ORIGINAL RESEARCH published: 02 December 2020 doi: 10.3389/fgene.2020.562114 MicroRNA-1291 Is Associated With Locoregional Metastases in Patients With Early-Stage Breast Cancer Daniel Escuin 1* † , Laura López-Vilaró 1,2† , Olga Bell 1 , Josefina Mora 2 , Antonio Moral 2,3 , José Ignacio Pérez 2 , Cristina Arqueros 2 , Teresa Ramón y Cajal 2 , Enrique Lerma 1,2,3 and Agustí Barnadas 1,2,3,4* 1 Institut d’Investigació Biomèdica Sant Pau, Barcelona, Spain, 2 Hospital de la Santa Creu i Sant Pau, Barcelona, Spain, 3 Department of Medicine, Universitat Autònoma de Barcelona, Bellaterra (Cerdanyola del Vallès), Spain, 4 Centro de Investigación Biomédica en Red Cáncer, Madrid, Spain Evidence that microRNAs (miRNAs) regulate the various steps of metastasis is increasing. Several studies have looked at the miRNA expression profile in primary breast tumors but few have compared primary tumor and sentinel lymph node (SLN) metastasis. We correlated the expression of miRNAs with the SLN status and the Edited by: outcome of axillary lymph node dissection (ALND) in 60 patients with early breast cancer. Dongqing Wei, We profiled the expression of miRNAs in paired breast tumor samples and SLNs using Shanghai Jiao Tong University, China the NextSeq500 Illumina platform and key findings were validated by qPCR. MultiMiR Reviewed by: Leda Torres, Bioconductor and Reactome pathways analysis were performed to identify target genes National Institute of Pediatrics, Mexico and signaling pathways affected by altered expressed miRNAs. Our results show that Jesús Espinal-Enríquez, Instituto Nacional de Medicina nine miRNAs were differentially expressed in tumor tissues (q ≤ 0.05). In tumor samples, Genómica (INMEGEN), Mexico a 13.5-fold up-regulation of miR-7641-2 (q < 0.001) and a 2.9-fold down-regulation of *Correspondence: miR-1291 (q < 0.001) were associated with tumors with positive SLNs. However, only Daniel Escuin down-regulation of miR-1291 (q = 0.048) remained significant in paired SLNs samples. descuin@santpau.cat Agustí Barnadas Interestingly, a 10.5 up-regulation of miR-1291 in SLNs samples was associated with abarnadasm@santpau.cat additional axillary lymph node involvement (q < 0.001). The enrichment analyses † These authors have contributed showed that canonical and non-canonical WNT pathways and negative regulation of equally to this work various receptor tyrosine kinases signaling pathways were targets of miR-1291 and Specialty section: supports the role of miR-1291 as a tumor suppressor gene (TSG). Further studies are This article was submitted to warranted to investigate the use of miR-1291 as a surrogate biomarker of SLN node Systems Biology, a section of the journal metastasis in patients with early-stage breast cancer. Frontiers in Genetics Keywords: axillary lymph node dissection, breast cancer, sentinel lymph node, metastasis, microRNAs Received: 20 May 2020 Accepted: 10 November 2020 Published: 02 December 2020 INTRODUCTION Citation: Escuin D, López-Vilaró L, Bell O, MicroRNAs (miRNAs) are a small (19–25 nt) non-coding RNA (ncRNA), expressed in a wide Mora J, Moral A, Pérez JI, variety of organisms and highly conserved across species. MiRNAs regulate the expression of target Arqueros C, Ramón y Cajal T, genes by binding to complementary regions of messenger transcripts to repress their translation Lerma E and Barnadas A (2020) or regulate their degradation. The number of human miRNAs is currently over 1900 and the MicroRNA-1291 Is Associated With Locoregional Metastases in Patients number of predicted target genes are in the range of thousands, with some estimates indicating With Early-Stage Breast Cancer. that miRNAs target over 30% of the human genome. The overall emerging picture is that of a Front. Genet. 11:562114. complex regulation level of gene expression, in which a single miRNA may control hundreds of doi: 10.3389/fgene.2020.562114 targets (Condrat et al., 2020). Frontiers in Genetics | www.frontiersin.org 1 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases Several studies have shown that the expression signatures MATERIALS AND METHODS of miRNAs in human cancers are distinct from those in normal tissues (Janssen et al., 2010). In breast cancer, various Ethics Approval, Consent to Participate studies have revealed a deregulation of miRNAs with clusters and Institutional Safety Procedures of miRNAs frequently being either over-expressed or down- This study was conducted according to the Declaration of regulated (Nana-Sinkam and Croce, 2011; Condrat et al., Helsinki principles, with approval from the Clinical Research 2020). In addition, many cellular pathways are affected by the Ethics Committee at “Institut d’Investigacions Biomèdiques Sant regulatory function of miRNAs and several human pathologies Pau” (IIB Sant Pau). Written informed consent was obtained including cancers, have been associated with deregulation of from all patients under institutional review board-approved the miRNAs (Dai and Ahmed, 2011) and play a pivotal protocols. All methods were performed in accordance with the role in various steps of the metastatic process (Nicoloso relevant guidelines and regulations. et al., 2009; Valastyan, 2012). Furthermore, analysis of miRNA expression correlated these with various clinicopathological factors (Blenkiron et al., 2007). However, few studies have Patient and Study Samples compared primary tumor and lymph node (LN) metastasis This study included 60 patients with early-stage breast cancer (Smeets et al., 2011; Gravgaard et al., 2012; Cascione et al., treated with surgery. None of the patients had prior treatment 2013). Gravgaard et al. (2012) found a differential expression with chemotherapy or radiation. Patients were chosen on the of miRNA-200 family and miRNA-9 in LN associated with the basis of paired biological samples availability and confirmed metastatic process. Cascione et al. (2013) reported analysis of diagnose based on the histopathology of primary tumors 173 formalin-fixed paraffin embedded (FFPE) tumors and 53 and OSNA-diagnosed SLNs. Information about completion matched LN in TNBC. They found two miRNA signatures of ALND was available for all patients. Samples included that were independent predictors for overall survival (OS) hormone receptor (HR) positive [estrogen receptor (ER) or and distant-disease free survival, respectively. In another progesterone receptor (PR) positive], HER2 positive and triple study, Smeets et al. (2011) identified eight with measurable negative (TN) tumors. We collected clinical and pathological differences in gene and miRNA expression between N0 parameters and clinical follow-up. Tumor stage was determined and N+ patients, suggesting that LN involvement is not a according to the AJCC/UICC system (Webber et al., 2014) genetically random process. and histological grade was determined using the Elston–Ellis Herein, we studied the miRNA expression profile in paired grading system (Elston and Ellis, 1991). We collected paired tumor and sentinel lymph node (SLN) from patients with tumor and SLNs for all 60 patients (n = 120 samples). All early breast cancer. We used the one-step nucleic amplification samples were classified according to the SLN status as negative (OSNA) assay (Tsujimoto et al., 2007) to accurately measure total (n = 20) or positive (n = 40). Positive samples were sub- metastatic volume in the SLN (Cserni, 2012), as an alternative classified as macrometastatic (n = 20) or micrometastatic (n = 20) to intraoperative microscopy-based pathological assessment of (Tsujimoto et al., 2007). the SLN. The OSNA assay is a rapid molecular detection of SLN metastasis based on the semi-quantification of cytokeratin Sample Processing 19 (CK19) mRNA copy numbers (Tsujimoto et al., 2007). Primary tumor samples were collected and processed within The CK19 mRNA copy number, also defined as total tumor 30 min after surgery. A portion of each tumor was rapidly load (TTL), has been shown to be a statistically significant embedded in Tissue-Tek OCT (Sakura Finetek, Netherlands) and parameter in predicting the risks of further positive axillary frozen using a histobath (Thermo Fisher Scientific, Waltham, LN (aLN). Thus, only patients diagnosed with more than MA, United States). Frozen tumor blocks were thin-sectioned two macrometastatic SLN are further treated with axillary and stained with hematoxylin/eosin and only those that were lymph node dissection (ALND) (Peg et al., 2013), the golden judged to contain at least 70% viable tumor by area were standard procedure for invasive breast cancer. The aLN status carried on for RNA extraction. Intraoperative SLN were evaluated is the most powerful prognostic factor in breast cancer and by OSNA assay (Tsujimoto et al., 2007). The entire LN was knowledge of this is essential for making decisions about adjuvant homogenized in 4 mL of a lysis solution (Sysmex, Spain) therapy (Shek and Godolphin, 1988). However, ALND has and centrifuged at 10,000 × g at room temperature. A 2 µL been questioned in recent years because of inherent morbidity sample of the supernatant was analyzed with the RD-100i following the procedure without directly contributing to survival system (Sysmex) and the remaining SLN homogenate was stored (Giuliano et al., 2010; Jagsi et al., 2014). immediately at −80C. In this study, we investigated the potential use of miRNAs as surrogate biomarkers for the presence of metastases in the SLN RNA Isolation and the outcome of ALND, in patients with early-stage breast Total RNA was purified using TRIzol (Invitrogen) and mirVana cancer. We found a differential expression of various miRNAs (Ambion) reagents according to manufacturer’s instructions with associated with both the metastatic status of the SLN and the some modifications. Briefly, Trizol homogenates were mixed with occurrence of further aLN metastases. This study provides a new 1/10 volume of mirVana homogenizing solution to retain low framework to study the role of miRNAs in the regulation of tumor molecular weight RNA species. The clear fraction containing total metastases and their impact on patient outcome. RNA was then purified using the mirVana miRNA isolation kit, Frontiers in Genetics | www.frontiersin.org 2 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases following manufacturer’s instructions. Total RNA was eluted in change (log2) between the two experimental groups on the 100 µl nuclease-free water. x-axis. Wherever indicated, we have used fold regulation throughout the text to represent positive FC values as up- RNA-Sequencing regulation (fold regulation is equal to 2FC ) and negative Total RNA concentration and purity was measured using FC values to indicate a down-regulation (fold regulation is the ribogreen RNA assay kit (Life Technologies) and the equal to 2−FC ). integrity was visualized with a TapeStation Bioanalyzer (Agilent Technologies, Inc.). Next, RNA pools were precipitated overnight Quantitative Real-Time RT–PCR with 2× volumes of absolute ethanol and 0.1× volume of 0.3 Validation Analysis M sodium acetate at −80◦ C for cDNA library construction. Selected miRNAs were validated by quantitative real-time RT– Double-stranded cDNA libraries were constructed using the PCR (qPCR) in an ABI Prism 7500 Sequence Detection System NEBNext Small RNA library Prep Set for Illumina (New England using specific LNA PCR primers (Exiqon). The cDNA was Biolabs, Ipswich, MA, United States) following manufacturer’s constructed using the miRCURY LNATM Universal RT cDNA instructions. A quality check (QC) and size selection of Synthesis Kit (Exiqon), diluted 40× and assayed in 10 µl PCR the PCR amplified cDNA construct was performed using reactions according to manufacturer’s instructions. Each qPCR 6% polyacrylamide gel. Two biological replicates for each was assayed in triplicates and a no-template control (NTC) of developmental stage were separately sequenced by the MiSeq water was purified and profiled like the rest of the samples. (Illumina, San Diego, CA, United States) platform using Analysis of the data was performed using the relative miRNA sequenced runs of 2 × 75 paired-end reads and 1000× coverage expression according to the comparative Ct (11Ct) method to ensure proper quantification of the miRNA expression. A total using negative metastatic samples as reference. We used the of 59 tumors and 58 SLNs (117 paired-end, 2 × 75) sequences geNorm (Andersen et al., 2004) or the Normfinder algorithm were successfully sequenced. (Vandesompele et al., 2002) to select the best combination of two reference genes based on our qPCR data. Data from Genome Annotation and Quantification multiples plates were normalized using UniSp3 spike-in as of MiRNAs interplate calibrators. Paired-end (forward-reverse) sample merging and initial bioinformatics analysis were performed with the CLCBio Gene Targets Prediction and Enrichment Genomics Workbench version 8.0.2 (Qiagen, Germany). R Analysis of Gene Targets A total of 234 fastq input files were generated and used in the The multiMiR Bioconductor’s package (Ru et al., 2014) was analyses. The CLCBio software was used to align and map the used to retrieve miRNA-target interactions from 14 external trimmed reads to the human and mouse miRBases (version 19) databases2 and the Reactome pathway database (Jassal et al., 2020) and the Homo_sapiens.GRCh37.57 tracks from Ensembl. Up was use to performed enrichment analysis of target pathways to two mismatches were allowed on the sequences. Mapping and genes.3 options were set as the program’s default. Count tables were generated with R programming language (R Development Core Statistics Team, 2011) and the EdgeR package (Bioconductor repository) Differentially expressed miRNAs obtained by next generation (Robinson et al., 2010), using non-specific filtering for sequences sequencing (NGS) were detected by an exact test based on having a reads-per-million value higher than 0, in at least half conditional maximum likelihood (CML) included in the R of the samples included on each experimental comparison. Bioconductor package edgeR (Robinson et al., 2010). p-Values Transcript per million (TPM) was used as a normalization from NGS were corrected (q-values) for multiple testing using procedure to correct for differences in sequencing depth and to the Benjamini–Hochberg method (Benjamini and Hochberg, quantified RNA species. 1995). A q-value ≤ 0.05 was considered significant. In all group comparisons missing expression values were treated as Differential Expression Analysis zero. Differences in total numbers of miRNAs between groups were analyzed by two-sided parametric t-tests. The analysis of Differential expression analyses were performed using the clinicopathological was performed using the Student’s t-test to trimmed mean of M-values normalization method (TMM) compare quantitative variables, and the X2 or Fisher exact tests (Robinson and Oshlack, 2010), based on the log-fold and to compared qualitative variables. Disease free survival (DFS) absolute gene-wise changes in expression between samples. was defined as the time from diagnosis to date of first relapse Differential expression analysis was performed using the (local, regional, contralateral, or metastatic) or second primary EdgeR statistical software package (Bioconductor1 ). Principal cancer. OS was defined as the time from sample collection component analysis (PCA) was performed using R programming to death resulting from any cause. Patients lost to follow- and TMM-normalized quantifications from defined collections up were censored at the last contact. Kaplan–Meier and log- of samples as input. Volcano plots were constructed plotting rank analyses were used to compare DFS and OS. Differential the p-value (−log10) on the y-axis and the expression fold 2 http://multimir.ucdenver.edu 1 3 http://www.bioconductor.org/ https://reactome.org/ Frontiers in Genetics | www.frontiersin.org 3 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases expression by qPCR of selected miRNAs was analyzed using RESULTS an independent sample t-test with a Levene’s test for equality of variances. The p-values were calculated using a Student’s Patients Characteristics t-test of the replicate 2−1CT values for each miRNA in the Patient and tumor characteristics are summarized in Table 1. different groups compared. A two-sided p-value ≤ 0.05 was We have analyzed 60 patients for whom paired tumor and considered significant. SLN samples were available. A total of 20 patients had negative TABLE 1 | Basic patient and tumor characteristics. Variable Total (%) LN negative (%) LN positive (%) N N (%) 60 (100) 20 (100) 40 (100) Age (years) ≤50 9 (15) 2 (10) 7 (18) >50 51 (85) 18 (90) 33 (82) Mean + SD 63.6 ± 14.2 68.8 ± 11.9 60.9 ± 14.7 Median (range) 64 (26-88) 71 (47-87) 60 (26-88) Tumor stage I 18 (30) 12 (60) 6 (15) II 40 (67) 8 (40) 32 (80) III 2 (3) 0 (0) 2 (5) Tumor status T1 32 (53) 12 (60) 20 (50) T2 27 (45) 8 (40) 19 (48) T3 1 (2) 0 (0) 1 (2) Tumor grade 1 3 (5) 1 (5) 2 (5) 2 35 (58) 14 (70) 21 (52) 3 22 (37) 5 (25) 17 (43) Node status N0 20 (33) 20 (100) 0 (0) N+ 40 (67) 0 (0) 40 (100) SLN OSNA diagnosis Negative 20 (33) 20 (100) 0 (0) Micrometastasis 20 (33) 0 (0) 20 (50) Macrometastasis 20 (33) 0 (0) 20 (50) aLN status* Negative 15 (75) 0 (0) 15 (75) Positive 5 (25) 0 (0) 5 (25) CK19 status** Low 12 (60) 0 (0) 12 (60) High 8 (40) 0 (0) 8 (40) Tumor type Unifocal 36 (60) 13 (65) 23 (58) Multifocal 20 (33) 7 (35) 13 (32) Multicentric 4 (7) 0 (0) 4 (10) ER status Negative 6 (10) 2 (10) 4 (10) Positive 54 (90) 18 (90) 36 (90) PR status Negative 18 (30) 7 (35) 11 (28) Positive 42 (70) 13 (65) 29 (72) Receptor status HR-positive 46 (77) 16 (80) 30 (75) HER-2 positive 9 (15) 3 (15) 6 (15) TN 5 (8) 1 (5) 4 (10) Ki67 status >20% 14 (23) 4 (20) 10 (25) ≤20% 46 (77) 16 (80) 30 (75) Lymphovascular invasion Negative 48 (80) 18 (90) 30 (75) Positive 12 (20) 2 (10) 10 (25) Menopausal status Premenopausal 10 (17) 2 (10) 8 (20) Postmenopausal 50 (83) 18 (90) 32 (80) Breast affected Left 41 (68) 12 (60) 29 (73) Right 19 (32) 8 (40) 11 (27) Breast surgery Mastectomy 19 (32) 5 (25) 14 (35) Lumpectomy 41 (68) 15 (75) 26 (65) aLN, axillary lymph node; ER, estrogen receptor; LN, lymph node; PR, progesterone receptor; TN, triple negative. *CK19 mRNA copy number >15,000 in the subgroup of patients with macrometastatic SLNs. **Presence of additional aLN in the subgroup of patients with macrometastases. Frontiers in Genetics | www.frontiersin.org 4 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases SLNs (33%) and 40 patients had positive SLNs (67%). Of those to negative samples was retained when samples were sub- patients with positive SLNs, 20 (33%) SLNs were diagnosed classified as either macro- or micrometastasis. We observed a as macrometastasis and 20 (33%) SLNs as micrometastasis. All similar deregulation pattern in tumor samples. Macrometastatic patients diagnosed with macrometastasis received ALND of tumors showed an 18.4-fold up-regulation (q < 0.001) of miR- whom 5 (25%) had further aLN involvement. Our study included 7641-2, a 3.2-fold down-regulation of miR-1291 (q = 0.014) 46 patients with HR-positive tumors (77%), 9 patients with HER- and a 2.3-fold down-regulation of miR-6240 (q = 0.038). In 2 positive carcinomas (15%), and 5 patients with TN (8%) tumors. contrast, no differential expression was observed for SLNs diagnosed as macrometastasis (Figure 1C and Supplementary RNA-Sequencing Table 2). Similar results were obtained when we compared Prior to sequencing the samples were subjected to a QC. Three micrometastatic and negative tumors or SLNs (Figure 1D samples (one tumor and two SLNs) were excluded from further and Supplementary Table 2). Interestingly, no significant analyses because they did not pass all of the QC metrics, differences were observed when comparing macrometastatic including the average read quality, the average base quality to micrometastatic tumor or SLNs (Supplementary Table 2), and the read length distribution with a Q-score > 30 (99.9% suggesting that the volume of the metastatic lesion do not correct) (Cock et al., 2010). The remaining 59 tumors and translate in differences in the expression profile of miRNAs. 58 SLNs were successfully sequenced resulting in a total of It has been reported that patients with T1−2 N0−1 invasive 117 paired samples. All samples were sequenced in nine runs non-metastatic breast cancer treated with breast conserving with a minimum and maximum read number of 0.33 and 34.5 surgery and randomized to undergo ALND after SLN dissection million reads number, respectively. This resulted in a median vs. SLN dissection alone, showed no significant differences in 2.5 million read number per sample (Supplementary Table 1). local recurrence or regional recurrence (Giuliano et al., 2010; Following sequence trimming, all reads containing identical Jagsi et al., 2014). Therefore, we investigated the expression of insert sequences were collapsed into a single read, which were miRNAs in tumors (n = 19) or SLNs (n = 20) from patients passed into the analysis pipeline. On average we obtained 0.92 diagnosed as macrometastasis and with additional positive aLNs million reads for each sample resulting in an average genome (n = 5). Our results show that in SLNs samples, a 10.6-fold up- mapping rate of 26.9%. The remaining unmapped reads were regulation of miR-1291 (q < 0.001) was associated with additional usually from degraded RNAs that could not be uniquely mapped. aLN metastases (Figure 1E). In tumor samples, we found a After mapping and counting to relevant entries in mirbase_20 differential expression of miR-182 (p = 0.013) and miR-7641- database, the number of known miRNAs was calculated using 2 (p = 0.029) but none were significant after FDR correction TPM to measure expression. (q > 0.05) (Figure 1E). We have further analyzed this subset of samples in relation to the TTL since it has been reported Differential Expression that a TTL cut-off value >15,000 copies of CK19 is associated To identify differentially expressed miRNAs between paired with additional aLN metastasis (Peg et al., 2013). Our results samples. First, we performed a data reduction analysis to compare show a negative correlation of miR-3535 and miR-3653 in tumors the miRNA expression profile in paired tumor and SLN samples. and a positive association of miR-1291in SLNs in relation to the Our results show that the two types of samples separated into two TTL. However, none of these miRNAs passed the FDR correction different groups, suggesting that the miRNA expression profile (q > 0.05) (data not shown). between paired tumor and SLNs samples from a same patient is Our tumor samples included 46 HR-positive (77%), 9 HER2- different. Other prognostic factor such as tumor stage or tumor positive (15%), and 5 TN (8%) tumors (Table 1). No differential grade were in part responsible for grouping the tumor samples expression of miRNAs was found between molecular subtypes (Supplementary Figure 1). We observed significant differences in tumor samples (n = 59). However, in SLN samples, miR- in the expression of 9 miRNAs (q < 0.05) in tumors samples 1291 showed a 10-fold up-regulation in HER2-positive samples compared to SLNs. Six miRNAs were up-regulated (miR-182, (q = 0.0046) compared to HR-positive SLNs and a 3.5-fold up- miR-1291, miR-3651, miR-6240, miR-7641-1, and miR-6516) and regulation of miR-3607 in HER2-positive SLNs compared to TN three miRNAs were down-regulated (miR-3653, miR-3535, and SLNs (q = 0.033) (Table 2). Next, we investigated the expression miR-3607) (Figure 1A). We further validated the expression for profile of miRNAs according to the molecular subtype and the all nine miRNAs using specific qPCR assays. SLN metastatic status. In TN tumor samples with positive SLNs, Next, we analyzed tumor and SLN samples independently an 18.1 down-regulation of miR-3535 (q = 0.048) was observed in relation to the metastatic SLN status. We first compared compared with HR-positive tumors. A similar non-significant negative (n = 20) vs. positive (n = 40) samples (Figure 1B 20.2-fold down-regulation was observed when comparing TN and Supplementary Table 2). Our results show a 13.9-fold up- with HER-2 positive tumors with positive LNs (q = 0.06). No regulation of miR-7641-2 (q < 0.001) and a fourfold down- differences were observed in SLN samples (data not shown). regulation of miR-1291 (q < 0.001) in positive tumor samples. A non-significant 1.9-fold down-regulation was also observed for miR-6240 (q = 0.076). In contrast, in positive SLNs samples only Gene Target Prediction and Enrichment a 2.8-fold down-regulation of miR-1291 remained significant Analyses (q = 0.048). Next, we determined whether the differential To investigate target genes of miR-1291 we used the multiMiR expression of miRNAs observed in positive samples compared Bioconductor’s package to retrieve validated targets from Frontiers in Genetics | www.frontiersin.org 5 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases FIGURE 1 | MiRNA expression profile in paired tumor and SLN samples from breast cancer patients. (A) Total number of miRNAs with significant differential expression (q < 0.05) in tumor samples compared to paired SLN samples. Data is expressed as log2 fold change relative to tumor samples. *q < 0.001, **q < 0.05. The volcano plots show differentially expressed miRNAs in tumor (blue) and SLN (red) samples according to the locoregional metastatic status: positive vs. negative (B), macrometastasis vs. negative (C), micrometastasis vs. negative (D), and subgroup of patients with macrometastasis that were treated with ALND and grouped according to the aLN status (positive vs. negative) (E). The data show the relationship between the p-values (y-axis) and the fold change (x-axis) between the experimental groups. Only miRNAs with q < 0.05 are shown in the plots. Frontiers in Genetics | www.frontiersin.org 6 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases TABLE 2 | Differential expression of miRNAs in tumor samples (n = 59) and SLN (n = 58) according to the breast cancer molecular subtypes. Sample Comparison miRNA log2 FC p-Value q-Value Fold regulation Tumor HR+ vs. HER2+ mir-21 −1.28 0.012 0.175 2.4 Down-regulation HR+ vs. TNBC mir-3651 1.75 0.053 0.596 3.4 Up-regulation HER2+ vs. TNBC mir-3651 2.13 0.032 0.352 4.4 Up-regulation mir-21 1.88 0.050 0.352 3.7 Up-regulation SLN HR+ vs. HER2+ mir-1291 3.323 0.0005 0.005 10.0 Up-regulation mir-3607 −0.756 0.032 0.143 1.7 Down-regulation HR+ vs. TNBC mir-3607 1.03 0.051 0.319 2.1 Up-regulation mir-6516 1.56 0.074 0.319 3.0 Up-regulation HER2+ vs. TNBC mir-3607 1.79 0.004 0.033 3.5 Up-regulation mir-6516 2.27 0.017 0.075 4.9 Up-regulation mir-3653 1.09 0.044 0.132 2.1 Up-regulation mir-1291 −1.98 0.073 0.164 3.9 Down-regulation FC, fold change. mirecords, mirtarbase, and tarbase databases and predicted DISCUSSION targets from diana-microt, elmmo, microcosm, miranda, mirdb, pictar, pita, and targetscan databases. A total of 171 In this study we have correlated the expression of miRNAs validated targets were found (Supplementary Table 3). with SLN metastatic status and the outcome of ALND. We To investigate the pathways and genes associated with profiled the miRNA expression using NGS and key findings miR-1291 we used the Reactome pathway database. The were validated by qPCR. We analyzed samples from 60 patients analyses show the most relevant pathways and the number with early breast cancer for whom paired primary tumor and of significant target genes within each category (Figure 2A SLNs were available. Overall, we found a poor correlation and Supplementary Table 4), the interactions between these in the miRNA expression profile from SLNs compared to pathways (Figure 2B) and the targets genes associated with match primary tumors samples from the same patient. These these pathways and how these genes and signaling pathways differences are likely due to histological differences between the are linked (Figure 2C). Overall, our results show that the top LN and the tumor. Whereas the SLN is constituted mainly five significant pathways regulated by miR-1291 are signaling by lymphoid and monocytes cells the tumor tissue is formed by WNT, planar cell polarity/convergent extension (PCP/CE) mostly by epithelial and mesenchymal cells. Nonetheless, we pathway, β-catenin independent WNT signaling, diseases of found nine differentially expressed miRNAs in tumor tissues signal transduction and signaling by receptor tyrosine kinases compared to SLN (q ≤ 0.05). Of those, only four miRNAs (RTKs). Interestingly, all target genes within these categories showed a fold de-regulation ≥2.0 (either up-regulation or down- (Figure 2C) were included in the validated mirtabase and tabase regulation). The two most de-regulated miRNAs were miR- databases (Supplementary Table 3). 182 (6.2-fold up-regulation, p < 0.001) and miR-3607 (2.1-fold down-regulation, p < 0.001). Our data agrees with a previous Clinical Status report showing that miR-182 directly targets MIM (Missing Our series include 60 patients with early breast cancer and in Metastasis), which suppresses metastasis by inhibiting Ras we reported recurrence in 10/60 (17%) patients. Four patients homolog family member A (RhoA) activity and stress fiber had locoregional recurrences (breast) and six patients had formation in breast cancer cells (Lei et al., 2014). In addition, distant metastasis (two cases in the liver and one case miR-182 expression was reported to be higher in primary tumors in the lung, pancreas, bone, or brain). At last follow-up, with positive SLNs metastasis and correlated with earlier relapse- three patients with recurrences in their breast were reported free and metastasis-free survivals (Lei et al., 2014). On the other alive without any evidence of disease, whereas 5/10 other hand, miR-3607 has been shown to be widely attenuated in recurrences (breast, lung, bone, and pancreas) were alive prostate cancer (PCa) patients and low expression levels were with disease. We reported seven deaths in our series, three significantly associated with higher PCa stage, Gleason score, deaths were due to complications related to the disease serum PSA levels, tumor progression, and poor survival outcome (two deaths were related to liver metastases and one death (Saini et al., 2014). related to lung metastases), and four deaths non-related to Given that paired tumor and SLN samples in our study showed disease (general deterioration related to age and/or associated different miRNA expression profiles, we analyzed both types of comorbidity). We performed a survival analysis using the samples independently in relation to the locoregional metastatic expression of miR-1291. The Kaplan–Meier and log-rank status. We show that down-regulation of miR-1291 in both tumor analyses showed that the expression levels of miR-1291 (high vs. and SLN samples was associated with positive SLN metastases. low expression) were not significantly associated with DFS nor Interestingly, we found that in the subgroup of positive patients OS (Figure 3). that underwent ALND and had further aLN metastases the Frontiers in Genetics | www.frontiersin.org 7 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases FIGURE 3 | Association of miR-1291 expression with patient outcome. The expression levels of log2-transformed miR-1291 was correlated with (A) disease-free survival and (B) overall survival. Kaplan–Meier curves and log-rank tests are shown. FIGURE 2 | Reactome pathway enrichment in miR-1291 targeted genes. (A) Dot plot graph shows the 15 most significant enriched pathways and the expression of miR-1291 was significantly higher compare to ratio between the genes that belongs to that category and the total number of genes in the pathway. The color of the nodes indicates the p-value and the patient with negative aLNs. However, these results must be size of the nodes the number of genes in the data that belong to that pathway interpreted with caution since only five patients fulfilled those (B) Enrichment map of the top 15 signaling pathways grouped by similarity. characteristics in our dataset and thus, further clinical validation Nodes are colored by p-value and their size reflects the number of genes in a larger cohort of patients will be required to establish these found in that term. (C) Network plot of the genes found in the top five enriched observations. If confirmed, these results would suggest that the signaling pathways showing the linkages between genes and pathways. The size of nodes indicates the number of genes found in that pathway. degree of down-regulation might have an effect on how miR- 1291 modulates its target genes and affect various signaling Frontiers in Genetics | www.frontiersin.org 8 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases pathways. Nonetheless, our study did not correlate the miRNA (FDZ5) and FDZ3 are another non-canonical WNT members expression with the transcriptional expression in the same targeted by miR-1291 that bind to the FZD receptor and leads samples. Therefore we could not determine if miR-1291 function to activation of small Rho GTPases (RAC2) and JNK, which as an oncogene as it has been described for hepatocellular regulate the cytoskeleton and coordinate cell migration and carcinomas and liver cirrhosis patients (Hagag et al., 2020) or in polarity (Schlessinger et al., 2009). Overexpression of FDZ5 microvesicles derived from medulloblastoma cell lines with stem- and FDZ3 has been associated with cancer aggressiveness in like properties (Kaid et al., 2020). Nevertheless, our data suggest human breast carcinomas (da Cunha et al., 2013; Kazanietz the contrary and agrees with previous studies showing that the and Caloca, 2017). Furthermore, we showed that miR-1291 is expression of miR-1291 is significantly down-regulated in various associated with negative regulation of various RTKs pathways, human cancers, including pancreatic cancer (Bi et al., 2014; Tu such as the PI3K/AKT, IGFR1, and FGFR3 signaling pathways et al., 2016) esophageal squamous cell carcinoma (ESCC) (Luo that have a pivotal role in the regulation of cancer proliferation, et al., 2015), PCa (Cai et al., 2019) and renal cell carcinoma (RCC) angiogenesis, and metastasis in breast cancer (Butti et al., 2018). (Yamasaki et al., 2013). In these studies, loss of tumor-suppressive We have identified several well-known oncogenes that participate miR-1291 have been reported to enhanced cell proliferation, in these pathways, including MAPK3, RAF1, AKT2, or TGFA migration, and invasion through up-regulation of various miR- as validated targets of miR-1291. Collectively, our data supports 1291 targets, whereas restoration of miR-1291 expression in the role of miR-1291 as a TSG in the onset of locoregional cancer cell lines and animal models reverted those effects. metastasis in early breast cancer patients. Further research For example, miR-1291 restoration repressed tumorigenesis in is warranted to investigate the expression of miR-1291 as a prostate and pancreatic xenograft tumor models via inhibition of potential surrogate marker of SLN involvement in early breast Mediator of RNA polymerase II transcription subunit 1 (MED1) cancer patients. (Cai et al., 2019), N-methylnicotinamide (NMN) (Bi et al., Our results complement other studies that have compared 2014). In addition, miR-1291 reduced the protein levels of target the miRNA expression profile between primary tumor and LNs genes including ATP Binding Cassette Subfamily C Member 11 from the same patients (Smeets et al., 2011; Gravgaard et al., (ABCC1), Forkhead box protein A2 (FOXA2), Anterior Gradient 2012; Cascione et al., 2013). Gravgaard et al. (2012) found a 2 (AGR2), methyl CpG binding protein 2 (MeCP2) and carnitine differential expression of miRNA-200 family and miRNA-9 in palmitoyltransferase 1C (CPT1C) resulting in the suppression LN associated with the metastatic process. Cascione et al. (2013) of growth and tumorigenesis of human breast and pancreatic reported the analysis of 173 FFPE tumors and 53 matched cell lines (Bi et al., 2014; Li et al., 2015; Tu et al., 2016; LN in TNBC. They found two miRNA signatures (miR-16, Chen et al., 2020). Similar effects have been reported in RCC 155, 125b, 374a and miR-16, 125b, 374a, 374b, 421, 655, 497) through targeting SLC2A1/GLUT1 (Yamasaki et al., 2013) which that were independent predictors for OS and distant-DFS, it has also been reported to be overexpressed in human breast respectively. In another study, Smeets et al. (2011) identified eight carcinomas (da Cunha et al., 2013), in esophageal carcinomas miRNAs (miR-195, miR-191, miR-132, miR-203, miR-431, miR- by inhibiting mucin 1 (MUC1) (Luo et al., 2015), whereas miR- 16, miR-30c, miR-30a) with measurable differences in gene and 1291 acts upstream of the Rho GTPase-activating protein 29 miRNA expression between N0 and N+ patients. However, the (ArhGAP29) to negatively regulate the RhoA/ROCK1 epithelial deregulated miRNAs described here is different from previously mesenchymal transition (EMT) pathway, ultimately leading to published (Smeets et al., 2011; Gravgaard et al., 2012; Cascione endometrial fibrosis (Xu et al., 2017). Interestingly, our target et al., 2013). Various reasons might explain this discrepancy. prediction analysis identified the reported genes as miR-1291 First, the different molecular breast cancer subtypes investigated targets, whereas our enrichment analysis pointed out various in each study. Our cohort of patients comprised mostly luminal signaling pathways and several miR-1291 target genes that are tumors, whereas others analyzed only TNs tumors (Cascione commonly overexpressed in breast cancer, suggesting that in fact et al., 2013) or a mix of molecular subtypes (Smeets et al., 2011; miR-1291 acts as a tumor suppressor gene (TSG). For instance, Gravgaard et al., 2012). Second, the type of starting biological our data shows that miR-1291 targets genes that are involved in material used in previous studies, which was either archive FFPE both canonical and non-canonical WNT signaling and regulate specimens (Gravgaard et al., 2012; Cascione et al., 2013) or frozen many aspects of cell polarity, morphogenesis, and development sections (Smeets et al., 2011). In contrast, we used the entire (Saito-Diaz et al., 2013). In some contexts, both the canonical and fresh SLN homogenate employed in the OSNA assay (Cserni, non-canonical WNT signaling contribute to tumor formation 2012). The OSNA assay was developed as an alternative to intra- by promoting cell migration, invasiveness, and metastasis (van operative microscopy-based pathological assessment of the SLN Amerongen, 2012; VanderVorst et al., 2019). One of the non- because such methods (frozen sections, touch imprints scrapes, canonical WNT pathways we identified in relation to miR-1291 or a combination of these) have a limited ability to accurately is the PCP pathway, which contains core Wnt/PCP components measure total metastatic volume in a LN (Cserni, 2012). Thus, that are overexpressed in a variety of solid tumors and have our study was not limit by sampling bias that might have occurred been directly implicated in promoting tumor cell migration and from using different parts of the LN. Moreover, the OSNA metastasis (VanderVorst et al., 2018). Our data shows that two of technique allows for the immediate sample processing required the Wnt/PCP genes (WNT11 and VANGL2) are targets of miR- to preserve RNA for research purposes. Therefore, we did not 1291 and both have been reported to promote cell motility and encounter the limitations of using RNA extracted from archival metastasis in breast cancer (Luga et al., 2012). The Frizzled-5 FFPE tissues, which has often suffered chemical modification, Frontiers in Genetics | www.frontiersin.org 9 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases cross-linking, and degradation over time as a result of the DATA AVAILABILITY STATEMENT fixation and archiving methods (Esteve-Codina et al., 2017). Finally, a recent study has shown that differences on the protocol The original contributions presented in the study are publicly used for the preparation of the NGS libraries results in over- available. The datasets can be found in the Sequence Research or underrepresentation of miRNAs in the sequencing library Archive under ID PRJNA663033 and are available for download (Coenen-Stass et al., 2018). The use of different NGS library here http://www.ncbi.nlm.nih.gov/bioproject/663033. preparation kits and protocols could result on an inefficient size- selection or in an enrichment of other RNA species therefore decreasing the number of usable miRNAs reads. That could be ETHICS STATEMENT our case, given that the library kit used to generate our RNA- seq data (NEBNext) has been recently reported to produce the The studies involving human participants were reviewed and least mirRNA with the greatest coefficient of variation (Coenen- approved by the Comité Ètic d’Investigació Clínica (CEIC) Stass et al., 2018). In fact, a preliminary analysis of our RNA-seq Hospital de la Santa Creu i Sant Pau Sant Quintí 77-79, 08041 data in tumor samples shows an enrichment in small nucleolar Barcelona, Catalunya, Spain. The patients/participants provided RNAs (snoRNAs) (data not published). Interestingly, several their written informed consent to participate in this study. of the miRNAs described in this study appear to be derived from snoRNAs small ncRNAs (Thorenoor and Slaby, 2015). Other studies have reported human snoRNAs fragments that AUTHOR CONTRIBUTIONS are annotated as miRNAs, including miR-1291 (SNORA2C), miR-1259 (SNORD12B), miR-1248 (SNORA81), miR-3535, and DE, LL-V, and AB were involved in the conceptualization of miR3653 (SNORD125), all of which are reported in our study. the project. DE, LL-V, OB, JM, JP, AM, CA, TR, and EL were Similar reports have described other human snoRNAs showed involved in resources, investigation, and methodology. DE, LL-V, miRNA-like processing signatures that can inhibit the expression EL, and AB were involved in analysis and interpretation of data. of targets genes (Ender et al., 2008; Brameier et al., 2011; All authors have contributed to the writing of the manuscript and Makarova et al., 2013; Thorenoor and Slaby, 2015). Further have critically reviewed it. studies are warranted to understand the role of snoRNAs in the mechanisms of invasion in breast cancer. FUNDING CONCLUSION Instituto de Salud Carlos III and European Regional Development Fund (PI13/00110 to DE); Centro de Investigación To summarize, in this study we have analyzed the miRNA Biomédica en Red Cáncer (CB 16/12/00471 to AB); and Red expression profile in paired tumor and SLN from patients with Temática de Investigación Cooperativa de Cáncer (RTICC; early breast cancer. We found that miR-1291 is differentially RD12/0036/0076 to all authors). expressed in breast cancer patients with positive SLN and further involvement of aLN metastases. Our prediction and enrichment analyses provides a comprehensive understanding ACKNOWLEDGMENTS of the involvement of miR-1291 in promoting metastasis. Altogether, our data supports the role of miR-1291 as We thank the patients and their caregivers for their participation. a TSG. Given the increasing importance of miRNAs in regulating cellular processes and the clinical success of the OSNA assay in detecting LN metastases, further research is SUPPLEMENTARY MATERIAL warranted to investigate the direct implications of our results (Condrat et al., 2020) in treatment decision-making and The Supplementary Material for this article can be found (Janssen et al., 2010) in developing more effective blood-based online at: https://www.frontiersin.org/articles/10.3389/fgene. clinical tests. 2020.562114/full#supplementary-material REFERENCES Bi, H. C., Pan, Y. Z., Qiu, J. X., Krausz, K. W., Li, F., Johnson, C. H., et al. (2014). N- methylnicotinamide and nicotinamide N-methyltransferase are associated with Andersen, C. L., Jensen, J. L., and Orntoft, T. F. (2004). Normalization of real-time microRNA-1291-altered pancreatic carcinoma cell metabolome and suppressed quantitative reverse transcription-PCR data: a model-based variance estimation tumorigenesis. Carcinogenesis 35, 2264–2272. doi: 10.1093/carcin/bgu174 approach to identify genes suited for normalization, applied to bladder and Blenkiron, C., Goldstein, L. D., Thorne, N. P., Spiteri, I., Chin, S. F., Dunning, M. J., colon cancer data sets. Cancer Res. 64, 5245–5250. doi: 10.1158/0008-5472.can- et al. (2007). MicroRNA expression profiling of human breast cancer identifies 04-0496 new markers of tumor subtype. Genome Biol. 8:R214. Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a Brameier, M., Herwig, A., Reinhardt, R., Walter, L., and Gruber, J. (2011). practical and powerful approach to multiple testing. J. R. Stat. Soc. Series B Stat. Human box C/D snoRNAs with miRNA like functions: expanding the range Methodol. 57, 289–300. doi: 10.1111/j.2517-6161.1995.tb02031.x of regulatory RNAs. Nucleic Acids Res. 39, 675–686. doi: 10.1093/nar/gkq776 Frontiers in Genetics | www.frontiersin.org 10 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases Butti, R., Das, S., Gunasekaran, V. P., Yadav, A. S., Kumar, D., and Kundu, Kazanietz, M. G., and Caloca, M. J. (2017). The Rac GTPase in cancer: from old G. C. (2018). Receptor tyrosine kinases (RTKs) in breast cancer: signaling, concepts to new paradigms. Cancer Res. 77, 5445–5451. doi: 10.1158/0008- therapeutic implications and challenges. Mol. Cancer 17:34. 5472.can-17-1456 Cai, Q., Zhao, A., Ren, L., Chen, J., Liao, K., Wang, Z., et al. (2019). MicroRNA- Lei, R., Tang, J., Zhuang, X., Deng, R., Li, G., Yu, J., et al. (2014). Suppression of 1291 mediates cell proliferation and tumorigenesis by downregulating MED1 MIM by microRNA-182 activates RhoA and promotes breast cancer metastasis. in prostate cancer. Oncol. Lett. 17, 3253–3260. Oncogene 33, 1287–1296. doi: 10.1038/onc.2013.65 Cascione, L., Gasparini, P., Lovat, F., Carasi, S., Pulvirenti, A., Ferro, A., et al. Li, M. M., Addepalli, B., Tu, M. J., Chen, Q. X., Wang, W. P., Limbach, P. A., (2013). Integrated MicroRNA and mRNA signatures associated with survival et al. (2015). Chimeric microRNA-1291 biosynthesized efficiently in Escherichia in triple negative breast cancer. PLoS One 8:e55910. doi: 10.1371/journal.pone. coli is effective to reduce target gene expression in human carcinoma cells and 0055910 improve chemosensitivity. Dis. Markers 43, 1129–1136. doi: 10.1124/dmd.115. Chen, Y., Zhou, Y., Han, F., Zhao, Y., Tu, M., Wang, Y., et al. (2020). A novel miR- 064493 1291-ERRα-CPT1C axis modulates tumor cell proliferation, metabolism and Luga, V., Zhang, L., Viloria-Petit, A. M., Ogunjimi, A. A., Inanlou, M. R., Chiu, tumorigenesis. Theranostics 10, 7193–7210. doi: 10.7150/thno.44877 E., et al. (2012). Exosomes mediate stromal mobilization of autocrine Wnt-PCP Cock, P. J., Fields, C. J., Goto, N., Heuer, M. L., and Rice, P. M. (2010). The Sanger signaling in breast cancer cell migration. Cell 151, 1542–1556. doi: 10.1016/j. FASTQ file format for sequences with quality scores, and the Solexa/Illumina cell.2012.11.024 FASTQ variants. Nucleic Acids Res. 38, 1767–1771. doi: 10.1093/nar/gkp1137 Luo, H., Guo, W., Wang, F., You, Y., Wang, J., Chen, X., et al. (2015). miR- Coenen-Stass, A. M. L., Magen, I., Brooks, T., Ben-Dov, I. Z., Greensmith, 1291 targets mucin 1 inhibiting cell proliferation and invasion to promote cell L., Hornstein, E., et al. (2018). Evaluation of methodologies for microRNA apoptosis in esophageal squamous cell carcinoma. Oncol. Rep. 34, 2665–2673. biomarker detection by next generation sequencing. RNA Biol. 15, 1133–1145. doi: 10.3892/or.2015.4206 Condrat, C. E., Thompson, D. C., Barbu, M. G., Bugnar, O. L., Boboc, A., Cretoiu, Makarova, J. A., Ivanova, S. M., Tonevitsky, A. G., and Grigoriev, A. I. (2013). D., et al. (2020). miRNAs as biomarkers in disease: latest findings regarding their New functions of small nucleolar RNAs. Biochem. Biokhim. 78, 638–650. doi: role in diagnosis and prognosis. Cells 9:276. doi: 10.3390/cells9020276 10.1134/s0006297913060096 Cserni, G. (2012). Intraoperative analysis of sentinel lymph nodes in breast cancer Nana-Sinkam, S. P., and Croce, C. M. (2011). MicroRNAs as therapeutic targets in by one-step nucleic acid amplification. J. Clin. Pathol. 65, 193–199. doi: 10. cancer. Transl. Res. 157, 216–225. 1136/jclinpath-2011-200301 Nicoloso, M. S., Spizzo, R., Shimizu, M., Rossi, S., and Calin, G. A. (2009). da Cunha, J. P., Galante, P. A., de Souza, J. E., Pieprzyk, M., Carraro, D. M., Old, MicroRNAs–the micro steering wheel of tumour metastases. Nat. Rev. Cancer L. J., et al. (2013). The human cell surfaceome of breast tumors. Biomed Res. Int. 9, 293–302. doi: 10.1038/nrc2619 2013:976816. Peg, V., Espinosa-Bravo, M., Vieites, B., Vilardell, F., Antunez, J. R., de Salas, Dai, R., and Ahmed, S. A. (2011). MicroRNA, a new paradigm for understanding M. S., et al. (2013). Intraoperative molecular analysis of total tumor load immunoregulation, inflammation, and autoimmune diseases. Transl. Res. 157, in sentinel lymph node: a new predictor of axillary status in early breast 163–179. doi: 10.1016/j.trsl.2011.01.007 cancer patients. Breast Cancer Res. Treat. 139, 87–93. doi: 10.1007/s10549-013- Elston, C. W., and Ellis, I. O. (1991). Pathological prognostic factors in breast 2524-z cancer. I. The value of histological grade in breast cancer: experience from a R Development Core Team (2011). R: A Language and Environment for Statistical large study with long-term follow-up. Histopathology 19, 403–410. doi: 10.1111/ Computing. Vienna: The R Foundation for Statistical Computing. j.1365-2559.1991.tb00229.x Robinson, M. D., McCarthy, D. J., and Smyth, G. K. (2010). edgeR: a Bioconductor Ender, C., Krek, A., Friedlander, M. R., Beitzinger, M., Weinmann, L., Chen, W., package for differential expression analysis of digital gene expression data. et al. (2008). A human snoRNA with microRNA-like functions. Mol. Cell 32, Bioinformatics 26, 139–140. doi: 10.1093/bioinformatics/btp616 519–528. doi: 10.1016/j.molcel.2008.10.017 Robinson, M. D., and Oshlack, A. (2010). A scaling normalization method for Esteve-Codina, A., Arpi, O., Martinez-Garcia, M., Pineda, E., Mallo, M., Gut, M., differential expression analysis of RNA-seq data. Genome Biol. 11:R25. et al. (2017). A comparison of RNA-seq results from paired formalin-fixed Ru, Y., Kechris, K. J., Tabakoff, B., Hoffman, P., Radcliffe, R. A., Bowler, R., et al. paraffin-embedded and fresh-frozen glioblastoma tissue samples. PLoS One (2014). The multiMiR R package and database: integration of microRNA-target 12:e0170632. doi: 10.1371/journal.pone.0170632 interactions along with their disease and drug associations. Nucleic Acids Res. Giuliano, A. E., McCall, L., Beitsch, P., Whitworth, P. W., Blumencranz, P., 42:e133. doi: 10.1093/nar/gku631 Leitch, A. M., et al. (2010). Locoregional recurrence after sentinel lymph node Saini, S., Majid, S., Shahryari, V., Tabatabai, Z. L., Arora, S., Yamamura, S., et al. dissection with or without axillary dissection in patients with sentinel lymph (2014). Regulation of SRC kinases by microRNA-3607 located in a frequently node metastases: the American College of Surgeons Oncology Group Z0011 deleted locus in prostate cancer. Mol. Cancer Ther. 13, 1952–1963. doi: 10.1158/ randomized trial. Ann. Surg. 252, 426–432; discussion 432–433. 1535-7163.mct-14-0017 Gravgaard, K. H., Lyng, M. B., Laenkholm, A. V., Sokilde, R., Nielsen, B. S., Litman, Saito-Diaz, K., Chen, T. W., Wang, X., Thorne, C. A., Wallace, H. A., Page-McCaw, T., et al. (2012). The miRNA-200 family and miRNA-9 exhibit differential A., et al. (2013). The way Wnt works: components and mechanism. Growth expression in primary versus corresponding metastatic tissue in breast cancer. Factors 31, 1–31. doi: 10.3109/08977194.2012.752737 Breast Cancer Res. Treat. 134, 207–217. doi: 10.1007/s10549-012-1969-9 Schlessinger, K., Hall, A., and Tolwinski, N. (2009). Wnt signaling pathways meet Hagag, N. A., Ali, Y. B. M., Elsharawy, A. A., and Talaat, R. M. (2020). Clinical Rho GTPases. Genes Dev. 23, 265–277. doi: 10.1101/gad.1760809 impact of circulated miR-1291 in plasma of patients with liver cirrhosis (LC) Shek, L. L., and Godolphin, W. (1988). Model for breast cancer survival: and hepatocellular carcinoma (HCC): implication on glypican-3 expression. relative prognostic roles of axillary nodal status, TNM stage, estrogen receptor J. Gastrointest. Cancer. 51, 234–241. doi: 10.1007/s12029-019-00234-9 concentration, and tumor necrosis. Cancer Res. 48, 5565–5569. Jagsi, R., Chadha, M., Moni, J., Ballman, K., Laurie, F., Buchholz, T. A., et al. (2014). Smeets, A., Daemen, A., Vanden Bempt, I., Gevaert, O., Claes, B., Wildiers, H., et al. Radiation field design in the ACOSOG Z0011 (Alliance) trial. J. Clin. Oncol. 32, (2011). Prediction of lymph node involvement in breast cancer from primary 3600–3606. doi: 10.1200/jco.2014.56.5838 tumor tissue using gene expression profiling and miRNAs. Breast Cancer Res. Janssen, E. A., Slewa, A., Gudlaugsson, E., Jonsdottir, K., Skaland, I., Soiland, H., Treat. 129, 767–776. doi: 10.1007/s10549-010-1265-5 et al. (2010). Biologic profiling of lymph node negative breast cancers by means Thorenoor, N., and Slaby, O. (2015). Small nucleolar RNAs functioning and of microRNA expression. Mod. Pathol. 23, 1567–1576. doi: 10.1038/modpathol. potential roles in cancer. Tumour Biol. 36, 41–53. doi: 10.1007/s13277-014- 2010.177 2818-8 Jassal, B., Matthews, L., Viteri, G., Gong, C., Lorente, P., Fabregat, A., et al. (2020). Tsujimoto, M., Nakabayashi, K., Yoshidome, K., Kaneko, T., Iwase, T., Akiyama, The reactome pathway knowledgebase. Nucleic Acids Res. 48, D498–D503. F., et al. (2007). One-step nucleic acid amplification for intraoperative detection Kaid, C., Assoni, A., Marçola, M., Semedo-Kuriki, P., Bortolin, R. H., Carvalho, of lymph node metastasis in breast cancer patients. Clin. Cancer Res. 13, V. M., et al. (2020). Proteome and miRNome profiling of microvesicles derived 4807–4816. from medulloblastoma cell lines with stem-like properties reveals biomarkers Tu, M. J., Pan, Y. Z., Qiu, J. X., Kim, E. J., and Yu, A. M. (2016). MicroRNA- of poor prognosis. Brain Res. 1730:146646. doi: 10.1016/j.brainres.2020.146646 1291 targets the FOXA2-AGR2 pathway to suppress pancreatic cancer cell Frontiers in Genetics | www.frontiersin.org 11 December 2020 | Volume 11 | Article 562114 Escuin et al. MiRNAs Role in Locoregional Breast Metastases proliferation and tumorigenesis. Oncotarget 7, 45547–45561. doi: 10.18632/ continuous literature review. Int. J. Cancer 135, 371–378. doi: 10.1002/ijc.2 oncotarget.9999 8683 Valastyan, S. (2012). Roles of microRNAs and other non-coding RNAs in breast Xu, Q., Duan, H., Gan, L., Liu, X., Chen, F., Shen, X., et al. (2017). MicroRNA-1291 cancer metastasis. J. Mammary Gland Biol. Neoplasia 17, 23–32. doi: 10.1007/ promotes endometrial fibrosis by regulating the ArhGAP29-RhoA/ROCK1 s10911-012-9241-9 signaling pathway in a murine model. Mol. Med. Rep. 16, 4501–4510. doi: van Amerongen, R. (2012). Alternative Wnt pathways and receptors. Cold Spring 10.3892/mmr.2017.7210 Harb. Perspect. Biol. 4:a007914. doi: 10.1101/cshperspect.a007914 Yamasaki, T., Seki, N., Yoshino, H., Itesako, T., Yamada, Y., Tatarano, S., VanderVorst, K., Dreyer, C. A., Konopelski, S. E., Lee, H., Ho, H. H., and Carraway, et al. (2013). Tumor-suppressive microRNA-1291 directly regulates glucose K. L. III (2019). Wnt/PCP signaling contribution to carcinoma collective cell transporter 1 in renal cell carcinoma. Cancer Sci. 104, 1411–1419. migration and metastasis. Cancer Res. 79, 1719–1729. doi: 10.1158/0008-5472. can-18-2757 Conflict of Interest: The authors declare that the research was conducted in the VanderVorst, K., Hatakeyama, J., Berg, A., Lee, H., and Carraway, K. L. III (2018). absence of any commercial or financial relationships that could be construed as a Cellular and molecular mechanisms underlying planar cell polarity pathway potential conflict of interest. contributions to cancer malignancy. Semin. Cell Dev. Biol. 81, 78–87. doi: 10.1016/j.semcdb.2017.09.026 Copyright © 2020 Escuin, López-Vilaró, Bell, Mora, Moral, Pérez, Arqueros, Ramón Vandesompele, J., De Preter, K., Pattyn, F., Poppe, B., Van Roy, N., De Paepe, y Cajal, Lerma and Barnadas. This is an open-access article distributed under the A., et al. (2002). Accurate normalization of real-time quantitative RT-PCR terms of the Creative Commons Attribution License (CC BY). The use, distribution data by geometric averaging of multiple internal control genes. Genome Biol. or reproduction in other forums is permitted, provided the original author(s) and 3:RESEARCH0034. the copyright owner(s) are credited and that the original publication in this journal Webber, C., Gospodarowicz, M., Sobin, L. H., Wittekind, C., Greene, F. L., Mason, is cited, in accordance with accepted academic practice. No use, distribution or M. D., et al. (2014). Improving the TNM classification: findings from a 10-year reproduction is permitted which does not comply with these terms. Frontiers in Genetics | www.frontiersin.org 12 December 2020 | Volume 11 | Article 562114