kruskal-Wallis test was used to identify differences in the expression pattern of 23 lncRNAs among the intrinsic tumor subtypes of the PAM50 classification. Statistically significant differences were found for a total of 21 lncRNAs (S4 Table). These 21 lncRNAs were further evaluated by the Mann-Whitney test and multiple comparisons in order to find transcripts with differential expression in the basal subtype compared to the other subtypes (S5 Table). Four sets of lncRNAs were established for the comparisons: set 1, concerns to transcripts with a p-value less than 0.005 in the Basal vs. HER2 + comparison; set 2, Basal vs. Luminal A; set 3, Basal vs. Luminal B; and set 4, Basal vs. Normal-like. The four sets of tlncRNAs transcripts were used to construct a Venn Diagram, where the central intersection represents the transcripts differentially expressed in the basal subtype compared to the other breast cancer subtypes (Fig 3).
A) Of the 35 differentially expressed lncRNAs identified in our microarray, 23 were found in the TANRIC database. These 23 transcripts were evaluated by the Mann-Whitney statistical test. The sets where the basal subtype is evidenced were used to construct the Venn diagram to identify transcripts differentially expressed primarily in basal tumors. A total of 5 transcripts, represented in the central intersection, were identified with a differential expression preferentially in basal tumors, among other subtypes. Set 1: Basal vs. HER2; set 2: Basal vs. Luminal A; set 3: Basal vs. Luminal B; set 4: Basal vs. Normal-like. B) Elements in the central intersection.
The lncRNAs WDFY3-AS2, BDNF-AS, and AFAP1-AS1, identified as differentially expressed in basal tumors, were associated with poor survival and selected for subsequent analyses. WDFY3-AS2, BDNF-AS, and AFAP1-AS1 were classified into low or high expression according to the median expression value. These data were used to construct the contingency table according to the clinicopathological features of 828 patients coming from the TANRIC-TCGA platform. A descriptive analysis of the analyzable cases is presented in S6 and S7 Tables. In the analysis with all breast cancer subtypes, we observed significant differences in the status of hormone receptors (RP, p <0.0001; RE, p <0.0001), HER2 (p <0.0001), and PAM50 classification (p <0.0001) for those three lncRNAs. For the basal subtype tumors (139 cases), statistically significant differences were found regarding the patient’s age and BDNF-AS expression (p = 0.038) and AFAP1-AS1 (p = 0.011) (Table 2).
The long non-coding WDFY3-AS2 RNA was identified as downregulated in breast tumors relative to normal tissue (Fig 4A), and according to previous analyses in this study, low expression of this transcript was associated with worse prognosis in women with breast cancer (Fig 5A and 5B; Table 1; S1 and S2 Tables). Therefore, we conducted a multiple analysis (S5 Table) using the expression data from 828 patients and stratified the groups according to the PAM50 classification. The WDFY3-AS2 lncRNA was observed with the lowest expression in the basal subtype (Fig 4B).
Expression in normal versus tumor tissue (A, C, E); Expression pattern of WDFY3-AS2, BDNF-AS, and AFAP1-AS1 in different subtypes of breast cancer (B, D, F).
Relapse-free survival for patients stratified by the expression of WDFY3-AS2 (A), BDNF-AS (C), and AFAP1-AS1 (E); Overall survival of patients stratified by the expression. WDFY3-AS2 (B), BDNF-AS (D), and AFAP1-AS1 (F). Survival analysis was conducted on the KM Plotter database (www.kmplot.com).
Our analysis using cBioPortal databases showed that the WDFY3-AS2 lncRNA is positively correlated with the expression of WDFY3 (R = 0.7; S1A Fig). We further investigated WDFY3 expression in breast tumors from the UALCAN platform, and similar to WDFY3-AS2, low expression of WDFY3 was observed in these tumors (p = 1.7E-12; S2A Fig), being significantly low expressed in the triple-negative subtypes (p = 1,6E-12; S2B Fig). WDFY3 low expression was associated with poor overall and relapse-free survival in breast cancer patients with tumors of the Luminal subtype. However, the survival probabilities between patients with tumors of the basal subtype with low and high WDFY3 expression were not statistically significant (S3 Fig).
We identified BDNF-AS transcripts as down-regulated in breast tumors (Fig 4C). Regarding the PAM50 classification, low BDNF-AS expression was observed in the basal subtype (Fig 4D). The heat map of 104 cases of matched normal and tumor tissue visually shows us a reduced expression of BDNF-AS in normal tissue relative to tumor tissue (S4A Fig). Survival analysis showed that low BNDF-AS expression is associated with worse prognosis in patients with basal breast cancer (Fig 5C and 5D). A correlation analysis was conducted, and a significant result was found between the expression of BDNF-AS and BDNF (S4B Fig; R2 = 0.3; p <0.0001). Reduced BDNF expression was associated with worse disease-free survival and overall survival in breast cancer patients (TNBC, RFS, log-rank p = 0.0094; S5 Fig).
AFAP1-AS1 was observed with high expression in breast tumors compared to the normal tissue (Fig 4E). Regarding the PAM50 classification, AFAP1-AS1 overexpression was observed in the basal subtype compared to the other subtypes (Fig 4F). Survival analysis showed that low AFAP1-AS1 expression is associated with a worse prognosis in patients with basal breast cancer (Fig 5E and 5F). We conducted a correlation analysis with paired data from TANRIC-cBioPortal databases, and statistically significant associations were found (S6 Fig; R2 = 0.33; p <0.001). Survival curves obtained by the KM Plotter platform show that increased expression of AFAP1 transcripts is associated with short disease-free interval in patients with Luminal B (p = 0.0057), HER2 + (p = 0.017), and basal (p = 0.0052) tumors and worse overall survival Luminal A (0.02) and HER2 + (p <0.0001) tumor subtypes.
Discussion
In the present study, we sought to identify differentially expressed lncRNAs from a microarray data set available in the GEO database, generated in a previous study from our group [42]. The differentially expressed lncRNAs were further evaluated using different databases to assess their potential prognostic value for breast cancer patients. The transcripts of three lncRNAs stood out in our analysis, namely WDFY3-AS2, AFAP1-AS1, and BDNF-AS, for showing differential expression in the basal subtype compared to the other subtypes of the PAM50 classification.
Downregulation of WDFY3-AS2 expression has been reported in different types of tumors, including breast cancer [50–52]. WDFY3-AS2 was found down-regulated in glioma and was associated with poor patient survival [51]. An in silico study identified WDFY3-AS2 as down-regulated and associated with a worse prognosis in esophageal cancer [50]. WDFY3-AS2 knockdown resulted in decreased expression of N-cadherin in liver carcinoma cell lines, suggesting that WDFY3-AS2 is associated with the process of epithelial-mesenchymal transition (EMT) in hepatocellular carcinoma (HCC) cells [53]. These authors also demonstrated a significant decrease in the invasive and migratory capability of HCC cells after WDFY3-AS2 knockdown [53]. In ovarian carcinoma, Li et al. [54] identified reduced WDFY3-AS2 expression in tumor tissue compared to adjacent normal tissue [54]. They also demonstrated that WDFY3-AS2 overexpression led to inhibition of cell proliferation, migration, and invasion and decreased protein levels of N-cadherin and increased E-cadherin expression [54]. Recently, Deva Magendhra Rao et al. [52], using RNA sequencing to determine the expression profile of lncRNA in breast cancer, identified WDFY3-AS2 as down-regulated in early-stage breast tumors, which was validated in 52 tumor samples. Performing data mining on the TCGA dataset, these authors also found that altered expression of WDFY3-AS2 transcripts is associated with a worse prognosis for breast cancer patients [52]. Here, we found that low WDFY3-AS2 expression is associated with worse prognosis in breast cancer patients, including those classified as a basal or triple-negative subtype, suggesting that WDFY3-AS2 may act as a tumor suppressor gene for breast cancer. Those data provide new insights on the role played by WDFY3-AS2 in the tumorigenic process, including TNBC breast tumors, which warranted furthers experimental and clinical studies to better understand the potential role of WDFY3-AS2 as a breast cancer biomarker.
BDNF-AS is a tumor suppressor originally described as antisense to the brain-derived neurotrophic factor (BDNF) gene [55–57]. Zhi and Lian [57] identified reduced expression of BDNF-AS in colorectal tumors, and they demonstrated that BDNF-AS overexpression inhibited proliferation and decrease the migratory and invasive capability of LoVo colon cancer cell [57]. On the other hand, BDNF-AS silencing in HCT116 cells resulted in increased proliferation, migration, and invasion, suggesting that BDNF-AS acts as a tumor suppressor in colorectal cancer. The authors also found that BDNF-AS works by regulating the expression of GSK-3β –a known oncogene [57]. BDNF-AS reduced expression was found in osteosarcoma and was associated with worse prognosis in a cohort of 114 patients [55]. HUANG et al. [55] also demonstrated that BDNF-AS overexpression results in the inhibition of proliferation in osteosarcoma cells. A study published by Li et al. [58] identified the down-regulation of BDNF-AS in a cohort of 141 prostate cancer patients [58]. Overexpression of BDNF-AS transcripts in PSA negative and PSA positive prostate cancer cell lines (PC-3 and LNCaP, respectively) resulted in a lower proliferative rate and significantly decreased the invasive capability of prostate cancer cells [58]. In the present study, we found that BDNF-AS transcripts were significantly down-regulated in breast tumors compared with normal tissue, and we further revealed that low BNDF-AS expression is associated with worse prognosis in patients with basal breast cancer subtype. However, further clinical and experimental studies are needed to elucidate the biologic and clinical significance of BDNF-AS expression in breast cancer before to define the clinical utility of this potential biomarker.
The lncRNA AFAP1-AS1 (actin filament associated protein 1 antisense RNA 1) overlaps the exons 2, 14, 15, and 16 of the AFAP1 gene and is the most investigated among lncRNAs found in our study. AFAP1-AS1 is reported in the literature as an oncogene, including in triple-negative breast cancer [59, 60]. In the study by Zhang et al. [53], AFAP1-AS1 up-regulation was associated with TNBC. Analysis of overall survival and disease-free survival in a cohort of 238 patients undergoing mastectomy and chemotherapy revealed that AFAP1-AS1 overexpression was associated with poor prognosis in TNBC patients. In vitro assays, demonstrated that AFAP1-AS1 knockdown resulted in reduced proliferation and invasion and increased apoptosis rates in MDA-MB-231 and BT-549 cell lines [60]. In vivo, overexpression of AFAP1-AS1 promoted tumor growth and activated Wnt/β-catenin pathway to promote tumorigenesis and cell invasion by increasing the expression of C-myc and EMT related genes [60]. Here, we also found overexpression of AFAP1-AS1 in the basal subtype compared to the other breast cancer subtypes; however, our in silico survival analysis revealed that low AFAP1-AS1 expression is associated with a worse prognosis in patients with breast cancer of the basal subtype. These findings imply that altered expression of AFAP1-AS1 transcripts is associated with tumorigenesis in TNBC cancer; however, the role played by AFAP1-AS1 in the tumorigenic process of TNBC remains to be further verified by the larger sample scale.
Our study was able to identify three lncRNAs as potential biomarkers in breast cancer, and which functions have not been already fully understood or elucidated in the context of breast cancer. Although preliminary, our in silico results are promising, and further clinical and functional assays are necessary to understand the particularities of the cellular effect achieved at the expense of modulating the expression of these transcripts, as well as their interaction with other biomolecules, such as microRNAs and transcription factors.
Supporting information
S1 Fig. WDFY3-AS2 expression in breast cancer.
A) correlation analysis between WDFY3 and WDFY3-AS2 expression z-score. Analyses were performed using 1108 cases of breast tumors from cBioPortal. B) Heat Map with paired cases of 104 samples from the TANRIC database.
https://doi.org/10.1371/journal.pone.0232284.s001
(JPG)
S2 Fig. WDFY3 expression in breast cancer.
A) Box plot showing the WDFY3 expression between normal and tumor tissue. B) WDFY3 expression level in breast cancer subtypes. Charts from the UALCAN online platform.
https://doi.org/10.1371/journal.pone.0232284.s002
(TIFF)
S3 Fig. Relapse-free survival and overall survival for patients stratified by the expression of WDFY3.
A, RFS for all breast cancer patients; B, RFS for TNBC patients; C, OS for all breast cancer patients; D, OS for TNBC patients.
https://doi.org/10.1371/journal.pone.0232284.s003
(JPG)
S4 Fig. BDNF-AS expression in breast cancer.
A) correlation analysis between BDNF and BDNF-AS expression z-score. Analyses were performed using 1108 cases of breast tumors from cBioPortal. B) Heat Map with paired cases of 104 samples from the TANRIC database.
https://doi.org/10.1371/journal.pone.0232284.s004
(JPG)
S5 Fig. Relapse-free survival and overall survival for patients stratified by the expression of BNDF.
A, RFS for all breast cancer patients; B, RFS for TNBC patients; C, OS for all breast cancer patients; D, OS for TNBC patients.
https://doi.org/10.1371/journal.pone.0232284.s005
(JPG)
S6 Fig. Correlation analysis between AFAP1 and AFAP1-AS1 expression z-score.
Analyses were performed using 1108 cases of breast tumors from cBioPortal.
https://doi.org/10.1371/journal.pone.0232284.s006
(PNG)
S1 Table. Association of lncRNAs expression with relapse-free survival in breast cancer patients.
https://doi.org/10.1371/journal.pone.0232284.s007
(DOCX)
S2 Table. Association of lncRNAs expression with overall survival in breast cancer patients.
https://doi.org/10.1371/journal.pone.0232284.s008
(DOCX)
S3 Table. Statistical analysis comparing lncRNA expression differences between normal tissue (104 cases) and tumor tissue (828 cases).
https://doi.org/10.1371/journal.pone.0232284.s009
(DOCX)
S4 Table. Statistical analysis comparing differences in lncRNA expression between intrinsic subtypes of the PAM50 classification.
https://doi.org/10.1371/journal.pone.0232284.s010
(DOCX)
S5 Table. Multiple analysis comparing lncRNAs expression in different breast cancer subtypes.
https://doi.org/10.1371/journal.pone.0232284.s011
(DOCX)
S6 Table. Frequency of cases by lncRNA according to expression data available from TANRIC database.
https://doi.org/10.1371/journal.pone.0232284.s012
(DOCX)
S7 Table. Clinicopathological data of breast cancer patients from TANRIC-TCGA bank.
https://doi.org/10.1371/journal.pone.0232284.s013
(DOCX)
References
- 1. Bray F., Ferlay Jacques, Soerjomataram Isabelle, Siegel RL, Torre LA, Jemal A. Global Cancer Statistics 2018: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2018;
- 2. Maxwell C. Biomarker research in breast cancer. Clin J Oncol Nurs. 2010 Dec;14(6):771–83. pmid:21112854
- 3. Berse B, Lynch JA. Molecular diagnostic testing in breast cancer. Semin Oncol Nurs. W.B. Saunders; 2015 May;31(2):108–21.
- 4. Perou CM. Molecular Stratification of Triple-Negative Breast Cancers. Oncologist [Internet]. 2011;16(Supplement 1):61–70. Available from: http://theoncologist.alphamedpress.org/cgi/doi/10.1634/theoncologist.2011-S1-61
- 5. Perou CM, Sørile T, Eisen MB, Van De Rijn M, Jeffrey SS, Ress CA, et al. Molecular portraits of human breast tumours. Nature. 2000 Aug;406(6797):747–52. pmid:10963602
- 6. Sorlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A [Internet]. 2001;98(19):10869–74. Available from: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=11553815%5Cnhttp://www.ncbi.nlm.nih.gov/pmc/articles/PMC58566/pdf/pq010869.pdf pmid:11553815
- 7. Santonja A, Sánchez-Muñoz A, Lluch A, Rosario Chica-Parrado M, Albanell J, Chacón JI, et al. Triple negative breast cancer subtypes and pathologic complete response rate to neoadjuvant chemotherapy. Vol. 9, Oncotarget. 2018.
- 8. Lehmann BD, Jovanović B, Chen X, Estrada M V., Johnson KN, Shyr Y, et al. Refinement of Triple-Negative Breast Cancer Molecular Subtypes: Implications for Neoadjuvant Chemotherapy Selection. Sapino A, editor. PLoS One. 2016 Jun;11(6):e0157368. pmid:27310713
- 9. Shang M, Chang C, Pei Y, Guan Y, Chang J, Li H. Potential Management of Circulating Tumor DNA as a Biomarker in Triple-Negative Breast Cancer. J Cancer. Ivyspring International Publisher; 2018;9(24):4627–34.
- 10. Vikas P, Borcherding N, Zhang W. The clinical promise of immunotherapy in triple-negative breast cancer. Cancer Manag Res. Dove Press; 2018;10:10–6823.
- 11. Lebert JM, Lester R, Powell E, Seal M, McCarthy J. Advances in the systemic treatment of triple-negative breast cancer. Vol. 25, Current Oncology. Multimed Inc.; 2018. p. S142–50.
- 12. Collignon J, Lousberg L, Schroeder H, Jerusalem G. Triple-negative breast cancer: Treatment challenges and solutions. Vol. 8, Breast Cancer: Targets and Therapy. Dove Press; 2016. p. 93–107.
- 13. Neophytou C, Boutsikos P, Papageorgis P. Molecular Mechanisms and Emerging Therapeutic Targets of Triple-Negative Breast Cancer Metastasis. Front Oncol. Frontiers Media SA; 2018;8:31.
- 14. Han L, Zhang H-C, Li L, Li C-X, Di X, Qu X. Downregulation of Long Noncoding RNA HOTAIR and EZH2 Induces Apoptosis and Inhibits Proliferation, Invasion, and Migration of Human Breast Cancer Cells. Cancer Biother Radiopharm. 2018 Aug;33(6):cbr.2017.2432.
- 15. Tian T, Gong Z, Wang M, Hao R, Lin S, Liu K, et al. Identification of long non-coding RNA signatures in triple-negative breast cancer. Cancer Cell Int. BioMed Central; 2018;18:103.
- 16. Wang J, Li B, Wang C, Luo Y, Zhao M, Chen P. Long noncoding RNA FOXD2-AS1 promotes glioma cell cycle progression and proliferation through the FOXD2-AS1/miR-31/CDK1 pathway. J Cell Biochem. 2019 Jul;jcb.29284.
- 17. Xu ST, Xu JH, Zheng ZR, Zhao QQ, Zeng XS, Cheng SX, et al. Long non-coding RNA ANRIL promotes carcinogenesis via sponging miR-199a in triple-negative breast cancer. Biomed Pharmacother. 2017 Dec;96:14–21. pmid:28961506
- 18. Zhang H, Liu C, Yan T, Wang J, Liang W. Long noncoding RNA BDNF-AS is downregulated in cervical cancer and has anti-cancer functions by negatively associating with BDNF. Arch Biochem Biophys. 2018 May;646:113–9. pmid:29572178
- 19. Zheng P, Dong L, Zhang B, Dai J, Zhang Y, Wang Y, et al. Long noncoding RNA CASC2 promotes paclitaxel resistance in breast cancer through regulation of miR-18a-5p/CDK19. Histochem Cell Biol. 2019 Jul;
- 20. Richard JLC, Eichhorn PJA, Lalith Charles Richard J, Johan Adam Eichhorn P, Richard JLC, Eichhorn PJA. Deciphering the roles of lncRNAs in breast development and disease. Oncotarget. Impact Journals; 2018 Apr;9(28):20179–212.
- 21. Balas MM, Johnson AM. Exploring the mechanisms behind long noncoding RNAs and cancer. Non-coding RNA Res. KeAi Publishing; 2018 Sep;3(3):108–17.
- 22. Ransohoff JD, Wei Y, Khavari PA. The functions and unique features of long intergenic non-coding RNA. Nat Rev Mol Cell Biol. NIH Public Access; 2018 Mar;19(3):143–57.
- 23. Tomar D, Yadav AS, Kumar D, Bhadauriya G, Kundu GC. Non-coding RNAs as potential therapeutic targets in breast cancer. Biochim Biophys Acta—Gene Regul Mech. 2019 Apr;
- 24. Youness RA, Gad MZ. Long non-coding RNAs: Functional regulatory players in breast cancer. Non-coding RNA Res. KeAi Publishing; 2019 Mar;4(1):36–44.
- 25. Yang R, Xing L, Wang M, Chi H, Zhang L, Chen J. Comprehensive Analysis of Differentially Expressed Profiles of lncRNAs/mRNAs and miRNAs with Associated ceRNA Networks in Triple-Negative Breast Cancer. Cell Physiol Biochem. 2018;50(2):489–500.
- 26. Gupta RA, Shah N, Wang KC, Kim J, Horlings HM, Wong DJ, et al. Long non-coding RNA HOTAIR reprograms chromatin state to promote cancer metastasis. Nature. 2010 Apr 15;464(7291):1071–6. pmid:20393566
- 27. Yu F, Wang L, Zhang B. Long non-coding RNA DRHC inhibits the proliferation of cancer cells in triple negative breast cancer by downregulating long non-coding RNA HOTAIR. Oncol Lett. Spandidos Publications; 2019 Oct;18(4):3817–22.
- 28. Liang H, Huang W, Wang Y, Ding L, Zeng L. Overexpression of MiR-146a-5p Upregulates lncRNA HOTAIR in Triple-Negative Breast Cancer Cells and Predicts Poor Prognosis. Technol Cancer Res Treat. NLM (Medline); 2019 Jan;18:1533033819882949.
- 29. Miao Z, Ding J, Chen B, Yang Y, Chen Y. HOTAIR overexpression correlated with worse survival in patients with solid tumors. Vol. 107, Minerva Medica. Edizioni Minerva Medica; 2016. p. 392–400.
- 30. Lu R, Zhang J, Zhang W, Huang Y, Wang N, Zhang Q, et al. Circulating HOTAIR expression predicts the clinical response to neoadjuvant chemotherapy in patients with breast cancer. Cancer Biomarkers. IOS Press; 2018;22(2):249–56.
- 31. Gao H, Hao G, Sun Y, Li L, Wang Y. Long noncoding RNA H19 mediated the chemosensitivity of breast cancer cells via wnt pathway and EMT process. Onco Targets Ther. Dove Medical Press Ltd.; 2018;11:8001–12.
- 32. Li L, Jian S, Yupeng Z, Liu S, Yanan P, Zhou Z, et al. MEG3 is a prognostic factor for CRC and promotes chemosensitivity by enhancing oxaliplatin-induced cell apoptosis. Oncol Rep. Spandidos Publications; 2017 Sep;38(3):1383–92.
- 33. Li J, Li L, Yuan H, Huang XW, Xiang T, Dai S. Up-regulated lncRNA GAS5 promotes chemosensitivity and apoptosis of triple-negative breast cancer cells. Cell Cycle. Taylor and Francis Inc.; 2019 Aug;18(16):1965–75.
- 34. Chen W, Xu XK, Li JL, Kong KK, Li H, Chen C, et al. MALAT1 is a prognostic factor in glioblastoma multiforme and induces chemoresistance to temozolomide through suppressing miR-203 and promoting thymidylate synthase expression. Oncotarget. Impact Journals LLC; 2017 Apr;8(14):22783–99.
- 35. Zhang T, Hu H, Yan G, Wu T, Liu S, Chen W, et al. Long Non-Coding RNA and Breast Cancer. Vol. 18, Technology in cancer research & treatment. NLM (Medline); 2019. p. 153303381984388.
- 36. Bawazeer S, Sabry D, Mahmoud RH, Elhanbuli HM, Yassen NN, Abdelhafez MN. Association of SPARC gene polymorphisms rs3210714 and rs7719521 with VEGF expression and utility of Nottingham Prognostic Index scoring in breast cancer in a sample of Egyptian women. Mol Biol Rep. Springer Netherlands; 2018 Dec;45(6):2313–24.
- 37. Ma J, Gao S, Xie X, Sun E, Zhang M, Zhou Q, et al. SPARC inhibits breast cancer bone metastasis and may be a clinical therapeutic target. Oncol Lett. Spandidos Publications; 2017 Nov;14(5):5876–82.
- 38. Arnold SA, Brekken RA. SPARC: A matricellular regulator of tumorigenesis. Vol. 3, Journal of Cell Communication and Signaling. Springer; 2009. p. 255–73.
- 39. dos Santos ML, Palanch CG, Salaorni S, Da Silva WA, Nagai MA. Transcriptome characterization of human mammary cell lines expressing different levels of ERBB2 by serial analysis of gene expression. Int J Oncol. 2006 Jun;28(6):1441–61. pmid:16685446
- 40. dos Santos ML, Gimenes KP, Silva WA, Nagai MA. Transcriptome changes induced by docetaxel in human mammary cell lines expressing different levels of ERBB2. Int J Mol Med. 2009 Jun;23(6):733–43. pmid:19424599
- 41. Nagai MA, Gerhard R, Fregnani JHTG, Nonogaki S, Rierger RB, Netto MM, et al. Prognostic value of NDRG1 and SPARC protein expression in breast cancer patients. Breast Cancer Res Treat. 2011 Feb;126(1):1–14. pmid:20369286
- 42. de Alcantara Filho PRD, Mangone FR, Pavanelli AC, De SA, De Andrade VP. Gene expression profiling of triple-negative breast tumors with different expression of secreted protein acidic and cysteine rich (SPARC). Breast Cancer Manag. 2018 May;7(2):BMT09.
- 43. Györffy B, Lanczky A, Eklund AC, Denkert C, Budczies J, Li Q, et al. An online survival analysis tool to rapidly assess the effect of 22,277 genes on breast cancer prognosis using microarray data of 1,809 patients. Breast Cancer Research and Treatment. 2010.
- 44. Li J, Han L, Roebuck P, Diao L, Liu L, Yuan Y, et al. TANRIC: An interactive open platform to explore the function of IncRNAs in cancer. Cancer Res. American Association for Cancer Research Inc.; 2015 Sep;75(18):3728–37.
- 45. Goldman M, Craft B, Hastie M, Repečka K, McDade F, Kamath A, et al. The UCSC Xena platform for public and private cancer genomics data visualization and interpretation. bioRxiv. Cold Spring Harbor Laboratory; 2019 Sep;(Schroeder 2015):326470.
- 46. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013 Apr;6(269).
- 47. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. The cBio Cancer Genomics Portal: An open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2012 May;2(5):401–4. pmid:22588877
- 48. Chandrashekar DS, Bashel B, Balasubramanya SAH, Creighton CJ, Ponce-Rodriguez I, Chakravarthi BVSK, et al. UALCAN: A Portal for Facilitating Tumor Subgroup Gene Expression and Survival Analyses. Neoplasia (United States). Neoplasia Press, Inc.; 2017 Aug;19(8):649–58.
- 49. Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, et al. STRING v11: Protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. 2019;
- 50. Liu H, Zhang Q, Lou Q, Zhang X, Cui Y, Wang P, et al. Differential Analysis of lncRNA, miRNA and mRNA Expression Profiles and the Prognostic Value of lncRNA in Esophageal Cancer. Pathol Oncol Res. Springer Netherlands; 2019 Apr;1–11.
- 51. Wu F, Zhao Z, Chai R, Liu Y, Wang K, Wang Z, et al. Expression profile analysis of antisense long non-coding RNA identifies WDFY3-AS2 as a prognostic biomarker in diffuse glioma. Cancer Cell Int. BioMed Central; 2018 Dec;18(1):107.
- 52. Deva Magendhra Rao AK, Patel K, Korivi Jyothiraj S, Meenakumari B, Sundersingh S, Sridevi V, et al. Identification of lnc RNA s associated with early-stage breast cancer and their prognostic implications. Mol Oncol. 2019 May;13(6):1878–0261.12489.
- 53. Zhang Z, Wang S, Liu W. EMT-related long non-coding RNA in hepatocellular carcinoma: A study with TCGA database. Biochem Biophys Res Commun. 2018 Sep;503(3):1530–6. pmid:30037433
- 54. Li W, Ma S, Bai X, Pan W, Ai L, Tan W. Long noncoding RNA WDFY3-AS2 suppresses tumor progression by acting as a competing endogenous RNA of microRNA-18a in ovarian cancer. J Cell Physiol. 2019 Jul;jcp.29028.
- 55. Huang Q, Yang J, He X, Shi S, Xing S. LncRNA BDNF-AS is associated with the malignant status and regulates cell proliferation and apoptosis in osteosarcoma. Biosci Rep. 2018 Dec;38(6):BSR20181498.
- 56. Modarresi F, Faghihi MA, Lopez-Toledano MA, Fatemi RP, Magistri M, Brothers SP, et al. Inhibition of natural antisense transcripts in vivo results in gene-specific transcriptional upregulation. Nat Biotechnol. 2012 May;30(5):453–9. pmid:22446693
- 57. Zhi H, Lian J. LncRNA BDNF-AS suppresses colorectal cancer cell proliferation and migration by epigenetically repressing GSK-3β expression. Cell Biochem Funct. 2019 May;cbf.3403.
- 58. Li W, Dou Z, We S, Zhu Z, Pan D, Jia Z, et al. Long noncoding RNA BDNF-AS is associated with clinical outcomes and has functional role in human prostate cancer. Biomed Pharmacother. 2018 Jun;102:1105–10. pmid:29710528
- 59. Dianatpour A, Faramarzi S, Geranpayeh L, Mirfakhraie R, Motevaseli E, Ghafouri-Fard S. Expression analysis of AFAP1-AS1 and AFAP1 in breast cancer. Cancer Biomarkers. 2018 May;22(1):49–54. pmid:29439313
- 60. Zhang K, Liu P, Tang H, Xie X, Kong Y, Song C, et al. AFAP1-AS1 Promotes Epithelial-Mesenchymal Transition and Tumorigenesis Through Wnt/β-Catenin Signaling Pathway in Triple-Negative Breast Cancer. Front Pharmacol. 2018 Nov;9:1248. pmid:30505272