Introduction
Breast cancer (BCa) is a leading cause of cancer-related death among women in the United States and is a major public health concern worldwide, with an estimated 1.7 million newly diagnosed cases in 2015 1 . The American Cancer Society estimates about 246,660 BCa cases and 61,000 non-invasive new cases will lead to 40,450 US deaths in 2016 1 . Triple-Negative Breast Cancer (TNBC) is a subtype of BCa defined by the lack of estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2) expression and represents a focus of increasing interest 2 . This is due in part to the aggressive nature of TNBC, which is noted to have higher recurrence and lower survival rates than other BCa subtypes and the lack of targeted therapies 3 . There has been intensified interest in further understanding their molecular profile to identify unique therapeutic target sites which could improve overall survival and quality of life in patients diagnosed with TNBC.
Technological advances in next generation sequencing (NGS) have helped to profile the whole transcriptomic expression of diseased biological systems at the molecular level and increase our fundamental understanding. Small RNAs are non-coding RNAs (sncRNA) consisting of 17-250 nucleotides in length that perhaps play a crucial role in disease development 4 . Secondary to accumulating NGS investigations, a nearly comprehensive repertoire of small RNAs i.e. miRNAs (17-22 nucleotides) 5 , piRNAs (26-33 nucleotides) 6 , lncRNAs (more than 200 nucleotides) 7 and small nuclear/nucleolar RNAs (70-120 nucleotides) 8 has been collected and their roles in disease development analyzed. Over the last decade there has been a great deal of focus on revealing miRNAs role in transcriptional regulation. Furthermore, piwi-interacting RNAs (piRNA) are the largest class of the small non-coding RNA family and are implicated in epigenetic and post transcriptional regulation, however most of their functions are still unknown 9 . Long non-coding RNAs (lncRNA) are a diverse class of RNAs that are believed to be functional; however, little biological relevance has been established thus far 10 . snoRNAs are perhaps the most ancient and highly conserved class of sncRNAs present in eukaryotes which carry out a fundamental role of modification and processing of ribosomal RNAs (rRNA), transfer RNAs (tRNA) and small nuclear RNAs (snRNA) 11 . Two well-known classes of snoRNAs are C/D snoRNAs and box H/ACA snoRNAs which primarily differ in sequence and structure 11 .
The present study focuses on in-depth analysis of small RNA sequencing data obtained from TNBC patient samples compared to normal tissue. We aimed to identify differential molecular signature expressions in whole small RNA groups. We further sought to identify the most predominant non-coding RNA expressions along with the corresponding molecular pathways that may be involved in the development of TNBC and serve as biomarkers.
Materials and Methods
Samples and Data Assembly
Triple-negative breast cancer (TNBC) small RNA sequencing raw sample datasets (project no: PRJNA172756; GEO: GSE40049 ) were downloaded from the NIH bioproject 12 which included 24 TNBC human tissue samples (stage I=7; stage II=15; stage III=2) and 14 normal adjacent samples. Detailed clinical information about the patients was reported by Chang et al. 12 . The small RNA library was constructed from total RNA using the SOLiD Total RNA-Seq Kit (Applied Biosystems, Foster City, CA, USA). This method utilizes SOLiD's Adaptor Mix which is a set of oligonucleotides that can bind to an RNA molecule with a 5'-monophosphate and a 3'-hydroxyl end; therefore, most small RNAs can participate in this reaction, and intact mRNA molecules with a 5' cap structure are excluded. The ligated small RNAs were reverse transcribed, size-selected, and amplified, before SOLiD sequencing. Raw files were downloaded as s equence r aw a rchive (SRA) files then converted to FASTQ, using the SRA toolkit version 2.5.7. Data was assembled using PartekFlow ® software, version 5.0 (Partek, Inc., St. Louis, MO, USA). Converted FASTQ files were uploaded to the PartekFlow ® server and remapped to human genome hg19. Transcript abundances were determined and expression values were represented using reads per million (RPM) mapped reads which normalizes for sequencing depth. All small RNA with expression RPM values >1 in at least 10% of the samples was considered robustly expressed and used for further analysis. Expression matrices were aligned to clinicopatholgical features in order to compare miRNA, piRNA, lncRNA and sn/snoRNA levels for association with specific TNBC phenotypes. Statistical analyses were carried out using the non-parametric Mann-Whitney U test followed by false discovery rate (FDR) correction through the Benjamin-Hochberg method, with the default p -value <0.05 considered statistically significant 13 . A Circos plot 14 was generated for differential expression of all small RNAs.
Assembly of miRNA, piRNA, lncRNA and snoRNA Annotations
Small RNA sequencing data was trimmed and aligned to the whole human genome (hg19), and BWA-0.7.12 aligner (BWA-MEM) with a few modifications (mismatch penalty 2, gap open penalty 6, clipping penalty 4 and alignment score cutoff 15) for short read mapping. miRNAs were annotated from miRBase version 21 ( http://www.mirbase.org/ ), which contains more than 1900 high confidence miRNAs 15 . piRNA data was generated and annotated from piRBase ( http://regulatoryrna.org/database/piRNA ), which is manually curated with a focus on piRNA functional analysis 16 . lncRNAs were quantified using reference annotation LNCipedia ( http://www.lncipedia.org ) version 3.1, downloaded from all coordinates relative to the hg19 reference genome 17 . Total small RNA (including miRNA, piRNA, snRNA, scRNA, snoRNA, piRNA, tRF3, tRF5, tRNA, rRNA) was annotated using Gencode version 19 annotation file ( www.gencodegenes.org ), which provides comprehensive information on human small non-coding RNAs, with specific regards to small nuclear and nucleolar RNAs.
Biological Processes and Gene Network Visualization by MetaCore
Biological pathway interactions of small RNA expression were analyzed using MetaCore pathway analysis of differentially expressed genes (Thomson Reuters, New York, NY) 18 with p < 0.05 from each group (TNBC vs adjacent normal tissue; along with different stages). Functional gene networks were built based on differentially regulated gene lists as input to generate disease biomarkers and GO terms (Data analyzed by Gene Arrays, Entity of Vedic Research, Inc., New York, USA).
Statistical Analysis
All experiments calculated p -values using paired student's t test, with p < 0.05 considered statistically significant. Furthermore, the Benjamini and Hochberg multiple testing adjustment method has been applied for all small RNA sequencing studies and pathway analysis.
Results and Discussion
miRNA Expression in Triple Negative Breast Cancer Tissue Samples
The miRNA raw data from patients with TNBC and adjacent normal tissue samples (control) were first downloaded from GEO series accession number GSE40049 which was then converted to FASTQ files and uploaded to the PartekFlow server for miRNA data analysis. We identified 99 miRNAs enriched in TNBC vs adjacent normal tissues that were statistically significant ( p < 0.05). Applying more stringent statistical analysis ( p < 0.005) to create hierarchical clustering, we found 55 dysregulated miRs ( p < 0.005; Fig. 1 A) of which eight were upregulated and 47 miRs were downregulated. Top five upregulated/downregulated miRNAs are shown in Figure 1 B, C, respectively. Our miRNA findings were in line with that originally published by Chang et al 12 i.e. the top upregulated microRNA was miR-301b in TNBC (2.73 fold, p < 0.0001) along with miR-183 ( p < 0.004), miR-18a ( p < 0.003), miR-3074 ( p < 0.003) and miR-96 ( p < 0.001). While, miR-122, miR-204, miR-135a-2, miR-139 and miR-215 were substantially downregulated in TNBC samples (Figure 1 B). In other findings, miR-301 was highly upregulated in hepatocellular cancer 19 , while specific miR-301b promotes cell invasiveness through targeting TP63 in pancreatic carcinoma cells 20 . Similarly, miR-183, a cluster (miR-96 and miR-182) of family genes, which is a potent prognostic marker for lung adenocarcinoma and other cancers, was found to be highly upregulated in our TNBC samples 21 - 27 . Likewise, a well-known onco-miR miR-18a belongs to miR-17-92 cluster that consists of seven miRNAs (miR-17-5p, miR-17-3p, miR-18a, miR-19b, miR-20a and miR-92) 28 and is considered as a biomarker for several cancers, particularly, esophageal, nasopharyngeal, colorectal and gastric cancer 29 - 34 . Another miR that was highly dysregulated in TNBC samples was miR-96. Previous studies have shown miR-96 as a potential biomarker for bladder, colorectal, non-small cell lung cancer and acute myeloid leukemia 35 - 40 . On the other hand, many of the miRs were found to be significantly downregulated in TNBC samples compared to adjacent normal tissues. The top five downregulated miRs, miR-122 (~45 fold, p < 0.00003), miR-204 (~26 fold, p < 0.0000001), miR-135a-2 (16 fold, p < 0.0000003), miR-139 (~14 fold, p < 0.0000001) and miR-215 (~11 fold, p < 0.000000001), are listed in Figure 1 C. Stage-wise analysis of TNBC samples identified significantly dysregulated miRs in TNBC (sub-stages with differential expression of miRs are shown as a volcano plot in Figure 1 D). We found 64 miRs in stage I, 63 miRs in stage II and 15 miRs in stage III, whereas, we found 36 miRs differentially expressed commonly in stage I & II, three miRs in stage II & III and only one miRs in stage I & III. Overall, there were six miRs commonly regulated in all three stages; miR-135a-2, miR-16-1, miR-215, miR-301b, miR-486 and miR-517b (Venn diagram in Fig. 1 E). The reduction of miR-122 (Fig. 1 C), which is known as a crucial member with a wide variety of functionality 41 , is consistent with the recent similar study in breast cancer specimens 42 and in hepatocellular carcinoma and with poor prognosis and metastasis 41 , 43 - 49 . Furthermore, miR-204, which was downregulated in TNBC (Fig. 1 C), is considered to have tumor suppressor activity 50 . The low expression was observed in the breast cancer cells, prostate cancer cells, hepatoma, esophageal and gastric cancer cells 51 - 54 with notable correlation with advancement of cancer stages. We similarly plotted the differentially expressed miRNAs by both p -value (0.05 cutoff) and fold change (2 fold) for the stage-by-stage comparisons (Fig. 1 D). We further classified differentially expressed miRNAs in each stage (stages I, II and III) of the TNBC (Fig. 1 E). Chang et al, showed only 25 miRNAs that were significantly affected, however, in our data analysis, we found 98 miRs [ p < 0.05; 55 miRs ( p < 0.005)]. Out of the 25 miRs identified by Chang et al ., we were able to identify 22 of them along with an additional 76 more miRs.
Following the miRNA expression analysis, using MetaCore software, we looked at various biological pathways to identity the molecular signature of the most affected and crucial signaling proteins in TNBC pathogenesis. Pathway analysis of differentially expressed biological pathways in TNBC compared to adjacent normal tissue samples clearly showed involvement of signaling molecules in the cell regulation, cell proliferation, migration and angiogenesis of TNBC (Fig. 2 A). Response to amino acid was most significantly over-represented by Gene Ontology (GO) Biological Processes analysis (Fig. 2 B) and many cancer-related categories were over-represented by Disease Stages by Biomarkers analysis (Fig. 2 C). Enrichment analysis of the miRNA in TNBC GO Localization analysis revealed utmost of the miRNA located in the cytoplasm and intracellular compartments (Fig. 2 D). While GO Molecular Function analysis showed enrichment of protein binding and a functional pathway was depicted around let 7a-1 and let 7a-2 miRNAs, which includes Interleukin-1 alpha, NSF, DSPP, MMP-11 and MMP-19 as extracellular regulators, by way of intercellular signaling molecules such as Caspase 3, ERK1/2, p38/MAPK, and PDE, and then resulting in involving nuclear transcription factors such as GATA Group, SP3, BRD4/NUT fusion protein, RBP-J kappa, and NF-AT (Fig. 2 E, F). In-depth pathway analysis showed the TGF-beta signaling pathway to be majorly affected in the TNBC samples (Fig. 2 G). Biological network analysis of miRNA involved with the highest affected network processes were shown in Figure 2 H-K, respectively.
Differentially Expressed piRNAs in Triple Negative Breast Cancer
Piwi-interacting RNAs (piRNAs) are the largest class of endogenous non-coding small RNAs. The piRNAs have been recently shown to play important biological roles as RNA silencers, where piwi-proteins form RNA-protein complexes and are required for both epigenetic and post-transcriptional gene silencing of retrotransposons and other genetic elements in germ line cells, particularly during spermatogenesis 55 . Aligning small RNA sequencing data and remapping with piRBase annotation that contains thousands of the piRNAs, we identified more than 139 differentially expressed piRNAs that were statistically significant ( p < 0.05; Fig. 3 A-C) in TNBC samples compared to normal adjacent tissue samples. Out of 139, 103 piRNAs were upregulated and 36 piRNAs were downregulated in TNBC samples. A number of groups have directed their attention to understand the biological and epigenetic functions of piRNAs. Since, detection of piRNAs in cancer correlates with poorer prognostics and clinical outcomes, suggesting that piRNAs play an important functional role in the cancer biogenesis, there is an ongoing effort to utilize piRNAs for developing diagnostic, prognostic and therapeutic tools. However, until the gap in our understanding of mechanisms underlying piRNA biogenesis and functions is filled, it will be difficult to entirely discriminate between a 'passenger' role for the ectopic expression of piRNAs and PIWI proteins in cancer from a 'driver' role in the pathogenesis of these diseases 56 . We listed the top five upregulated piRNAs (Fig. 3 B) i.e.piR-21131 (8 fold, p < 0.00004), piR-32745 (6 fold, p < 0.00001), piR-21131 (5 fold, p < 0.003), piR-1282 (5 fold, p < 0.0003) and piR-23672 (4 fold, p < 0.0006) and downregulated (Fig. 3 C) piR-23662 (15 fold, p < 0.003), piR-26526 (10 fold, p < 0.00003), piR-26527 (9 fold, p < 0.00009), piR-30293 (9 fold, p < 0.00009) and piR-26528 (8 fold, p < 0.000008) in TNBC compared to normal tissue. Recent studies have also reported the involvement of specific piRNAs in different cancers, for example, piR-4987, piR-20365, piR-20485, piR-20582 and piR-932 in breast cancer 57 , 58 , piR-823 and piR651 in gastric cancer 59 , 60 , upregulated piR-32051, piR-39894, piR-43607 and downregulated piR-38756, piR-57125, piR-30924 in kidney cancer 61 , piR-Hep1 in liver cancer 62 , downregulated piR-017061 in pancreatic cancer 63 , downregulated piR-823 in multiple myeloma, downregulated piR-L-163 in lung cancer 64 , and piR-59056, piR-32105 and piR-58099 in colon cancer 65 . Analyzing stage-wise differential expression of piRNAs, we found 46 piRNAs were shared between stages I & II; three shared piRNAs in stage II & III whereas no common piRNAs in stages I & III. However, in all three stages eight piRNAs were differentially expressed (Fig. 3 D).
Differentially Expressed Long Non-coding RNAs in TNBC Tissue Samples
Long non-coding RNAs (lncRNAs) are the most recent and the least characterized class of the sncRNAs family, being larger than 200 nucleotides and non-conserved among species 55 . lncRNAs are known to have tissue-specific expression in a regulated manner correlatingwith distinct gene sets that influence cellular function 66 , in addition to, acting as a tumor suppressor or promoter 67 - 69 . Our small RNA sequencing data remapped to identify differentially expressed lncRNAs in TNBC patient's tissue samples found 258 lncRNAs ( p <0.05) significantly expressed. When more stringent p -values ( p < 0.001) were applied, we identified 61 lncRNAs (Fig. 4 A) where 28 lncRNAs were upregulated (Top five: lnc-DNAJC16-1:1, lnc-SC5DL-3:1, lnc-PURA-2:1, lnc-EIF2C2-1:1 and lnc-ELP4-3:1; Fig. 4 B) and 33 lncRNAs were downregulated (Top five: lnc-PAPLN-2:1, lnc-FLT3LG-1:7, lnc-NEK8-2:1, lnc-FLOT2-1:1 and lnc-ZNF75D-2:2; Fig. 4 C). lncRNAs are a heterogeneous groups of transcripts with diverse mechanisms and are differentially expressed in many diseases including cancer 68 . DNAJC16 belonging to the DnaJ Heat shock protein family, functions in protein translation, translocation and degradation and this lnc-DNAJC16-1:1 was significantly upregulated (17 fold, p < 0.0001) in TNBC tissue samples. PURA is a sequence-specific, multi-functional single-stranded-DNA/RNA-binding protein and RNA-binding protein which can act as a transcriptional activator and repressor and we observed lnc-PURA-2:1 (5 fold, p < 0.002) upregulated in TBNC, which has not been reported in the literature, however, recently, Inoue et al ., showed PURA over expression leads to activation of PSA production in PC3 cells 70 . We furthermore analyzed the data for stage wise assertion of lncRNA profiling and found 160 lncRNAs in stage I, 155 lncRNAs in stage II and 27 lncRNAs in stage III, whereas, 79 were commonly regulated in stage I & II, only one lncRNA in stage II & III and 3 lncRNAs in stage I & III. Overall, in all three stages, 16 lncRNAs were differentially regulated (Fig. 4 D). Further investigation is needed to check their functions in TNBC and other diseases.
Differentially Expressed snRNA and snoRNAs in TNBC Tissue Samples
Small nuclear RNAs (snRNAs) form a class of RNA molecules that localize within the nucleus of eukaryotic cells 55 . Their primary function is pre-mRNA processing, for which they are always associated with a set of specific proteins. These complexes are referred to as small nuclear ribonucleoproteins (snRNP). The small nucleolar RNAs (snoRNAs) are another subclass of snRNA that are localized in the nucleolus and are associated with the maturation of RNA molecules through chemical modifications targeting mainly rRNAs, tRNAs and snRNAs 55 . We remapped aligned reads to GenCode database, which contains most of the curated small RNAs in order to specifically look for snRNAs and snoRNAs. We identified 28 snRNAs ( p < 0.05; Fig. 5 A) and 123 snoRNAs ( p < 0.05; Fig. 5 D). Out of 28 snRNAs, only three snRNAs were downregulated (Fig. 6 C; RNU11, RNU2-7P and RNU2-48P) and the remaining 25 were upregulated (Top five upregulated snRNAs are listed in Fig. 5 B; i.e. RNU1-149P, RNU4-19P, RNU1-77P, RNU4-58P and RNU1-136P). U1 small nuclear RNA is a multigene family located on the small arm of chromosome 1 and in addition to this, U1 snRNA pseudogenes are present throughout the genome 71 . Many of the snRNAs which are upregulated in TNBC samples indicated that these snRNAs may be involved in oncogenesis, however, this needs to be further validated and investigated to be developed as biomarkers. RNU11 plays a role in splicing the U12-dependent class of eukaryotic nuclear introns and was found to be downregulated in TNBC samples. Although previous studies have shown the RNU family of snRNAs are expressed in pancreatic, colorectal and lung cancers, no concrete evidence is available to define the function of RNU2 snRNAs 72 .
Accumulating evidence through recent reports indicate the role of snoRNA in regulation of cell fate and tumorigenesis 73 . We identified 123 snoRNAs that were differentially affected in the TNBC samples ( p < 0.05; Fig. 5 D), where only 4 out of 123 snoRNAs (scaRNA9, snoU2_19, snoRD88C and SNHG6) were upregulated (Fig. 5 E) while the remaining 119 snoRNAs were downregulated (Fig. 5 F). Recent studies have shown that, the SNGH family of snoRNAs plays a crucial role in oncogenesis. For example, Zhang et al., have shown long noncoding RNA SNHG1 affects tumor suppressor gene p53, thus promoting hepatocellular carcinoma tumorigenesis. They proposed SNHG1 could serve as a predictor of poor prognosis 74 . Additionally, Zhao et al, showed SNHG5 suppresses gastric cancer progression by trapping MTA2 in the cytosol 75 . Conversely, in our study, SNHG6 was found to be upregulated in TNBC samples. Similarly, snoU2_19, a modifying snRNA (function as modification of other snRNAs) was also found to be upregulated in the TNBC tissue samples. In our analysis downregulated snoRNAs spotted were snoRD113 (~10 fold, p < 0.0000001, multiple transcript types), snoRD114 (~9 fold, p < 0.0000001), snoRD42A (~9 fold, p < 0.0001), snoRD96A (~8 fold, p < 0.000001) and snoRD34 (~7 fold, p < 0.0000001). snoRD113, snoRD114 and snoRD116, that belong to the C/D box class of snoRNAs were found to be remarkably downregulated in TNBC samples. The snoRNA113, which was significantly downregulated in TNBC samples has been previously reported to suppress tumorigenesis in hepatocellular carcinoma 76 . While it was downregulated in our TNBC samples, snoRD114 was previously reported to be overexpressed in acute promyelocytic leukemia (APL) 77 , suggesting, tissue specific action of this cluster of snoRNAs. The role of the snoRD113, snoRD114 and snoRD116 cluster region is becoming an important molecular target 78 . For example, genetic mapping of snoRD116 has been recently reported to be absent or change its expression and affect other snoRNAs in Prader-Willi syndrome (PWS), a complex rare genetic condition caused by a loss of normally expressed genes on chromosome 15 from the paternal parent 79 , 80 . It is known that snoRD42A regulates snoRD116 clusters 81 , and in our TNBC samples, both snoRNAs (snoRD42A and snoRD116) were found to be drastically under expressed. Purified box C/D snoRNPs were predicted to function in site specific methylation of substrate RNA 82 . At the moment, we do not understand the biological and clinical implication of this finding, however, it is anticipated that better understanding of these molecules will lead to finding a prognostic marker or novel drug target for specific diseases.
Differential Expression of rRNAs, miscRNAs and Nonsense Mediated Decay (NMD) RNAs in TNBC
We extended our analysis to other different types of small non-coding RNAs, i.e. rRNAs, miscRNAs and nonsense mediated decay (NMD) RNAs. Ribosomal RNAs are known to be essential tools for protein biosynthesis 83 , 84 . In our analysis, we found that 13 rRNAs were differentially expressed ( p < 0.05; Fig. 6 A), with 11 upregulated (RNA5SP358, RNA5SP214, RNA5SP150, RNA5S9, RNA5SP204, RNA5SP248, RNA5SP500, RNA5-8SP2, RNA5SP207, RNA5SP132 and RNA5SP19; Fig. 6 B) and two downregulated (RNA5SP444 and RNA5SP211; Fig. 6 C). Presently, none of these rRNAs have been reported to have any specific function in any known disease status. Very few studies have been reported in this area and so further in-depth investigation is needed to link the disease status and the rRNA expression.
We also identified the 53 differentially expressed miscRNAs, and most of the them are described as Y RNAs that are required for DNA replication 85 ( p < 0.05; Fig. 6 D) in TNBC tissue samples. Only four miscRNAs were found to be downregulated (Fig. 6 F), while the remaining 49 miscRNAs were significantly upregulated (Fig. 6 E).
Nonsense-mediated decay RNAs (nmdRNAs) are involved in surveillance pathways in all eukaryotes and reducing errors in gene expression by eliminating mRNA transcripts that contain premature stop codons. However, in some cases translation of these aberrant mRNAs could lead to deleterious dominant-negative activity or gain-of-function of the resulting abnormal proteins 86 . In our data analysis, we found 20 substantially deregulated nmdRNAs, out of which, two nmdRNAs were downregulated and 18 nmdRNAs were upregulated (Fig. 6 G).
Further validation and clinical studies on large clinical samples are needed to confirm the biological function of dysregulated small non-coding RNAs identified in this study and to use as future prognostic, diagnostic and therapeutic biomarkers in TNBC.
We further designed a Circos plot for comprehensive genomic data visualization allowing for decreased graph complexity and better readability. Circos plot (v0.69-4) was used to generate plots, upregulated and downregulated genes shown as darker and lighter colors respectively. Comprehensive data visualization shows, enrichment of each sncRNAs type in chromosome view. Most of the differential expression of miRNAs was enriched in chromosome 14 and chromosome 11 while expressions of piRNAs were enriched in chromosome 14, chromosome 1 and chromosome 16. Further observing all the data, lncRNAs were enriched on chromosome 11, chromosome 1 and chromosome 17 whereas sn/snoRNAs on chromosome 15, chromosome 14 and chromosome 1. Overall, the most affected chromosomes in all the sncRNAs were on chromosomes 14, 11 and 1. Extending inward from the outer ring of the Circos plot, miRNAs were followed in order by piRNAs, lncRNAs, snRNAs, snoRNAs, rRNAs, miscRNAs and nmdRNAs. Black lines inside the respective rings shows the affected gene location on chromosome number (Fig. 7 ).
Conclusion
High-throughput sequencing technologies have effectively enabled the unbiased and sensitive detection of non-coding RNAs in disease as well as in normal cellular biology. It is presumed that small RNA plays a crucial role in disease network and pathogenesis. However, our understanding is presently limited mostly towards the biological role of small RNAs, and little is known about the expression pattern of non-coding RNAs in pathological conditions. We believe that by identifying the specific molecular gene signature in distinct disease states will allow for better diagnosis and optimization of treatment modalities. In summary, the present study analyzed all sncRNA sequencing data from TNBC and adjacent normal tissue samples for the first time. We have identified several differentially regulated miRNAs, piRNAs, lncRNAs and sn/snoRNAs in TNBC that could serve as new biomarkers for easy and early disease detection while offering new therapeutic targets, pending experimental validation. Our investigation demonstrates a comprehensive screening of non-coding RNA signature in publicly available NGS resources. Our approach seeks to maximize the utilization of established datasets to understand the biological role of non-coding RNAs in tumorigenesis, disease prognosis and treatment outcome.
We thank Gene Arrays (An entity of Vedic Research, Inc., USA) for MetaCore data analysis free of charge. This work is supported by the Department of Surgery at Penn State Milton S Hershey Medical Center. Authors also want to acknowledge Mr. Clark Royden & Ms. Anna Salzberg for Circos plot technical assistance and Kimberly Walker for manuscript formatting to journal style.
1 Siegel RL Miller KD Jemal A Cancer statistics, 2016 CA: A Cancer Journal for Clinicians 2016 66 7 30 26742998 10.3322/caac.21332
2 Irvin Jr WJ Carey LA What is triple-negative breast cancer? European Journal of Cancer 2008 44 2799 805 19008097 10.1016/j.ejca.2008.09.034
3 Kaplan N Moore IK Fondufe-Mittendorf Y Gossett AJ Tillo D Field Y The DNA-encoded nucleosome organization of a eukaryotic genome Nature 2009 458 362 6 19092803 10.1038/nature07667 PMC2658732
4 Eddy SR Non-coding RNA genes and the modern RNA world Nat Rev Genet 2001 2 919 29 11733745 10.1038/35103511
5 Lee RC Feinbaum RL Ambros V The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14 Cell 1993 75 843 54 8252621 10.1016/0092-8674(93)90529-y
6 Sai Lakshmi S Agrawal S piRNABank: a web resource on classified and clustered Piwi-interacting RNAs Nucleic Acids Research 2008 36 D173 D7 17881367 10.1093/nar/gkm696 PMC2238943
7 Cao J The functional role of long non-coding RNAs and epigenetics Biological Procedures Online 2014 16 11 25276098 10.1186/1480-9222-16-11 PMC4177375
8 Scott MS Ono M Yamada K Endo A Barton GJ Lamond AI Human box C/D snoRNA processing conservation across multiple cell types Nucleic Acids Research 2011 10.1093/nar/gkr1233 PMC3333852 22199253
9 Thomson T Lin H The Biogenesis and Function PIWI Proteins and piRNAs: Progress and Prospect Annual review of cell and developmental biology 2009 25 355 76 10.1146/annurev.cellbio.24.110707.175327 PMC2780330 19575643
10 Derrien T Johnson R Bussotti G Tanzer A Djebali S Tilgner H The GENCODE v7 catalog of human long noncoding RNAs: Analysis of their gene structure, evolution, and expression Genome Res 2012 22 1775 89 22955988 10.1101/gr.132159.111 PMC3431493
11 Scott MS Ono M From snoRNA to miRNA: Dual function regulatory non-coding RNAs Biochimie 2011 93 1987 92 21664409 10.1016/j.biochi.2011.05.026 PMC3476530
12 Chang Y-Y Kuo W-H Hung J-H Lee C-Y Lee Y-H Chang Y-C Deregulated microRNAs in triple-negative breast cancer revealed by deep sequencing Molecular Cancer 2015 14 36 25888956 10.1186/s12943-015-0301-9 PMC4351690
13 Firmino N Martinez VD Rowbotham DA Enfield KS Bennewith KL Lam WL HPV status is associated with altered PIWI-interacting RNA expression pattern in head and neck cancer Oral Oncol 2016 55 43 8 26852287 10.1016/j.oraloncology.2016.01.012 PMC4808439
14 Krzywinski M Schein J Birol İ Connors J Gascoyne R Horsman D Circos: An information aesthetic for comparative genomics Genome Research 2009 19 1639 45 19541911 10.1101/gr.092759.109 PMC2752132
15 Kozomara A Griffiths-Jones S miRBase: annotating high confidence microRNAs using deep sequencing data Nucleic Acids Research 2014 42 D68 D73 24275495 10.1093/nar/gkt1181 PMC3965103
16 Zhang P Si X Skogerbø G Wang J Cui D Li Y piRBase: a web resource assisting piRNA functional study. Database 2014 2014 10.1093/database/bau110 PMC4243270 25425034
17 Volders P-J Verheggen K Menschaert G Vandepoele K Martens L Vandesompele J An update on LNCipedia: a database for annotated human lncRNA sequences Nucleic Acids Research 2015 43 D174 D80 25378313 10.1093/nar/gku1060 PMC4383901
18 Sehgal K Guo X Koduru S Shah A Lin A Yan X Plasmacytoid dendritic cells, interferon signaling, and FcgammaR contribute to pathogenesis and therapeutic response in childhood immune thrombocytopenia Sci Transl Med 2013 5 193ra89 10.1126/scitranslmed.3006277 23843450
19 He K Hu Z Ruan J Ma Q Zhong F Cheng X MicroRNA301 is a potential diagnostic biomarker for hepatocellular cancer International Journal of Clinical and Experimental Pathology 2015 8 5603 8 26191271 PMC4503142
20 Funamizu N Lacy CR Parpart ST Takai A Hiyoshi Y Yanaga K MicroRNA-301b promotes cell invasiveness through targeting TP63 in pancreatic carcinoma cells International journal of oncology 2014 44 725 34 24398967 10.3892/ijo.2014.2243
21 Zhu W Liu X He J Chen D Hunag Y Zhang YK Overexpression of members of the microRNA-183 family is a risk factor for lung cancer: A case control study BMC Cancer 2011 11 393 - 21920043 10.1186/1471-2407-11-393 PMC3183044
22 Zhang Q-H Sun H-M Zheng R-Z Li Y-C Zhang Q Cheng P Meta-analysis of microRNA-183 family expression in human cancer studies comparing cancer tissues with noncancerous tissues Gene 2013 527 26 32 23791657 10.1016/j.gene.2013.06.006
23 Xu F Zhang H Su Y Kong J Yu H Qian B Up-regulation of microRNA-183-3p is a potent prognostic marker for lung adenocarcinoma of female non-smokers Clinical and Translational Oncology 2014 16 980 5 24805982 10.1007/s12094-014-1183-9
24 Yuan D Li K Zhu K Yan R Dang C Plasma miR-183 predicts recurrence and prognosis in patients with colorectal cancer Cancer Biology & Therapy 2015 16 268 75 25629978 10.1080/15384047.2014.1002327 PMC4622949
25 Zhang Q Di W Dong Y Lu G Yu J Li J High serum miR-183 level is associated with poor responsiveness of renal cancer to natural killer cells Tumor Biology 2015 36 9245 9 26091793 10.1007/s13277-015-3604-y
26 Chen J Gu L Ni J Hu P Hu K Shi Y-L MiR-183 Regulates ITGB1P Expression and Promotes Invasion of Endometrial Stromal Cells BioMed Research International 2015 2015 340218 26357653 10.1155/2015/340218 PMC4556833
27 Pak MG Lee C-H Lee W-J Shin D-H Roh M-S Unique microRNAs in lung adenocarcinoma groups according to major TKI sensitive EGFR mutation status Diagnostic Pathology 2015 10 99 26170125 10.1186/s13000-015-0339-4 PMC4501046
28 KOMATSU S ICHIKAWA D TAKESHITA H MORIMURA R HIRAJIMA S TSUJIURA M Circulating miR-18a: A Sensitive Cancer Screening Biomarker in Human Cancer In Vivo 2014 28 293 7 24815829
29 Su Z-X Zhao J Rong Z-H Wu Y-G Geng W-M Qin C-K Diagnostic and prognostic value of circulating miR-18a in the plasma of patients with gastric cancer Tumor Biology 2014 35 12119 25 25416437 10.1007/s13277-014-2516-6
30 Tsujiura M Komatsu S Ichikawa D Shiozaki A Konishi H Takeshita H Circulating miR-18a in plasma contributes to cancer detection and monitoring in patients with gastric cancer Gastric Cancer 2015 18 271 9 24626859 10.1007/s10120-014-0363-1
31 Yau TO Wu CW Dong Y Tang CM Ng SSM Chan FKL microRNA-221 and microRNA-18a identification in stool as potential biomarkers for the non-invasive diagnosis of colorectal carcinoma British Journal of Cancer 2014 111 1765 71 25233396 10.1038/bjc.2014.484 PMC4453736
32 Luo Z Dai Y Jiang C Li Z Yang J McCarthy JB miR-18a promotes malignant progression by impairing microRNA biogenesis in nasopharyngeal carcinoma Carcinogenesis 2013 34 415 25 23097559 10.1093/carcin/bgs329
33 Hirajima S Komatsu S Ichikawa D Takeshita H Konishi H Shiozaki A Clinical impact of circulating miR-18a in plasma of patients with oesophageal squamous cell carcinoma British Journal of Cancer 2013 108 1822 9 23579215 10.1038/bjc.2013.148 PMC3658511
34 Zheng Y Li S Ding Y Wang Q Luo H Shi Q The role of miR-18a in gastric cancer angiogenesis Hepato-gastroenterology 2013 60 1809 13 24624454
35 Li J Li P Chen T Gao G Chen X Du Y Expression of microRNA-96 and its potential functions by targeting FOXO3 in non-small cell lung cancer Tumor Biology 2015 36 685 92 25286764 10.1007/s13277-014-2698-y
36 Zhao J Lu Q Zhu J Fu J Chen Y-x Prognostic value of miR-96 in patients with acute myeloid leukemia Diagnostic Pathology 2014 9 76 - 24678958 10.1186/1746-1596-9-76 PMC3975266
37 Wang Y Luo H Li Y Chen T Wu S Yang L hsa-miR-96 up-regulates MAP4K1 and IRS1 and may function as a promising diagnostic marker in human bladder urothelial carcinomas Molecular medicine reports 2012 5 260 5 21993544 10.3892/mmr.2011.621
38 Gao F Wang W MicroRNA-96 promotes the proliferation of colorectal cancer cells and targets tumor protein p53 inducible nuclear protein 1, forkhead box protein O1 (FOXO1) and FOXO3a Molecular medicine reports 2015 11 1200 6 25369914 10.3892/mmr.2014.2854
39 Eissa S Habib H Ali E Kotb Y Evaluation of urinary miRNA-96 as a potential biomarker for bladder cancer diagnosis Medical Oncology 2014 32 1 7 10.1007/s12032-014-0413-x 25511320
40 Ress AL Stiegelbauer V Winter E Schwarzenbacher D Kiesslich T Lax S MiR-96-5p influences cellular growth and is associated with poor survival in colorectal cancer patients Molecular Carcinogenesis 2015 54 1442 50 25256312 10.1002/mc.22218
41 Bandiera S Pfeffer S Baumert TF Zeisel MB miR-122 - A key factor and therapeutic target in liver disease Journal of Hepatology 2015 62 448 57 25308172 10.1016/j.jhep.2014.10.004
42 Wang B Wang H Yang Z MiR-122 Inhibits Cell Proliferation and Tumorigenesis of Breast Cancer by Targeting IGF1R PLoS ONE 2012 7 e47053 23056576 10.1371/journal.pone.0047053 PMC3466252
43 Hsu S-h Wang B Kota J Yu J Costinean S Kutay H Essential metabolic, anti-inflammatory, and anti-tumorigenic functions of miR-122 in liver The Journal of Clinical Investigation 2012 122 2871 83 22820288 10.1172/JCI63539 PMC3408748
44 Coulouarn C Corlu A Glaise D Guénon I Thorgeirsson SS Clément B Hepatocyte-Stellate Cell Cross-Talk in the Liver Engenders a Permissive Inflammatory Microenvironment That Drives Progression in Hepatocellular Carcinoma Cancer Research 2012 72 2533 42 22419664 10.1158/0008-5472.CAN-11-3317 PMC3498759
45 Bai S Nasser MW Wang B Hsu S-H Datta J Kutay H MicroRNA-122 inhibits tumorigenic properties of hepatocellular carcinoma cells and sensitizes these cells to sorafenib Journal of Biological Chemistry 2009 284 32015 27 19726678 10.1074/jbc.M109.016774 PMC2797273
46 Fornari F Gramantieri L Giovannini C Veronese A Ferracin M Sabbioni S MiR-122/Cyclin G1 Interaction Modulates p53 Activity and Affects Doxorubicin Sensitivity of Human Hepatocarcinoma Cells Cancer Research 2009 69 5761 7 19584283 10.1158/0008-5472.CAN-08-4797
47 Nassirpour R Mehta PP Yin M-J miR-122 Regulates Tumorigenesis in Hepatocellular Carcinoma by Targeting AKT3 PLoS ONE 2013 8 e79655 24244539 10.1371/journal.pone.0079655 PMC3820664
48 Xu J Zhu X Wu L Yang R Yang Z Wang Q MicroRNA-122 suppresses cell proliferation and induces cell apoptosis in hepatocellular carcinoma by directly targeting Wnt/β-catenin pathway Liver International 2012 32 752 60 22276989 10.1111/j.1478-3231.2011.02750.x
49 Nakao K Miyaaki H Ichikawa T Antitumor function of microRNA-122 against hepatocellular carcinoma Journal of Gastroenterology 2014 49 589 93 24531873 10.1007/s00535-014-0932-4
50 Lee H Lee S Bae H Kang H-S Kim SJ Genome-wide identification of target genes for miR-204 and miR-211 identifies their proliferation stimulatory role in breast cancer cells Scientific reports 2016 6 10.1038/srep25287 PMC4848534 27121770
51 Turner DP Findlay VJ Moussa O Semenchenko VI Watson PM LaRue AC Mechanisms and functional consequences of PDEF protein expression loss during prostate cancer progression The Prostate 2011 71 1723 35 21446014 10.1002/pros.21389 PMC3128180
52 Ge Y Yan X Jin Y Yang X Yu X Zhou L fMiRNA-192 and miRNA-204 Directly Suppress lncRNA HOTTIP and Interrupt GLS1-Mediated Glutaminolysis in Hepatocellular Carcinoma PLoS Genet 2015 11 e1005726 26710269 10.1371/journal.pgen.1005726 PMC4692503
53 Sun Y Yu X Bai Q miR-204 inhibits invasion and epithelial-mesenchymal transition by targeting FOXM1 in esophageal cancer International journal of clinical and experimental pathology 2015 8 12775 26722467 PMC4680412
54 Li W Jin X Zhang Q Zhang G Deng X Ma L Decreased expression of miR-204 is associated with poor prognosis in patients with breast cancer Int J Clin Exp Pathol 2014 7 3287 92 25031750 PMC4097245
55 Morceau F Chateauvieux S Gaigneaux A Dicato M Diederich M Long and Short Non-Coding RNAs as Regulators of Hematopoietic Differentiation International Journal of Molecular Sciences 2013 14 14744 70 23860209 10.3390/ijms140714744 PMC3742271
56 Moyano M Stefani G piRNA involvement in genome stability and human cancer Journal of Hematology & Oncology 2015 8 10.1186/s13045-015-0133-5 PMC4412036 25895683
57 Huang G Hu H Xue X Shen S Gao E Guo G Altered expression of piRNAs and their relation with clinicopathologic features of breast cancer Clinical and Translational Oncology 2013 15 563 8 23229900 10.1007/s12094-012-0966-0
58 Zhang H Ren Y Xu H Pang D Duan C Liu C The expression of stem cell protein Piwil2 and piR-932 in breast cancer Surgical Oncology 2013 22 217 23 23992744 10.1016/j.suronc.2013.07.001
59 Cheng J Deng H Xiao B Zhou H Zhou F Shen Z piR-823, a novel non-coding small RNA, demonstrates in vitro and in vivo tumor suppressive activity in human gastric cancer cells Cancer Letters 2012 315 12 7 22047710 10.1016/j.canlet.2011.10.004
60 Cui L Lou Y Zhang X Zhou H Deng H Song H Detection of circulating tumor cells in peripheral blood from patients with gastric cancer using piRNAs as markers Clinical Biochemistry 2011 44 1050 7 21704610 10.1016/j.clinbiochem.2011.06.004
61 Busch J Ralla B Jung M Wotschofsky Z Trujillo-Arribas E Schwabe P Piwi-interacting RNAs as novel prognostic markers in clear cell renal cell carcinomas Journal of Experimental & Clinical Cancer Research CR 2015 34 10.1186/s13046-015-0180-3 PMC4467205 26071182
62 Lee JH Schütte D Wulf G Füzesi L Radzun H-J Schweyer S Stem-cell protein Piwil2 is widely expressed in tumors and inhibits apoptosis through activation of Stat3/Bcl-XL pathway Human Molecular Genetics 2006 15 201 11 16377660 10.1093/hmg/ddi430
63 Müller S Raulefs S Bruns P Afonso-Grunz F Plötner A Thermann R Next-generation sequencing reveals novel differentially regulated mRNAs, lncRNAs, miRNAs, sdRNAs and a piRNA in pancreatic cancer Mol Cancer 2015 14 10.1186/s12943-015-0358-5 PMC4417536 25910082
64 Mei Y Wang Y Kumari P Shetty AC Clark D Gable T A piRNA-like small RNA interacts with and modulates p-ERM proteins in human somatic cells Nature Communications 2015 6 10.1038/ncomms8316 PMC4557300 26095918
65 Ng KW Anderson C Marshall EA Minatel BC Enfield KSS Saprunoff HL Piwi-interacting RNAs in cancer: emerging functions and clinical utility Molecular Cancer 2016 15 1 13 26768585 10.1186/s12943-016-0491-9 PMC4714483
66 Guttman M Amit I Garber M French C Lin MF Feldser D Chromatin signature reveals over a thousand highly conserved large non-coding RNAs in mammals Nature 2009 458 223 7 19182780 10.1038/nature07672 PMC2754849
67 Huarte M Guttman M Feldser D Garber M Koziol MJ Kenzelmann-Broz D A large intergenic non-coding RNA induced by p53 mediates global gene repression in the p53 response Cell 2010 142 409 19 20673990 10.1016/j.cell.2010.06.040 PMC2956184
68 Huarte M The emerging role of lncRNAs in cancer Nat Med 2015 21 1253 61 26540387 10.1038/nm.3981
69 Zheng GXY Do BT Webster DE Khavari PA Chang HY Dicer-microRNA-Myc circuit promotes transcription of hundreds of long noncoding RNAs Nat Struct Mol Biol 2014 21 585 90 24929436 10.1038/nsmb.2842 PMC5509563
70 Inoue T Maeno A Talbot C Zeng Y Yeater DB Leman ES Purine-rich element binding protein (PUR) α induces endoplasmic reticulum stress response, and cell differentiation pathways in prostate cancer cells The Prostate 2009 69 861 73 19267365 10.1002/pros.20936
71 O'Reilly D Dienstbier M Cowley SA Vazquez P Drożdż M Taylor S Differentially expressed, variant U1 snRNAs regulate gene expression in human cells Genome research 2013 23 281 91 23070852 10.1101/gr.142968.112 PMC3561869
72 Köhler J Schuler M Gauler TC Nöpel-Dünnebacke S Ahrens M Hoffmann A-C Circulating U2 small nuclear RNA fragments as a diagnostic and prognostic biomarker in lung cancer patients Journal of Cancer Research and Clinical Oncology 2016 142 795 805 26687686 10.1007/s00432-015-2095-y PMC11819072
73 Chen L Han L Wei J Zhang K Shi Z Duan R SNORD76, a box C/D snoRNA, acts as a tumor suppressor in glioblastoma Scientific reports 2015 5 10.1038/srep08588 PMC5390076 25715874
74 Zhang M Wang W Li T Yu X Zhu Y Ding F Long noncoding RNA SNHG1 predicts a poor prognosis and promotes hepatocellular carcinoma tumorigenesis Biomedicine & Pharmacotherapy 2016 80 73 9 27133041 10.1016/j.biopha.2016.02.036
75 Zhao L Guo H Zhou B Feng J Li Y Han T Long non-coding RNA SNHG5 suppresses gastric cancer progression by trapping MTA2 in the cytosol Oncogene 2016 10.1038/onc.2016.110 27065326
76 Xu G Yang F Ding C-L Zhao L-J Ren H Zhao P Small nucleolar RNA 113-1 suppresses tumorigenesis in hepatocellular carcinoma Molecular Cancer 2014 13 216 25217841 10.1186/1476-4598-13-216 PMC4169825
77 Liuksiala T Teittinen KJ Granberg K Heinaniemi M Annala M Maki M Overexpression of SNORD114-3 marks acute promyelocytic leukemia Leukemia 2014 28 233 6 23979522 10.1038/leu.2013.250
78 Bieth E Eddiry S Gaston V Lorenzini F Buffet A Conte Auriol F Highly restricted deletion of the SNORD116 region is implicated in Prader-Willi Syndrome Eur J Hum Genet 2015 23 252 5 24916642 10.1038/ejhg.2014.103 PMC4297892
79 Cruvinel E Budinetz T Germain N Chamberlain S Lalande M Martins-Taylor K Reactivation of maternal SNORD116 cluster via SETDB1 knockdown in Prader-Willi syndrome iPSCs Human Molecular Genetics 2014 23 4674 85 24760766 10.1093/hmg/ddu187 PMC4481691
80 Falaleeva M Surface J Shen M de la Grange P Stamm S SNORD116 and SNORD115 change expression of multiple genes and modify each other's activity Gene 2015 572 266 73 26220404 10.1016/j.gene.2015.07.023 PMC5586535
81 Lestrade L Weber MJ snoRNA-LBME-db, a comprehensive database of human H/ACA and C/D box snoRNAs Nucleic Acids Research 2006 34 D158 D62 16381836 10.1093/nar/gkj002 PMC1347365
82 Galardi S Fatica A Bachi A Scaloni A Presutti C Bozzoni I Purified Box C/D snoRNPs Are Able To Reproduce Site-Specific 2′-O-Methylation of Target RNA In Vitro Molecular and cellular biology 2002 22 6663 8 12215523 10.1128/MCB.22.19.6663-6668.2002 PMC134041
83 Schultz J Wolf M ITS2 sequence-structure analysis in phylogenetics: A how-to manual for molecular systematics Molecular Phylogenetics and Evolution 2009 52 520 3 19489124 10.1016/j.ympev.2009.01.008
84 Son DJ Kumar S Takabe W Woo Kim C Ni C-W Alberts-Grill N The atypical mechanosensitive microRNA-712 derived from pre-ribosomal RNA induces endothelial inflammation and atherosclerosis Nature Communications 2013 4 10.1038/ncomms4000 PMC3923891 24346612
85 Christov CP Gardiner TJ Szüts D Krude T Functional Requirement of Noncoding Y RNAs for Human Chromosomal DNA Replication Molecular and cellular biology 2006 26 6993 7004 16943439 10.1128/MCB.01060-06 PMC1592862
86 Holbrook JA Neu-Yilik G Hentze MW Kulozik AE Nonsense-mediated decay approaches the clinic Nature genetics 2004 36 801 8 15284851 10.1038/ng1403