Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.


  • Login to see the comments

  • Be the first to like this


  1. 1. Dear Author, Here are the final proofs of your article. Please check the proofs carefully. All communications with regard to the proof should be sent to Please note that at this stage you should only be checking for errors introduced during the production process. Please pay particular attention to the following when checking the proof: - Author names. Check that each author name is spelled correctly, and that names appear in the correct order of first name followed by family name. This will ensure that the names will be indexed correctly (for example if the author’s name is ‘Jane Patel’, she will be cited as ‘Patel, J.’). - Affiliations. Check that all authors are cited with the correct affiliations, that the author who will receive correspondence has been identified with an asterisk (*), and that all equal contributors have been identified with a dagger sign (†). - Ensure that the main text is complete. - Check that figures, tables and their legends are included and in the correct order. - Look to see that queries that were raised during copy-editing or typesetting have been resolved. - Confirm that all web links are correct and working. - Ensure that special characters and equations are displaying correctly. - Check that additional or supplementary files can be opened and are correct. Changes in scientific content cannot be made at this stage unless the request has already been approved. This includes changes to title or authorship, new results, or corrected values. How to return your corrections Returning your corrections via online submission: - Please provide details of your corrections in the online correction form. Always indicate the line number to which the correction refers. Returning your corrections via email: - Annotate the proof PDF with your corrections. - Send it as an email attachment to: - Remember to include the journal title, manuscript number, and your name when sending your response via email. After you have submitted your corrections, you will receive email notification from our production team that your article has been published in the final version. All changes at this stage are final. We will not be able to make any further changes after publication. Kind regards, BioMed Central Production Team 2
  2. 2. 1 RESEARCH ARTICLE Open Access 2 Estrogen receptor beta impacts hormone-induced 3 alternative mRNA splicing in breast cancer cells 4Q1 Dougba Noel Dago1,2† , Claudio Scafoglio3† , Antonio Rinaldi1 , Domenico Memoli1 , Giorgio Giurato1 , Giovanni Nassa1 , 5 Maria Ravo1 , Francesca Rizzo1 , Roberta Tarallo1 and Alessandro Weisz1,4* 6789 10 Abstract 11 Background: Estrogens play an important role in breast cancer (BC) development and progression; when the two 12 isoforms of the estrogen receptor (ERα and ERβ) are co-expressed each of them mediate specific effects of these 13 hormones in BC cells. ERβ has been suggested to exert an antagonist role toward the oncogenic activities of ERα, 14 and for this reason it is considered an oncosuppressor. As clinical evidence regarding a prognostic role for this 15 receptor subtype in hormone-responsive BC is still limited and conflicting, more knowledge is required on the 16 biological functions of ERβ in cancer cells. We have previously described the ERβ and ERα interactomes from 17 BC cells, identifying specific and distinct patterns of protein interactions for the two receptors. In particular, we 18 identified factors involved in mRNA splicing and maturation as important components of both ERα and ERβ pathways. 19 Guided by these findings, here we performed RNA sequencing to investigate in depth the differences in the early 20 transcriptional events and RNA splicing patterns induced by estradiol in cells expressing ERα alone or ERα and ERβ. 21 Results: Exon skipping was the most abundant splicing event in the post-transcriptional regulation by estradiol. We 22 identified several splicing events induced by ERα alone and by ERα + ERβ, demonstrating for the first time that ERβ 23 significantly affects estrogen-induced splicing in BC cells, as revealed by modification of a subset of ERα-dependent 24 splicing by ERβ, as well as by the presence of splicing isoforms only in ERβ + cells. In particular, we observed that 25 ERβ + BC cell lines exhibited around 2-fold more splicing events than the ERβ- cells. Interestingly, we identified putative 26 direct targets of ERβ-mediated alternative splicing by correlating the genomic locations of ERβ and ERα binding sites 27 with estradiol-induced differential splicing in the corresponding genes. 28 Conclusions: Taken together, these results demonstrate that ERβ significantly affects estrogen-induced early 29 transcription and mRNA splicing in hormone-responsive BC cells, providing novel information on the biological role of 30 ERβ in these tumors. 31 Keywords: Breast cancer, Estrogen receptor beta, Alternative splicing, Alternative promoters, RNAseq 32 Background 33 Breast cancer (BC) is the most frequent cancer in women 34 worldwide [1], and its development and progression are 35 known to rely strongly on the stimulation by female sexual 36 hormones, especially estrogens. Estrogen receptors α (ERα) 37 and β (ERβ) are transcription factors that mediate the 38 actions of estrogens in target cells [2,3]. Ligand binding to 39ERα or ERβ induces receptor dimerization, either as homo- 40dimers (ERα/ERα or ERβ/ERβ) or heterodimers (ERα/ERβ) 41[4], and promotes its translocation to the nucleus and 42binding of target chromatin sites via Estrogen Response 43Elements (EREs) and other regulatory elements on DNA 44[5]. Although encoded by different genes, ERα and ERβ 45share the same general modular protein structure of the 46nuclear hormone receptor superfamily, and have almost 47100% amino acid sequence homology in their DNA- 48binding domain. They also show 59% amino acid hom- 49ology in their ligand-binding domains [4,5]. The two 50ERs are thus quite similar in sequence and structure, 51but ERβ has considerably different and, in most cases, 52opposite biological effects compared to ERα in BC cells, * Correspondence: † Equal contributors 1 Q2 Laboratory of Molecular Medicine and Genomics, Department of Medicine and Surgery, University of Salerno, Via S. Allende, 1, Baronissi, SA 84081, Italy 4 Molecular Pathology and Medical Genomics, “SS. Giovanni di Dio e Ruggi d’Aragona - Schola Medica Salernitana” Hospital of the University of Salerno, Salerno, Italy Full list of author information is available at the end of the article © 2015 Dago et al.; licensee BioMed Central. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated. Dago et al. BMC Genomics _#####################_ DOI 10.1186/s12864-015-1541-1
  3. 3. 53 both in vivo and in vitro. ERα regulates the transcription 54 of hundreds of genes [6], enhancing BC cell growth, 55 proliferation and survival in response to estrogens [7]. 56 The specific role of ERβ and its impact in BC are un- 57 clear. This ER subtype is expressed in 70% of human 58 breast tumors in combination with ERα, even if some 59 human breast tumors express only ERβ [8-10]. Several 60 reports have suggested that ERβ has anti-proliferative 61 action in BC cells, by increasing the expression of anti- 62 proliferative genes and/or decreasing the expression of 63 proliferative and anti-apoptotic genes [11-15] and the 64 ERα/ERβ ratio determines the cell-specific response to 65 estrogen. In BC this ratio is higher than in normal tis- 66 sues, due to up-regulation of ERα and down-regulation 67 of ERβ [16]. Loss of ERβ mRNA levels in cancer can 68 occur as a result of promoter methylation [17]. These 69 observations have suggested a positive prognostic value 70 of this receptor subtype [12,18]. However, several studies 71 have reported also a negative prognostic value for ERβ 72 expression [19,20], making the overall contribution of this 73 receptor isoform to BC biology unclear. 74 The exact mechanism of the antagonism between ERβ 75 and ERα is only partially known. As the two receptors 76 share only 30% homology in their transactivation do- 77 main AF1 [5], it is likely that they show different pat- 78 terns of interaction with coregulatory proteins. Indeed, 79 we have previously reported that the protein interac- 80 tomes of both ERβ and ERα show significant differences 81 of the protein complexes engaged by the two ER sub- 82 types [21-25]. A particularly interesting subset of inter- 83 acting proteins, with only partially overlapping interaction 84 patterns between the two receptors [26], comprises factors 85 involved in RNA maturation and splicing [Additional 86 file 1: Figure S1] [26]. 87 Alternative splicing is a mechanism by which cells can 88 increase the variability of their proteomes by changing 89 the composition of transcribed genes through differential 90 choice of exons to be included in the final mRNA mol- 91 ecule [27]. Almost 90% of human genes show alternative 92 splicing during development, cell differentiation and 93 disease [28]. Recent studies have shown the existence 94 of cancer-specific splicing events by which transformed 95 cells switch from the adult isoform of the gene to a 96 more embryonic one, contributing to the cancer pheno- 97 type [29-31]. 98 Alternative splicing events have been monitored in BC 99 and in numerous tumor types [32], and ERα itself has 100 been reported to induce alternative splicing of a spe- 101 cific set of genes [33-35]. Here, we investigated the 102 ability of ERβ to regulate mRNA maturation and spli- 103 cing in hormone-responsive BC cells. To this purpose, 104 we performed high-throughput RNA sequencing (RNA- 105 seq) analysis of human MCF-7 cell lines stably trans- 106 fected with ERβ, and compared them with the wild type 107line, expressing only ERα, upon 17β-estradiol (E2) 108stimulation. 109Results 110High-throughput sequencing in ERβ + and ERβ- human BC 111cell lines 112We have previously established and characterized sub- 113clones of the human BC cell line MCF-7 expressing hu- 114man ERβ fused to a Tandem Affinity Purification (TAP) 115tag, and have shown that the addition of the TAP-tag at 116either the N- or the C-terminus of the protein (indicated 117as Nt-ERβ and Ct-ERβ, respectively) does not alter sig- 118nificantly the receptor function, nor its ability to activate 119transcription or to antagonize ERα-dependent transcrip- 120tion [21,22,36]. In order to study the early events of 121hormone-induced pre-mRNA maturation, we used these 122stable cell lines to perform deep-sequencing analysis of 123estrogen-induced transcriptional events shortly after 124stimulation with 17β-estradiol (2 h), to focus mostly on 125primary transcriptional events [21,36]. For comparison, 126we also performed the experiment in wild-type MCF-7 127cells, which do not express endogenous ERβ. 128Almost 70 million reads/replicate were aligned against 129the reference human genome for ERβ- and ERβ + BC 130cell lines. The number of reads for genes and isoforms 131were normalized to “Fragment Per Kilobases of exon per 132Million of mapped reads” (FPKM). In order to analyze 133genes and isoforms, we set 0.5 FPKM (at least one ana- 134lyzed condition, with/without E2 stimulus) as the expres- 135sion level threshold. In this way, we identified 16,821 136(MCF-7 wt), 16,148 (Ct-ERβ) and 17,135 (Nt-ERβ) genes 137as expressed. The criteria for considering genes and iso- 138forms as significantly regulated by estradiol were: FPKM 139value ≥0.5 in at least one analyzed condition, q-value 140(FDR-adjusted p-value of the test statistics) ≤0.05 and 141|fold-change| (FC) ≥1.3 [Additional file 2: Table S1]. 142As shown in Figure Q3 F11A, 895 (MCF-7 wt), 2,899 (Ct-ERβ) 143and 3,043 (Nt-ERβ) genes were detected as significantly 144regulated by E2 in these BC cell lines. Expression of 145ERβ in MCF7 cells significantly affected the estrogen- 146dependent gene expression profile: the regulation of 147around 230 genes (≈25% of E2-regulated genes) was 148lost in both cell lines expressing ERβ, while a large 149number of genes which were not regulated in wt cells 150became significantly regulated in Ct-ERβ (2,396) and 151Nt-ERβ (2,463) clones (Figure 1). The genes regulated 152consistently in both ERβ + lines are reported in Additional 153file 2: Tables S1-D. Interestingly, expression of ERβ in this 154BC cell line had a stronger effect on inhibited genes than 155on the activated ones: regulation of 40% of genes inhibited 156by estradiol in wt cells was lost in both ERβ + cell lines, 157versus 14% for estrogen-activated genes. Gene Ontology 158analysis revealed that among the most enriched functions 159in the group of genes whose regulation by estradiol was Dago et al. BMC Genomics _#####################_ Page 2 of 13
  4. 4. 160 lost in both ERβ + cells there were, as expected, DNA 161 Replication, Recombination and Repair, as well as Cell 162 Cycle and Cell Morphology (data not shown). 163 Estrogen-dependent splicing events in ERβ + and ERβ- 164 human BC cell lines 165 Events of exon skipping, mutually exclusive exons, alterna- 166 tive start, stop splice site and intron retention were anno- 167 tated using the Multivariate Analysis of Transcript Splicing 168 (MATS) software [37] [Additional file 3: Tables S2]. Exon 169 skipping appeared to be the predominant splice event in all 170 cell lines analyzed. MATS reveled in detail 1,264 (Ct-ERβ), 171 1,402 (Nt-ERβ) and 975 (MCF-7 wt) exon skipping events 172 induced by estradiol, associated with 1,016 (Ct-ERβ), 1,117 173 (Nt-ERβ) and 816 (wt MCF-7) genes (FigureF2 2A). Five 174 hundred seventy-five events were common to all the cell 175 lines analyzed while 115 showed opposite exon inclusion 176 level in ERβ + lines compared to ERβ- wt cells. We also 177 observed high levels of retained intron and mutually ex- 178 clusive exons events, confirming a complex and significant 179 effect of ERs on the regulation of RNA splicing in these 180 cell lines (Figure 2A). 181 To focus on the differences in splicing patterns between 182 ERβ + and ERβ- cell lines, we first looked at the genes 183which were regulated by estradiol in opposite direction in 184ERβ + cells versus wt cells [Additional file 4: Table S3]. 185Among 298 regulated genes, 56 also underwent estradiol- 186induced alternative splicing in at least one of the cell lines, 187confirming that pre-mRNA maturation was regulated 188concurrently with transcription in a significant fraction 189of ERβ-regulated genes (Figure 2B). These ERβ-regulated 190genes undergoing alternative splicing included transcrip- 191tional regulators (NCOR2, ZNF189, MLXIP, ANKRD12, 192HSF1), enzymes involved in nucleoside/nucleotide metab- 193olism (GUK1, NME3, NME4), actin remodeling and cellu- 194lar transport processes (TNS3, TRAPPC6A, TMSB15B, 195KIF12), and protein translation (ZNF98, EEF1D, RPL10, 196RPL18, RPS18). 197Estrogen-induced differential splicing has been repor- 198ted also in genes independently on transcriptional regula- 199tion [35]. In order to find the splicing events differentially 200regulated by estradiol in ERβ + compared to wt cells, we 201scanned the whole list of expressed genes for splicing pat- 202terns occurring differentially in ERβ + vs ERβ- cells. In 203addition, we focused on those splicing events whose oc- 204currence significantly altered the ratio between different 205isoforms of the same gene. Therefore, for each isoform we 206calculated the percentage of gene expression associated 207with that particular isoform (FPKM ratio: FPKMisoform/ 208FPKMgene %) and selected for further analysis only those 209isoforms in which estradiol induced a change in FPKM 210ratio of at least 10% either in wt MCF-7 or in both ERβ + 211cell lines. Other criteria of inclusion were: occurrence 212of at least one splicing event as detected by MATS; 213at least one isoform of the gene with FPKM value ≥0.5 in 214at least one analyzed condition, q-value ≤0.05 either in wt 215or in both ERβ + cell lines; regulation in both ERβ + lines 216in opposite directions compared to the wt. In this 217way, we identified the quantitatively most relevant 218splicing events differentially regulated by E2 in ERβ + 219versus ERβ- cells, including 35 genes whose isoform 220composition changed significantly after E2 stimulation 221in an ERβ-dependent fashion (Figure F33). Among these, we 222found genes involved in apoptosis (BAD), lipid metabol- 223ism (ACADM, PLSCR1, SLC27A2, STARD4), nutrient 224transport (SLC25A19, SLC35C2), transmembrane receptor 225signaling (IFNGR2, LDLRAD4), Notch signaling (PSEN2, 226POGLUT1, SGK1, SLC35C2), as well as some non-coding 227RNAs (MCM3AP-AS1, SNHG17). An example of a gene 228whose splicing pattern was affected by estradiol in an ERβ- 229dependent fashion is reported in Figure F44A, showing the 230gene SGK1, which encodes a serum and glucocorticoid- 231induced serine/threonine protein kinase involved in ion 232transport affecting many cellular processes such as cell 233growth, proliferation, survival, apoptosis and migration 234[38,39]. The gene was induced by estradiol in both wt and 235ERβ + cells; however, expression of ERβ in the absence of 236hormonal stimulation induced a promoter usage switch Figure 1 Venn diagram of expressed genes in ERβ + and wt breast cancer cell lines. The Venn diagram shows the number of genes regulated by estradiol in the three cell lines, as indicated: wt (parental MCF-7 cells); Ct-ERβ (MCF-7 subclone stably expressing ERβ tagged with the TAP-tag at the C-terminus); Nt-ERβ (MCF-7 subclone stably expressing ERβ tagged with the TAP-tag at the N-terminus). The pie charts in the lower panels specify the direction of regulation by estradiol (induction or repression) of the genes whose regulation is present in wt cells but is lost in both ERβ + cell lines (left panel), or of the genes whose regulation by estradiol is not present in the wt cells but appears in both ERβ + lines (right panel). Dago et al. BMC Genomics _#####################_ Page 3 of 13
  5. 5. Figure 2 (See legend on next page.) Dago et al. BMC Genomics _#####################_ Page 4 of 13
  6. 6. 237 and retention of the first intron causing an alterna- 238 tive translation start site and therefore the expression 239 of an isoform with a different N-terminal sequence 240 (ENST00000367857), compared to the major isoform 241 expressed (ENST00000237305). It has been reported 242 that alternative isoforms at the N-terminus can affect 243 SGK1 localization and protein stability: the ERβ-specific 244 form with intron-retention misses a very crucial sequence 245 involved in targeting to the endoplasmic reticulum as well 246 as in proteasomal degradation [40]. 247 This data suggests that expression of ERβ causes switches 248 in estradiol-induced splicing patterns, potentially affecting 249 expression or function or ER targets. 250 Estrogen-dependent alternative promoter usage in ERβ + vs 251 ERβ- BC cells 252 As the usage of alternative promoters is a major deter- 253 minant of protein diversity, even more than alternative 254 splicing [41,42], we next focused on utilization of mul- 255 tiple promoters. We grouped the primary transcripts of 256 a gene based on the promoter used, and subsequently 257 tested changes in primary transcript abundance by 258 measuring the square root of the Jensen-Shannon diver- 259 gence that occurred within and between the analyzed 260 groups. Finally, we investigated the potential promoter 261 switch regulation for the genes comprising more than 262 one differentially expressed transcript initiating from 263 distinct genomic loci. There were 977 (Ct-ERβ), 402 264 (Nt-ERβ) and 222 (wt MCF-7) distinct promoter-switching 265 genes (FDR ≤ 0.05) in ERβ + and ERβ- BC cell lines, re- 266 spectively [Additional file 5: Tables S4A-C]. Of the 222 267 promoter-switching genes recorded in wt MCF-7 cells, 268 165 did not show promoter switch in both ERβ + cell lines, 269 while 61 new promoter-switching events, not present in 270 wt cells, were detected in both ERβ + lines. These 61 ERβ- 271 specific promoter-switching genes are involved in import- 272 ant cellular functions known to be controlled by E2 in BC 273 cells, such as transcription (FOXJ3, GTF2H, NR2C2AP), 274 DNA metabolism and repair (PRKDC, REV3L, SCAND3), 275 pre-mRNA maturation and splicing (PPM1G, PRPF38B, 276 RNMT, RPRD1A, SMG1), translation (FARSB, RARS, 277 RPS21, UTP20), protein ubiquitination and proteasome 278 pathway (CUL5, KLHL2, PSMB1, USP7), cytoskeleton and 279 cytokinesis (DCTN4, MYL12A, SEPT9, SYDE2), mem- 280 brane metabolism, remodeling and intracellular transport 281 (ATP9A, CAST, MAL2, PSD3, TMEM43, TRAPPC9), cell 282adhesion and polarity (ARHGAP12, CD9, CLDN7, 283EPB41L5, PERP), signal transduction (MAP3K5, NGFRAP1, 284TNFRSF12A, WWC3, ZDHHC5). 285As an example, we focused on the PSD3 gene, predicted 286to be a nucleotide exchange factor for ADP Ribosylation 287Factor (ARF) 6, a member of the RAS family involved in 288vesicular trafficking, remodeling of membrane lipids, and 289signal transduction [43]. As show in Figure 4B, for this 290gene we found 9 distinct primary transcripts (TSS01- 291TSS09) whose usage ratios changed with E2 stimulus 292(right panel) compared to the control untreated cells 293(left panel). In ERβ + cells, estradiol induced a switch 294from promoter TSS08 (ENST00000523619) to the down- 295stream promoter TSS02, resulting in a shorter transcript 296(ENST00000519653), which is predicted to undergo 297nonsense-mediated decay [Additional file 6: Table S5]. 298Estrogen-dependent splice ratios in ERβ + vs ERβ- cells 299To obtain a comprehensive view of the estrogen-induced 300differences in the splice ratios between wt, Ct-ERβ and 301Nt-ERβ cells, we employed Cuffdiff v2.1.1 [44], which 302calculates the changes in splice isoforms abundance, by 303quantifying the square root of the Jensen-Shannon diver- 304gence, considering each primary transcripts able to pro- 305duce multiple isoforms. We determined 217 (Ct-ERβ), 306241 (Nt-ERβ) and 95 (MCF-7 wt) differentially spliced 307genes (DSGs, i.e. genes for which estradiol challenge 308induced at least one splicing event) with a FDR 309value <0.05 [Additional file 5: Tables S4D-F]. Of the 31095 spliced genes detected in wt cells, sixty-nine lacked 311splicing events in both ERβ + cell lines, suggesting inhib- 312ition of ERα-dependent splicing by ERβ. Moreover, 28 313ERβ-specific DSGs, in which no estrogen-induced splicing 314was recorded in wt cells, were found in both ERβ + cell 315lines. These include genes involved in mitosis and cyto- 316kinesis (SEPT9, BICD2, ENSA, PDS5A), cell cycle control 317(CCNJ, RAN), transcription (C11orf30, EIF3M, HIPK1, 318PBX1, ZNF124, ZNF131), protein folding (DNAJB6, 319HSP90B1), ubiquitination and sumoylation (DCUN1D4, 320SENP5, TRIM33), and signal transduction (APBB2, RTKN2). 321Correlation between ER binding and ER-dependent splicing 322In order to identify direct splicing targets of ERα and ERβ, 323we next investigated the presence of ERα and ERβ binding 324sites in genomic locations close to the identified DSGs 325[Additional file 7: Table S6]. To verify the specificity of ER (See figure on previous page.) Figure 2 Annotation of splice events in ERα + and ERα + ERβ + BC cell lines. (A) The bar plot shows the number of all alternative splicing events occurring in the cell lines analyzed. Inclusion and exclusion behavior for each event are shown (FDR ≤ 0.05; c ≤ |0.1|). (B) Genes whose regulation has opposite direction in the ERβ + lines compared to the wt MCF-7. The heat map on the right side shows the gene expression fold changes induced by estradiol. The matrix on the left side shows in black those genes for which a splicing event was detected in at least one of the cell lines. The nomenclature for the cell lines is the following: wt (parental MCF-7 cells); Ct-ERβ (MCF-7 subclone stably expressing ERβ tagged with the TAP-tag at the C-terminus); Nt-ERβ (MCF-7 subclone stably expressing ERβ tagged with the TAP-tag at the N-terminus). Dago et al. BMC Genomics _#####################_ Page 5 of 13
  7. 7. 326binding sites, we created a heat-map based on previously 327published ChIP-Seq data [21,45]. The binding densities of 328ERα and ERβ were clustered according to the seqMINER 329platform [46]. In the clustering shown in [Additional file 3308: Figure S2], each line represents a genomic location of a 331binding site with its surrounding ±1.5 kb region. In the left 332panel, ERα binding sites were used as a reference to col- 333lect a ChIP-seq tag densities window in ERβ + and ERβ- 334cell lines, while in the right panel ERβ binding sites were 335used as a reference. The heat map, representing the clus- 336tered density matrix, confirmed that ERβ presence modi- 337fied a significant number of ERα binding sites. In order to 338investigate the correlation between ERα and ERβ DNA 339binding and splicing events, we compared our RNA-Seq 340data with the ChIP-Seq data we had previously obtained 341in these same cell lines [21]. We considered the binding 342sites included within regions spanning 10 kb upstream or 343downstream of all DSGs in each cell line. The Circos plot 344[47] in Figure F55 shows the DSGs for which estradiol chal- 345lenge induced at least one splicing event in both ERβ + cell 346lines with a nearby binding site for ERα and/or ERβ. The 347outer ring (blue) reports the ERα binding sites, while the 348inner ring (red) shows the ERβ binding sites. Based on 349ERα and ERβ binding, we distinguished three different 350DSGs groups. Group 1 DSGs were associated with both 351ERα and ERβ binding sites (black). Group 2 includes 352DSGs associated with ERα binding sites exclusively (blue) 353and group 3 DSGs associated with ERβ binding sites ex- 354clusively (red). The vast majority of genes exhibited both 355binding sites (Group 1), confirming the competing role of 356ERα and ERβ. Interestingly, among the putative direct 357ERβ splicing targets we found many genes involved in 358transcription: transcription factors (FOXN3, NFIB, TAF6, 359TCF12, ZNF295, ZNF438), histone methyltransferases Figure 3 Selected isoform switches affected by the expression of ERβ. The ERβ-dependent differential splicing events that affect most prominently the balance of different isoforms for each gene were identified by using the following parameters: (i) isoform ratio (FPKMisoform/FPKMgene %) changing of at least 10% after estradiol stimulation, either in the wt cells or in both the ERβ + cells in at least one of the isoforms of the gene; (ii) at least one isoform of the gene significantly regulated by estradiol (p-value ≤ 0.05; |FC| ≥ 1.3) either in the wt cells or in both the ERβ + cells; (iii) isoform ratio changing in opposite direction in the wt cells compared to the ERβ + cells; (iv) at least one splicing event identified by MATS analysis. Thirty-five genes satisfied all the selection requirements; of these, only 2 to 3 isoforms were selected for presentation, according to the expression levels and the regulation by estradiol. Left panel: heat map of regulation of the selected genes by estradiol in: Ct-ERβ (MCF-7 subclone stably expressing ERβ tagged with the TAP-tag at the C-terminus); Nt-ERβ (MCF-7 subclone stably expressing ERβ tagged with the TAP-tag at the N-terminus); wt (parental MCF-7 cells). Right panel: heat map with the change in isoform ratio (FPKMisoform/FPKMgene %) between E2-treated and non-treated cells in the indicated cell lines; for each gene, at least two different isoforms are presented, to show estrogen-induced switch from one isoform to the other. Dago et al. BMC Genomics _#####################_ Page 6 of 13
  8. 8. Figure 4 Examples of ERβ-specific splicing events. A) Example of alternative splicing in the SGK1 gene. The upper panel shows a schematic representation of two differentially regulated isoforms of the gene (same as those shown in Figure 3 for the same gene), differing in the transcription start site, in the inclusion of the first intron (the gene is encoded by the reverse strand) and in the transcription stop site. The lower panels show a representation of the RNA-Seq reads and junction reads associated with the gene in the different conditions: Ct-ERβ (MCF-7 subclone stably expressing ERβ tagged with the TAP-tag at the C-terminus) without or with E2 stimulation, and wt (parental MCF-7 line) without or with E2 stimulation. B) Example of ERβ-specific alternative promoter usage in the gene PSD3. The left panel shows a heat map with the fold change of all the primary transcripts associated with the gene in: Ct-ERβ (MCF-7 subclone stably expressing ERβ tagged with the TAP-tag at the C-terminus); Nt-ERβ (MCF-7 subclone stably expressing ERβ tagged with the TAP-tag at the N-terminus); wt (parental MCF-7 cells). The right panels show the logarithm base 10 (log10) FPKM for each of the different transcript start sites (from TSS01 to TSS09) corresponding with the different gene isoforms in the above described cell lines, without (left) or with (right) E2 stimulation. Dago et al. BMC Genomics _#####################_ Page 7 of 13
  9. 9. 360 (ASH1L, MLL5, SETD5, SETMAR) and acetyltransferase 361 (YEATS2), other transcriptional regulators (AKNA, BANP, 362 CHD3, MEIS2). Other interesting functions associated with 363 these genes are: apoptosis (BCL2L13, CASP8, C1orf201), 364 autophagy (AMBRA1, ATG12, ATG13, ATG16L2), splicing 365 (HNRNPH3, MBL2), protein ubiquitination (CNOT4, 366 FBXW11, RFWD2, WDR20) cytoskeleton and cytokin- 367 esis (AURKA, EPS8L2, EPB41L2, LARP4, NPHP4, PLEC, 368TLN2), primary cilium biogenesis [48] (BBIP1, IFT140, 369KIAA0586), intracellular trafficking/membrane traf- 370ficking (ASAP1, KIF13A, MCOLN1, RAB17, TBC1D1, 371VPS29), cell adhesion (ARMC8, CD151, ELMOD3, LLP, 372VMP1). 373Taken together, these analyses suggest the strong effect 374of ERs binding on alternative splicing and confirming 375the key role of ERβ in BC cells. Figure 5 Correlation between ERβ-specific splicing and ER binding. Circos plot of differential spliced genes (DSGs) common to both ERβ + cell lines which contain at least one ER binding event within a window of 10 kB around the gene. The outer ring shows chromosome ideograms with the relative genes located in their respective chromosomal locations, with the following color code: the genes that have both ERβ and ERα binding are in black; the genes which have only ERβ binding are in red; the genes which have only ERα binding are in blue. The two internal rings represent ERα and ERβ binding events in blue and red, respectively. Dago et al. BMC Genomics _#####################_ Page 8 of 13
  10. 10. 376 Discussion 377 In this study we investigated for the first time the effects 378 of ERβ expression on estrogen-dependent pre-mRNA 379 maturation and splicing. We used two different subclones 380 of ERα-positive and estrogen-dependent MCF-7 cells sta- 381 bly expressing ERβ, and we found that expression of ERβ 382 in these cells significantly affected the estradiol-dependent 383 early transcriptional program and splicing pattern. In par- 384 ticular, introduction of ERβ caused loss of regulation of 385 25% of the estrogen-regulated genes. Moreover, a com- 386 parison between the ERβ + and the wt ERβ- cells showed 387 that expression of ERβ caused loss of ERα-induced pro- 388 moter switching in 75% of the genes, and of ERα-induced 389 splicing in 72% of the genes. 390 Besides affecting ERα-dependent transcription and 391 splicing, expression of ERβ caused the appearance of 392 ERβ-specific estrogen-responsive transcription (550 genes) 393 [Additional file 2: Tables S1-D], promoter-switching (61 394 genes) and differential splicing (28 genes) events, counting 395 only the events that were present in both ERβ + lines. The 396 biological functions of the ERβ-specific genes included 397 DNA replication and repair, cell cycle, apoptosis and au- 398 tophagy, DNA transcription, lipid metabolism, membrane 399 metabolism, intracellular trafficking, mRNA maturation 400 and translation, protein ubiquitination and sumoylation, 401 cell signaling, confirming that a wide variety of cellular 402 processes are affected by ERβ. 403 Based on our data, there are at least three different 404 mechanisms by which ERβ can affect ERα-dependent 405 transcription: (1) competition with ERα for binding to 406 target gene promoters, in the form of competitive bind- 407 ing or heterodimerization, which can alter the recruit- 408 ment of coregulators: the comparison between Chip-seq 409 and RNA-seq data showed that the majority of the pri- 410 mary targets identified in this experiment were directly 411 targeted by both ERα and ERβ (Figure S1), as expected 412 also from data in the literature [49]; (2) gain of new 413 binding sites that are not bound by ERα alone: a subset 414 of binding sites appeared in the ERβ + cells but were not 415 present in wt cells, again consistently with a previous re- 416 port comparing ERα and ERβ genomic binding patterns 417 in MCF-7 cells [49]; (3) secondary effects: expression of 418 ERβ induced transcription and splicing of transcriptional 419 regulators (most interestingly, the corepressor NCOR2, 420 involved in gene repression by tamoxifen-bound estro- 421 gen receptor and by unliganded NRs such as the retinoic 422 acid receptors) and of splicing factors, which can in turn 423 affect estrogen-induced transcription and pre-mRNA 424 maturation. For instance, ERβ induced promoter switch- 425 ing in the PPM1G gene, encoding a protein phosphatase 426 responsible for dephosphorylation of pre-mRNA splicing 427 factors [50]; the ERβ-specific alternate protein isoform 428 from this gene (ENST00000350803) has an additional 17 429 amino acids in the N-terminal catalytic domain compared 430to the main isoform (ENST00000544412), which are likely 431to modulate the enzymatic function. 432Even if the sole detection of splicing events does not 433give information on the biological consequences of ERβ 434effects on RNA splicing, it is tempting to speculate on 435the possible implications of ERβ-dependent splicing 436events on BC cells. Changes in splicing patterns can 437affect biological processes by many different mechanisms, 438including gain-of-function or functional switches, altered 439cellular localization, dominant negative effect, changes in 440protein/mRNA stability. For instance, expression of ERβ 441induced a promoter switch in the USP7 gene, resulting in 442a shorter transcript (ENST00000535863) compared to the 443major isoform expressed in wt cells (ENST00000381886). 444This ERβ-specific isoform is missing the N-terminal 84 445amino acids, a region of the protein of critical importance 446for interaction with substrates. This suggests the possibil- 447ity of a switch in substrate affinity induced by ERβ. This is 448particularly interesting as USP7 is a deubiquitinating en- 449zyme, responsible for removing ubiquitin chains from 450both the tumor suppressor p53 and its negative regulator 451Mdm2 [51], which instead is an ubiquitin ligase inducing 452degradation of p53. As USP7 binds both p53 and Mdm2 453with the same N-terminal domain [51], the overall effect 454of its enzymatic activity is highly dependent on the relative 455affinity for the two targets. The ERβ-induced switch may 456alter this equilibrium, thus modulating such a relevant 457aspect of cancer biology as p53 stability. In the case of 458IFNγ Receptor 2, ERα induced expression of the full 459length transcript (ENST00000381995), while ERβ favored 460a switch toward truncated forms (ENST00000545369, 461ENST00000405436) lacking the transmembrane domain, 462and therefore predicted to be secreted as soluble forms in 463the extracellular environment, with the potential of acting 464as dominant negative modulators of interferon signaling. 465In another example (the gene PSEN1, involved in intra- 466membrane proteolysis and cleavage of the intracellular 467domain of transmembrane proteins such as amyloid 468precursor protein and Notch and therefore a potential 469therapeutic target in BC [52]), the ERβ-induced splicing 470switch did not alter the open reading frame but caused 471the expression of a longer mRNA (ENST00000366783) 472compared to the isoform that was favored by ERα 473(ENST00000422240), possibly affecting the rate of transla- 474tion or stability of the mRNA. 475Another way of finding hints to the biological conse- 476quences of ERβ-specific splicing events described here is 477the reconstruction of pathways whose genes are differen- 478tially spliced following ERβ introduction in the cells. For 479instance, we found many genes involved in the Notch sig- 480naling pathway differentially spliced by ERβ: the above- 481mentioned PSEN2 is the catalytic subunit of the γ-secretase 482complex, responsible for intramembrane proteolysis of 483transmembrane receptors including Notch, resulting in Dago et al. BMC Genomics _#####################_ Page 9 of 13
  11. 11. 484 the release of Notch Intracellular Domain (NICD), which 485 migrates to the nucleus and regulates transcription of tar- 486 get genes [53]. Also, NCOR2 is bound to the unliganded 487 CBF-1 transcription factor, a primary effector of Notch 488 signaling which acts as a repressor in unstimulated cells, 489 but is converted to an activator (dismissing the corepres- 490 sor NCOR2) after binding the NICD [54]. Furthermore, 491 POGLUT1 has recently been shown to be an endoplasmic 492 reticulum O-glycosyl-transferase responsible for glycosy- 493 lation of Notch and required for its function [55], and 494 SLC35C2 is an endoplasmic reticulum transporter respon- 495 sible for accumulation of GDP-fucose, which is used for 496 Notch fucosylation, required for full activation [56]. Fi- 497 nally, SGK1 has recently been shown to be a negative 498 regulator of Notch signaling by inhibiting γ-secretase ac- 499 tivity and promoting Notch degradation [57], and MAGI1 500 has been shown to recruit Notch ligand Dll1 to cadherin- 501 based adherens junctions, stabilizing it on the cell surface 502 [58]. It is worth noting that the predominant Notch recep- 503 tor expressed in these cell lines was Notch2, which was 504 induced by E2 in wt cells (to a level slightly below the 505 chosen cut-off threshold, but highly statistically signifi- 506 cant: FPKM without E2: 49.826; FPKM with E2: 61.583, 507 FC 1.236, q-value 0.0005), while its basal expression 508 was lowered and its up-regulation smoothened in the 509 ERβ + cells. 510 A possible modulation of the Notch pathway by ERβ is 511 especially interesting as Notch is a known regulator of 512 breast development and maintenance of breast stem 513 cells [59]; alterations in the Notch pathway have been in- 514 volved in breast carcinogenesis, and in particular the 515 Notch pathway has been implicated in the development 516 of triple negative BC (TNBC), a particularly aggressive 517 form of BC which does not express ERα, progesterone 518 receptor (PR) or HER2, and which has shown resistance 519 to all known therapies [60]. Targeting Notch signaling 520 has been proposed in TNBC. As these cancers are ERα- 521 negative, hormonal treatment is not currently used for 522 these patients; however, ERβ could be expressed in up to 523 50% of TNBCs [10,18], and its expression in TNBC has 524 been associated with better prognosis [18]. Therefore, 525 ERβ may represent a potential new therapeutic target in 526 TNBC. The interrelation between ERβ and Notch in the 527 development and prognosis of triple-negative BC should 528 be investigated further in future research. 529 Conclusions 530 In conclusion, whole-genome analysis of early transcrip- 531 tion evens and mRNA processing associated with ERβ 532 confirmed a relevant role for this receptor in modulating 533 ERα-dependent transcription and splicing, but also 534 identified novel, ERβ-specific transcription and splicing 535 events, confirming a wide range of actions of ERβ in the 536 biology of BC. 537The data reported here confirm the complexity of 538estrogen action in BC cells and provide a comprehensive 539description of the effects of ERα and by ERβ on early tran- 540scription and splicing in hormone-responsive BC cells. 541More importantly, they provide a starting point to identify 542the events of ERβ-dependent splicing which are most 543significant for cancer biology. 544Methods 545Human hormone-responsive BC cells 546Stable cell clones expressing either C-TAP-ERβ or N-TAP- 547ERβ (ERβ+) generated as previously described [23], and 548wild type (wt) MCF7 (ERβ-) cells were used for this 549study. All cell lines were maintained, propagated, hormone- 550starved for 5 days and analyzed for estrogen signaling as 551described earlier [21,45]. 552Illumina Genome Sequencing RNA sequencing library 553preparation 554Total RNA was extracted from hormone-starved cell 555cultures (+Ethanol -E2) or after 2 hours of stimulation 556with 10-8 M 17β-estradiol (+E2), as described previously 557[21]. RNA concentration in each sample was deter- 558mined with a NanoDrop-1000 spectrophotometer and 559the quality assessed with the Agilent 2100 Bioanalyzer 560and Agilent RNA 6000nano cartridges (Agilent Technolo- 561gies). Indexed libraries were prepared from 1 μg/ea. of puri- 562fied RNA with TruSeq Stranded total RNA Sample Prep 563Kit (Illumina) according to the manufacturer’s instructions. 564Libraries were sequenced (paired-end, 2×100 cycles) at a 565concentration of 8 pmol/L per lane on HiSeq1500 platform 566(Illumina) with a coverage of >70 million sequence reads/ 567sample on average. 568Read alignment and transcript assembly 569TopHat v.2.0.10 [61] was used to align all reads includ- 570ing junction-spanning reads back to the human genome 571(Homo sapiens Ensembl GRCh37, hg19). To identify the 572differentially expressed and spliced genes and isoforms 573between ERβ- and ERβ + cell lines we used Cuffdiff 574v2.1.1 [44]. The parameters to define genes and isoforms 575as differentially expressed were the following: expression 576level threshold of 0.5 FPKM; q-value (FDR-adjusted p- 577value of the test statistic) ≤ 0.05 and |FC| ≥ 1.3. Moreover, 578to detect the ERβ- and ERβ + BC-specific splice events 579such as exon skip, exon inclusion, alternative splice sites 580and intron retention, we performed a direct comparison 581analysis using MATS v3.0.8 [37]. To filter events with at 582least 10% change in exon inclusion level we set the MATS 583cutoff c, representing the extent of splicing change one 584wishes to identify, to 0.1, and FDR ≤ 0.05 to filter the iden- 585tified splice events. Dago et al. BMC Genomics _#####################_ Page 10 of 13
  12. 12. 586 Data access 587 RNA-Seq data have been deposited in the Gene Expression 588 Omnibus genomics data public repository (Q4 http://www. 589 with Accession Number GSE64590. 590Q5 Additional files 591 593 Additional file 1: Figure S1. Description: KEGG pathway (Kanehisa et al. 594 Nucleic Acids Res 2014, 42:D199) analysis of ERα and ERβ interactors 595 involved in Spliceosome Pathway. Blue and red boxes highlight ERα and 596 ERβ interacting proteins, respectively. 597 Additional file 2: Tables S1. A: Ct-ERβ gene list. Genes with a FPKM 598 value ≥0.5, q-value ≤0.05 and |FC| ≥ 1.3 in the Ct-ERβ BC cell line. B: 599 Nt-ERβ gene list. Genes with a FPKM value ≥0.5, q-value ≤0.05 and 600 |FC| ≥1.3 in the Nt-ERβ BC cell line. C: MCF-7 wt gene list. Genes with a 601 FPKM value ≥0.5, q-value ≤0.05 and |FC| ≥1.3 in the wt MCF-7 BC cell 602 line. D: Commonly regulated genes in both ERβ + BC cell lines. Genes 603 consistently regulated in both ERβ + BC cells with a q-value ≤0.05 and 604 |FC| ≥1.3. E: Ct-ERβ isoform list. Isoforms with a FPKM value ≥0.5, q-value 605 ≤0.05 and |FC| ≥1.3 in the Ct-ERβ BC cell line. F: Nt-ERβ isoform list. 606 Isoforms with a FPKM value ≥0.5, q-value ≤0.05 and |FC| ≥1.3 in the 607 Nt-ERβ BC cell line. G: MCF-7 wt isoform list. Isoforms with a FPKM 608 value ≥0.5, FDR ≤0.05 and |FC| ≥1.3 in the wt MCF-7 BC cell line. 609Q6 Additional file 3: Tables S2. A: Skipping Exon Ct-ERβ. List of skipping 610 exon events in the Ct-ERβ BC cell line (FDR ≤0.05 and |c| ≥0.1). B: Mutually 611 Exclusive Exons Ct-ERβ. List of mutually exclusive exons events in the 612 Ct-ERβ BC cell line (FDR ≤0.05 and |c| ≥0.1). C: Retained Intron Ct-ERβ. List of 613 retained intron events in the Ct-ERβ BC cell line (FDR ≤0.05 and |c| ≥0.1). 614 D: Alternative 3’ splice site Ct-ERβ. List of Alternative 3’ splice site events in 615 the Ct-ERβ BC cell line (FDR ≤0.05 and |c| ≥0.1). E: Alternative 5’ splice site 616 Ct-ERβ. List of Alternative 5’ splice site events in the Ct-ERβ BC cell line 617 (FDR ≤0.05 and |c| ≥0.1). F: Skipping Exon Nt-ERβ. List of skipping exon 618 events in the Nt-ERβ BC cell line (FDR ≤0.05 and |c| ≥0.1). G: Mutually Exclusive 619 Exons Nt-ERβ. List of mutually exclusive exons events in the Nt-ERβ BC cell line 620 (FDR ≤0.05 and |c| ≥0.1). H: Retained Intron Nt-ERβ. List of retained intron 621 events in the Nt-ERβ BC cell line (FDR ≤0.05 and |c| ≥0.1). I: Alternative 3’ splice 622 site Nt-ERβ. List of Alternative 3’ splice site events in the Nt-ERβ BC cell line 623 (FDR ≤0.05 and |c| ≥0.1). J: Alternative 5’ splice site Nt-ERβ. List of Alternative 624 5’ splice site events in the Nt-ERβ BC cell line (FDR ≤0.05 and |c| ≥0.1). K: 625 Skipping Exon MCF-7 wt. List of skipping exon events in the wt MCF-7 BC cell 626 line (FDR ≤0.05 and |c| ≥0.1). L: Mutually Exclusive Exons MCF-7 wt. List of 627 mutually exclusive exons events in the wt MCF-7 BC cell line (FDR ≤0.05 628 and |c| ≥0.1). M: Retained Intron MCF-7 wt. List of retained intron events in the 629 wt MCF-7 BC cell line (FDR ≤0.05 and |c| ≥0.1). N: Alternative 3’ splice site 630 MCF-7 wt. List of Alternative 3’ splice site events in the wt MCF-7 BC cell line 631 (FDR ≤0.05 and |c| ≥0.1). O: Alternative 5’ splice site MCF-7 wt. List of Alternative 632 5’ splice site events in the wt MCF-7 BC cell line (FDR ≤0.05 and |c| ≥0.1). 633 Additional file 4: Table S3. Discordant Isoforms ERβ+/ERβ-. List of all 634 isoforms with opposite regulation pattern in ERβ+/ERβ-. 635 Additional file 5: Tables S4. A: Differential Promoter Usage Ct-ERβ. 636 Gene list with differential promoter usage (DPU) in the Ct-ERβ BC cell line 637 (FDR ≤0.05). B: Differential Promoter Usage Nt-ERβ. Gene list with DPU in 638 the Nt-ERβ BC cell line (FDR ≤0.05). C: Differential Promoter Usage MCF-7 639 wt. Gene list with DPU in the wt MCF-7 BC cell line (FDR ≤0.05). D: Differential 640 Spliced Gene Ct-ERβ. Gene list differentially spliced (DSGs) in the Ct-ERβ BC 641 cell line (FDR ≤0.05). E: Differential Spliced Gene Nt-ERβ. Gene list differentially 642 spliced (DSGs) in the Nt-ERβ BC cell line (FDR ≤0.05). F: Differential Spliced 643 Gene MCF-7 wt. Gene list differentially spliced (DSGs) in the wt MCF-7 BC cell 644 line (FDR ≤0.05). 645 Additional file 6: Table S5. PSD3 TSSs expression values. Transcript 646 starting sites (TSSs) expression value list of the PSD3 gene, expressed in 647 FPKM with relative FC, FDR and p-value in Ct-ERβ, Nt-ERβ and MCF-7 wt 648 BC cell lines. 649 Additional file 7: Tables S6. A: ERα binding sites in Ct-ERβ. Gene list of 650 DSGs including ERα binding sites in a 10 kb window around the gene in 651 the Ct-ERβ BC cell line. B: ERβ binding sites in Ct-ERβ. Gene list of DSGs 652including ERβ binding sites in a 10 kb window around the gene in the 653Ct-ERβ BC cell line. C: ERα binding sites in Nt-ERβ. Gene list of DSGs including 654ERα binding sites in a 10 kb window around the gene in the Nt-ERβ BC cell 655line. D: ERβ binding sites in Nt-ERβ. Gene list of DSGs including ERβ binding 656sites in a 10 kb window around the gene in the Nt-ERβ BC cell line. E: ERα 657binding sites in MCF-7 wt. Gene list of DSGs including ERα binding sites in a 65810 kb window around the gene in the wt MCF-7 BC cell line. 659Additional file 8: Figure S2. Heat map representing the clustered density 660matrix of ERα and ERβ binding sites. Chromatin immunoprecipitation of ERα 661and ERβ in Ct-ERβ and wt MCF-7 cells show different ER binding profiles 662(highlighted by square brackets). In the clustering, each line represents a 663genomic location of a binding site with its surrounding ±1.5 kb region. This 664matrix was subjected to k-means clustering. 665Abbreviations 666ER: Estrogen receptor; Nt-ERβ: N-TAP-estrogen receptor beta; Ct-ERβ: 667C-TAP-estrogen receptor beta; E2: 17β-estradiol; BC: Breast cancer; 668FPKM: Fragment per kilobases of exon per million; DSGs: Differentially spliced 669genes; TSSs: Transcripts starting sites; DPU: Differential promoter usage. 670Competing interests 671The authors declare that they have no competing interests. 672Authors’ contributions 673All authors participated in conception and design of the study and 674manuscript drafting and revision, GN, MR, FR and RT preformed in vitro 675experimental work and sequencing, DND, AR, DM and GG performed the 676statistical and bioinformatic analyses, CS preformed functional investigations 677on alternatively splices genes, CS and AW coordinated and finalized figure 678preparation, manuscript drafting and revision. All authors read and approved 679the final manuscript. 680Acknowledgements 681Work supported by: Italian Association for Cancer Research (Grants IG-13176), 682Italian Ministries of Education, University and Research (Grants 2010LC747T, 683RBFR12W5V5_003 and PON03PE_00146_1) and Health (Young Researcher 684Grants GR-2011-02350476 and GR-2011-02347781), National Research Council 685(Flagship Project Interomics), the University of Salerno (FARB 2014). G.N. is 686supported by a ‘Mario e Valeria Rindi’ fellowship of the Italian Foundation for 687Cancer Research; A.R. is a PhD student of the Research Doctorate in ‘Molecular 688and Translational Oncology and Innovative Medical-Surgical Technologies’, 689University of Catanzaro ‘Magna Graecia’; C.S. is supported by the UCLA Scholars 690in Oncologic Molecular Imaging Fellowship (National Institutes of Health R25 691CA098010). 692Author details 6931 Q2Laboratory of Molecular Medicine and Genomics, Department of Medicine 694and Surgery, University of Salerno, Via S. Allende, 1, Baronissi, SA 84081, Italy. 6952 UFR Sciences Biologiques, Université Peleforo Gon Coulibaly, Korhogo, Ivory 696Coast. 3 Department of Molecular and Medical Pharmacology, University of 697California, Los Angeles, USA. 4 Molecular Pathology and Medical Genomics, 698“SS. Giovanni di Dio e Ruggi d’Aragona - Schola Medica Salernitana” Hospital 699of the University of Salerno, Salerno, Italy. 700Received: 30 December 2014 Accepted: 17 April 2015 701 702References 1. 703IARC GLOBOCAN 2012 Fact Sheet. [ Q8 704sheets_population.aspx] 2. 705Nilsson S, Makela S, Treuter E, Tujague M, Thomsen J, Andersson G, et al. 706Mechanisms of estrogen action. Physiol Rev. 2001;81(4):1535–65. 3. 707Deroo BJ, Korach KS. Estrogen receptors and human disease. J Clin Invest. 7082006;116(3):561–70. 4. 709Cowley SM, Hoare S, Mosselman S, Parker MG. Estrogen receptors alpha and 710beta form heterodimers on DNA. J Biol Chem. 1997;272(32):19858–62. 5. 711Pace P, Taylor J, Suntharalingam S, Coombes RC, Ali S. Human estrogen 712receptor beta binds DNA in a manner similar to and dimerizes with 713estrogen receptor alpha. J Biol Chem. 1997;272(41):25832–8. Dago et al. BMC Genomics _#####################_ Page 11 of 13
  13. 13. 6.714 Hah N, Danko CG, Core L, Waterfall JJ, Siepel A, Lis JT, et al. A rapid, 715 extensive, and transient transcriptional response to estrogen signaling in 716 breast cancer cells. Cell. 2011;145(4):622–34. 7.717 Frasor J, Danes JM, Komm B, Chang KC, Lyttle CR, Katzenellenbogen BS. 718 Profiling of estrogen up- and down-regulated gene expression in human 719 breast cancer cells: insights into gene networks and pathways underlying 720 estrogenic control of proliferation and cell phenotype. Endocrinology. 721 2003;144(10):4562–74. 8.722 Kurebayashi J, Otsuki T, Kunisue H, Tanaka K, Yamamoto S, Sonoo H. 723 Expression levels of estrogen receptor-alpha, estrogen receptor-beta, 724 coactivators, and corepressors in breast cancer. Clin Cancer Res. 725 2000;6(2):512–8. 9.726 Speirs V, Carder PJ, Lane S, Dodwell D, Lansdown MR, Hanby AM. 727 Oestrogen receptor beta: what it means for patients with breast cancer. 728 Lancet. 2004;5(3):174–81. 10.729 Skliris GP, Leygue E, Curtis-Snell L, Watson PH, Murphy LC. Expression of 730 oestrogen receptor-beta in oestrogen receptor-alpha negative human 731 breast tumours. Br J Cancer. 2006;95(5):616–26. 11.732 Lazennec G, Bresson D, Lucas A, Chauveau C, Vignon F. ER beta inhibits 733 proliferation and invasion of breast cancer cells. Endocrinology. 734 2001;142(9):4120–30. 12.735 Paruthiyil S, Parmar H, Kerekatte V, Cunha GR, Firestone GL, Leitman DC. 736 Estrogen receptor beta inhibits human breast cancer cell proliferation and 737 tumor formation by causing a G2 cell cycle arrest. Cancer Res. 2004;64(1):423–8. 13.738 Strom A, Hartman J, Foster JS, Kietz S, Wimalasena J, Gustafsson JA. Estrogen 739 receptor beta inhibits 17beta-estradiol-stimulated proliferation of the breast 740 cancer cell line T47D. Proc Natl Acad Sci U S A. 2004;101(6):1566–71. 14.741 Chang EC, Frasor J, Komm B, Katzenellenbogen BS. Impact of estrogen 742 receptor beta on gene networks regulated by estrogen receptor alpha in 743 breast cancer cells. Endocrinology. 2006;147(10):4831–42. 15.744 Williams C, Edvardsson K, Lewandowski SA, Strom A, Gustafsson JA. A 745 genome-wide study of the repressive effects of estrogen receptor beta on 746 estrogen receptor alpha signaling in breast cancer cells. Oncogene. 747 2008;27(7):1019–32. 16.748 Roger P, Sahla ME, Makela S, Gustafsson JA, Baldet P, Rochefort H. 749 Decreased expression of estrogen receptor beta protein in proliferative 750 preinvasive mammary tumors. Cancer Res. 2001;61(6):2537–41. 17.751 Zhao C, Lam EW, Sunters A, Enmark E, De Bella MT, Coombes RC, et al. 752 Expression of estrogen receptor beta isoforms in normal breast epithelial 753 cells and breast cancer: regulation by methylation. Oncogene. 754 2003;22(48):7600–6. 18.755 Honma N, Horii R, Iwase T, Saji S, Younes M, Takubo K, et al. Clinical 756 importance of estrogen receptor-beta evaluation in breast cancer patients 757 treated with adjuvant tamoxifen therapy. J Clin Oncol. 2008;26(22):3727–34. 19.758 Guo L, Zhu Q, Yilamu D, Jakulin A, Liu S, Liang T. Expression and prognostic 759 value of estrogen receptor beta in breast cancer patients. Int Journal Clin 760 Exp Med. 2014;7(10):3730–6. 20.761 Chantzi NI, Tiniakos DG, Palaiologou M, Goutas N, Filippidis T, Vassilaros SD, 762 et al. Estrogen receptor beta 2 is associated with poor prognosis in 763 estrogen receptor alpha-negative breast carcinoma. J Cancer Res Clin Oncol. 764 2013;139(9):1489–98. 21.765 Grober OM, Mutarelli M, Giurato G, Ravo M, Cicatiello L, De Filippo MR, et al. 766 Global analysis of estrogen receptor beta binding to breast cancer cell 767 genome reveals an extensive interplay with estrogen receptor alpha for 768 target gene regulation. BMC Genomics. 2011;12:36. 22.769 Paris O, Ferraro L, Grober OM, Ravo M, De Filippo MR, Giurato G, et al. 770 Direct regulation of microRNA biogenesis and expression by estrogen 771 receptor beta in hormone-responsive breast cancer. Oncogene. 772 2012;31(38):4196–206. 23.773 Nassa G, Tarallo R, Ambrosino C, Bamundo A, Ferraro L, Paris O, et al. A 774 large set of estrogen receptor beta-interacting proteins identified by 775 tandem affinity purification in hormone-responsive human breast cancer 776 cell nuclei. Proteomics. 2011;11(1):159–65. 24.777 Tarallo R, Bamundo A, Nassa G, Nola E, Paris O, Ambrosino C, et al. 778 Identification of proteins associated with ligand-activated estrogen receptor 779 alpha in human breast cancer cell nuclei by tandem affinity purification and 780 nano LC-MS/MS. Proteomics. 2011;11(1):172–9. 25.781 Ambrosino C, Tarallo R, Bamundo A, Cuomo D, Franci G, Nassa G, et al. 782 Identification of a hormone-regulated dynamic nuclear actin network 783 associated with estrogen receptor alpha in human breast cancer cell nuclei. 784 Mol Cell Proteomics. 2010;9(6):1352–67. 26. 785Nassa G, Tarallo R, Guzzi PH, Ferraro L, Cirillo F, Ravo M, et al. Comparative 786analysis of nuclear estrogen receptor alpha and beta interactomes in breast 787cancer cells. Mol Biosyst. 2011;7(3):667–76. 27. 788Black DL. Mechanisms of alternative pre-messenger RNA splicing. Annu Rev 789Biochem. 2003;72:291–336. 28. 790Wang ET, Sandberg R, Luo S, Khrebtukova I, Zhang L, Mayr C, et al. 791Alternative isoform regulation in human tissue transcriptomes. Nature. 7922008;456(7221):470–6. 29. 793Oltean S, Bates DO. Hallmarks of alternative splicing in cancer. Oncogene. 7942014;33(46):5311–8. Q730. 795Chen J, Weiss WA. Alternative splicing in cancer: implications for biology 796and therapy. Oncogene. 2014; 31. 797Carninci P, Kasukawa T, Katayama S, Gough J, Frith MC, Maeda N, et al. The 798transcriptional landscape of the mammalian genome. Science (New York, NY). 7992005;309(5740):1559–63. 32. 800Venables JP, Klinck R, Koh C, Gervais-Bird J, Bramard A, Inkel L, et al. 801Cancer-associated regulation of alternative splicing. Nat Struct Mol Biol. 8022009;16(6):670–6. 33. 803Salama SA, Mohammad MA, Diaz-Arrastia CR, Kamel MW, Kilic GS, Ndofor 804BT, et al. Estradiol-17beta upregulates pyruvate kinase M2 expression to 805coactivate estrogen receptor-alpha and to integrate metabolic reprogramming 806with the mitogenic response in endometrial cells. J Clin Endocrinol Metab. 8072014;99(10):3790–9. 34. 808Lal S, Allan A, Markovic D, Walker R, Macartney J, Europe-Finner N, et al. 809Estrogen alters the splicing of type 1 corticotropin-releasing hormone 810receptor in breast cancer cells. Sci Signal. 2013;6(282):ra53. 35. 811Bhat-Nakshatri P, Song EK, Collins NR, Uversky VN, Dunker AK, O’Malley BW, 812et al. Interplay between estrogen receptor and AKT in estradiol-induced 813alternative splicing. BMC Med Genomics. 2013;6:21. 36. 814Nassa G, Tarallo R, Giurato G, De Filippo MR, Ravo M, Rizzo F, et al. 815Post-transcriptional regulation of human breast cancer cell proteome by 816unliganded estrogen receptor beta via microRNAs. Mol Cell Proteomics. 8172014;13(4):1076–90. 37. 818Shen S, Park JW, Huang J, Dittmar KA, Lu ZX, Zhou Q, et al. MATS: a 819Bayesian framework for flexible detection of differential alternative splicing 820from RNA-Seq data. Nucleic Acids Res. 2012;40(8):e61. 38. 821Lang F, Shumilina E. Regulation of ion channels by the serum- and 822glucocorticoid-inducible kinase SGK1. Faseb J. 2012;27(1):3–12. 39. 823Lang F, Stournaras C. Serum and glucocorticoid inducible kinase, 824metabolic syndrome, inflammation, and tumor growth. Hormones (Athens). 8252013;12(2):160–71. 40. 826Simon P, Schneck M, Hochstetter T, Koutsouki E, Mittelbronn M, Merseburger 827A, et al. Differential regulation of serum- and glucocorticoid-inducible kinase 1 828(SGK1) splice variants based on alternative initiation of transcription. 829Cell Physiol Biochem. 2007;20(6):715–28. 41. 830Pal S, Gupta R, Kim H, Wickramasinghe P, Baubet V, Showe LC, et al. 831Alternative transcription exceeds alternative splicing in generating the 832transcriptome diversity of cerebellar development. Genome Res. 8332011;21(8):1260–72. 42. 834Shabalina SA, Ogurtsov AY, Spiridonov NA, Koonin EV. Evolution at protein 835ends: major contribution of alternative transcription initiation and 836termination to the transcriptome and proteome diversity in mammals. 837Nucleic Acids Res. 2014;42(11):7132–44. 43. 838Schweitzer JK, Sedgwick AE, D’Souza-Schorey C. ARF6-mediated endocytic 839recycling impacts cell movement, cell division and lipid homeostasis. 840Semin Cell Dev Biol. 2011;22(1):39–47. 44. 841Trapnell C, Hendrickson DG, Sauvageau M, Goff L, Rinn JL, Pachter L. 842Differential analysis of gene regulation at transcript resolution with RNA-seq. 843Nat Biotechnol. 2013;31(1):46–53. 45. 844Cicatiello L, Mutarelli M, Grober OM, Paris O, Ferraro L, Ravo M, et al. 845Estrogen receptor alpha controls a gene network in luminal-like breast 846cancer cells comprising multiple transcription factors and microRNAs. 847Am J Pathol. 2010;176(5):2113–30. 46. 848Ye T, Krebs AR, Choukrallah MA, Keime C, Plewniak F, Davidson I, et al. 849seqMINER: an integrated ChIP-seq data interpretation platform. 850Nucleic Acids Res. 2011;39(6):e35. 47. 851Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, et al. 852Circos: an information aesthetic for comparative genomics. Genome Res. 8532009;19(9):1639–45. 48. 854Menzl I, Lebeau L, Pandey R, Hassounah NB, Li FW, Nagle R, et al. Loss of 855primary cilia occurs early in breast cancer development. Cilia. 2014;3:7. Dago et al. BMC Genomics _#####################_ Page 12 of 13
  14. 14. 49.856 Madak-Erdogan Z, Charn TH, Jiang Y, Liu ET, Katzenellenbogen JA, 857 Katzenellenbogen BS. Integrative genomics of gene and metabolic 858 regulation by estrogen receptors alpha and beta, and their coregulators. 859 Mol Syst Biol. 2013;9:676. 50.860 Murray MV, Kobayashi R, Krainer AR. The type 2C Ser/Thr phosphatase 861 PP2Cgamma is a pre-mRNA splicing factor. Genes Dev. 1999;13(1):87–97. 51.862 Hu M, Gu L, Li M, Jeffrey PD, Gu W, Shi Y. Structural basis of competitive 863 recognition of p53 and MDM2 by HAUSP/USP7: implications for the 864 regulation of the p53-MDM2 pathway. PLoS Biol. 2006;4(2):e27. 52.865 Han J, Shen Q. Targeting gamma-secretase in breast cancer. Breast cancer 866 (Dove Medical Press). 2012;4:83–90. 53.867 Kopan R, Ilagan MX. The canonical Notch signaling pathway: unfolding the 868 activation mechanism. Cell. 2009;137(2):216–33. 54.869 Kao HY, Ordentlich P, Koyano-Nakagawa N, Tang Z, Downes M, Kintner CR, 870 et al. A histone deacetylase corepressor complex regulates the Notch signal 871 transduction pathway. Genes Dev. 1998;12(15):2269–77. 55.872 Acar M, Jafar-Nejad H, Takeuchi H, Rajan A, Ibrani D, Rana NA, et al. Rumi is 873 a CAP10 domain glycosyltransferase that modifies Notch and is required for 874 Notch signaling. Cell. 2008;132(2):247–58. 56.875 Lu L, Hou X, Shi S, Korner C, Stanley P. Slc35c2 promotes Notch1 876 fucosylation and is required for optimal Notch signaling in mammalian cells. 877 J Biol Chem. 2010;285(46):36245–54. 57.878 Mo JS, Ann EJ, Yoon JH, Jung J, Choi YH, Kim HY, et al. Serum- and 879 glucocorticoid-inducible kinase 1 (SGK1) controls Notch1 signaling by 880 downregulation of protein stability through Fbw7 ubiquitin ligase. J Cell Sci. 881 2010;124(Pt 1):100–12. 58.882 Mizuhara E, Nakatani T, Minaki Y, Sakamoto Y, Ono Y, Takai Y. MAGI1 recruits 883 Dll1 to cadherin-based adherens junctions and stabilizes it on the cell 884 surface. J Biol Chem. 2005;280(28):26499–507. 59.885 Reedijk M. Notch signaling and breast cancer. Adv Exp Med Biol. 886 2012;727:241–57. 60.887 Nwabo Kamdje AH, Seke Etet PF, Vecchio L, Muller JM, Krampera M, 888 Lukong KE. Signaling pathways in breast cancer: therapeutic targeting of 889 the microenvironment. Cell Signal. 2014;26(12):2843–56. 61.890 Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: 891 accurate alignment of transcriptomes in the presence of insertions, 892 deletions and gene fusions. Genome Biol. 2013;14(4):R36. 893 Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at 895895 896 Dago et al. BMC Genomics _#####################_ Page 13 of 13
  15. 15. Author Query Form Journal: BMC Genomics Title: Estrogen receptor beta impacts hormone-induced alternative mRNA splicing in breast cancer cells Authors: Dougba Noel Dago, Claudio Scafoglio, Antonio Rinaldi, Domenico Memoli, Giorgio - Giurato, Giovanni Nassa, Maria Ravo, Francesca Rizzo, Roberta Tarallo, Alessandro Weisz Article: 1541 Dear Authors, During production of your paper, the following queries arose. Please respond to these by annotating your proofs with the necessary changes/additions. If you intend to annotate your proof electronically, please refer to the E-annotation guidelines. We recommend that you provide additional clarification of answers to queries by entering your answers on the query sheet, in addition to the text mark-up. Query No. Query Remark Q1 Author names: Please confirm that the author names are presented accurately and in the correct sequence (given names/initials, family name). Author 1: Given name: Dougba Given name: Noel Family name: Dago Author 2: Given name: Claudio Family name: Scafoglio Author 3: Given name: Antonio Family name: Rinaldi Author 4: Given name: Domenico Family name: Memoli Author 5: Given name: Giorgio Family name: Giurato Author 6: Given name: Giovanni Family name: Nassa
  16. 16. Query No. Query Remark Author 7: Given name: Maria Family name: Ravo Author 8: Given name: Francesca Family name: Rizzo Author 9: Given name: Roberta Family name: Tarallo Author 10: Given name: Alessandro Family name: Weisz Q2 Please check the affiliations if captured correctly. Q3 Figure: Upon checking, it was noticed that Figure 1 has a panel A inside the main body of text; however, it was not found in the corresponding image and caption. Please provide us with an updated figure with corresponding panel matching their description in the figure caption and image. Otherwise, please advise us on how to proceed. Q4 URL: Please check that the following URLs are working. If not, please provide alternatives: Q5 Please check the caption of the additional files if captured correctly. Q6 Additional file: Journal standard requires Additional file legends of less than 300 words only; however, the provided data of Additional file 3 in the manuscript exceed the requirement. Please shorten the additional file legend accordingly. We suggest that any detailed information should be included as an introduction within the Additional file itself. Q7 Citation details for Reference [30] is incomplete. Please supply the volume id and page range of this reference. Otherwise, kindly advise us on how to proceed. Q8 URL: Please check that the following URLs are working. If not, please provide alternatives: