TRA2A Promoted Paclitaxel Resistance and Tumor Progression in Triple Negative Breast Cancers Via Regulating Alternative Splicing

Author Manuscript Published OnlineFirst on April 17, 2017; DOI: 10.1158/1535-7163.MCT-17-0026 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Title Page Title: TRA2A promoted paclitaxel resistance and tumor progression in triple negative breast cancers via regulating alternative splicing

Tieju Liu1,2, Huizhi Sun1, Dongwang Zhu3, Xueyi Dong1,2, Fang Liu1,2, Xiaohui Liang1,2, Chen Chen1, Bing Shao1, Meili Wang1, Yi Wang1, Baocun Sun1,2*

1 Department of Pathology, Tianjin Medical University, Tianjin 300070, China 2 Department of Pathology, General Hospital of Tianjin Medical University, Tianjin 300052, China 3 Stomatology Hospital of Tianjin Medical University, Tianjin, China

*Corresponding author: Prof. Baocun Sun, Department of Pathology and General Hospital of Tianjin Medical University, Tianjin, China; E-mail: [email protected]; [email protected] Tel:86-13602111192 Fax:86-22-83336813

Running title: TRA2A and paclitaxel resistance

Keywords: triple-negative breast cancer; alternative splicing; TRA2A; paclitaxel resistance

Downloaded from mct.aacrjournals.org on September 30, 2021. © 2017 American Association for Cancer Research. Author Manuscript Published OnlineFirst on April 17, 2017; DOI: 10.1158/1535-7163.MCT-17-0026 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Abbreviations AS, alternative splicing alt-exons, AS exons CE, cassette exon con-exons, constitutive exons DSS, disease-specific survival GO, Gene Ontology hnRNPs, heterogenous ribonucleoproteins IC50, inhibitory concentration 50 MTT, methylthiazol tetrazolium nt, nucleotide OS, overall survival pre-mRNAs, precursor messenger RNAs PTX, paclitaxel PTXR, PTX resistant ROC, receiver-operating characteristic SI, splicing index STRING, Search Tool for the Retrieval of Interacting Genes/Proteins TAC, Transcriptome Analysis Console TNBC, triple-negative breast cancer TRA2, Transformer2

Grant Support: This work was partly supported by a grant from The National Natural Science Foundation of China (No. 81672870 to T. Liu and No. 81572872 to X. Zhao), Key project of the National Natural Science Foundation of China (No. 81230050 to B. Sun), and National Undergraduate Training Program for Innovation and Entrepreneurship (No. 201510062001 to H. Sun).

Abstract Treatment of triple-negative breast cancer (TNBC) has been challenging and paclitaxel (PTX) resistance is one of the major obstacles to the better prognosis. Deregulation of alternative splicing (AS) may contribute to tumor progression and chemotherapy resistance. Human AS factor TRA2 has two separate gene paralogs encoding TRA2A and TRA2B proteins. TRA2B is associated with cancer cell survival and therapeutic sensitivity. However the individual role of TRA2A in cancer progression has not been reported. Here we report that TRA2A facilitates proliferation and survival, migration and invasion of TNBC cells. In addition, TRA2A promotes PTX resistance of TNBC by specifically controlling cancer-related splicing, which is independent of other splicing factors. TRA2A overexpression could promote AS of CALU, RSRC2 and PALM during PTX treatment of TNBC cells. The isoform shift of RSRC2 from RSRC2s to RSRC2l leads to a decreased RSRC2 protein expression, which could contribute to TNBC PTX resistance. TRA2A can regulate RSRC2 AS by specifically binding upstream intronic sequence of exon4. Strikingly, TRA2A expression is increased dramatically in TNBC patients, and has close relationship with decreased RSRC2 expression; both of them are associated with poor survival of TNBC. Collectively, our findings suggest that PTX targets the TRA2A-RSRC2 splicing pathway and deregulated TRA2A and RSRC2 expression may confer PTX resistance. In addition to providing a novel molecular mechanism of cancer-related splicing dysregulation, our study demonstrates that expression of TRA2A in conjunction with RSRC2 may provide valuable molecular biomarker evidence for TNBC clinical treatment decisions and patient outcome.

Introduction Triple-negative breast cancer (TNBC) patients are usually managed with chemotherapy including paclitaxel (PTX). PTX polymerizes tubulin and promotes microtubule assembly and stabilization to disrupt normal microtubule dynamics and arrest cells in mitosis. The ideal tumour responses are that cancer cells arrest in mitosis and die following PTX chemotherapy. However cancer cells can maintain viability by undergoing viable cellular responses and enhance the malignant phenotype after chemotherapy (1). Although, initially responsive to PTX, TNBC often recur and metastasize due to the development of chemoresistance (2). Alternative splicing (AS) is the process by which splice sites in precursor messenger RNAs (pre-mRNAs) are differentially selected and paired to produce multiple mature mRNAs and protein isoforms with distinct structural and functional properties. Regulation of AS is tightly controlled during normal tissue differentiation. Deregulation of AS can lead to production of aberrant protein isoforms, which may contribute to tumor establishment, progression and resistance to therapeutic treatments (3-7). In general, the regulation of AS patterns is achieved through complex interplay between cis regulatory elements within the pre-mRNAs and the trans protein factors that bind them (8, 9). The trans protein factors have been found to function in tumorigenesis and drug resistance (10). Transformer2 (TRA2) proteins that are first discovered in insects, could form an essential component of the splicing complex that controls fly sexual differentiation (11, 12), and play a role in the regulation of pre-mRNA splicing. Human TRA2 gene has two separate gene paralogs encoding TRA2A and TRA2B proteins. Both TRA2A and TRA2B contain RNA recognition motifs and extended regions of serine and arginine residues, resembling the well characterized trans protein factors (12-14). Current data implicate TRA2 proteins solely in AS rather than constitutive splicing (12, 15). There are studies that demonstrate TRA2B expression levels are upregulated in breast, cervical, ovarian, and lung cancer, and TRA2B is associated with cancer cell survival and drug sensitivity (16, 17). The known splicing targets of TRA2B identified in normal tissues are important for cancer cell biology and are particularly implicated in cell division, motility and invasion (11, 18). Recent study (12) found that simultaneous depletion of TRA2A and TRA2B induced substantial shifts in splicing

of endogenous TRA2B target exons, and that both constitutive and alternative target exons were under dual TRA2A-TRA2B control. Following depletion of TRA2B, upregulated TRA2A was able to functionally substitute for TRA2B and largely maintained TRA2B target exon inclusion, suggesting TRA2A has the same function as TRA2B and could modulate splicing events. However, the individual role of TRA2A in cancer related splicing has not been reported so far. Here we analyzed TRA2A-mediated changes of the transcriptome and assessed the role of TRA2A in PTX resistance and cancer progression of TNBC.

Materials and Methods Cell culture and lentiviral transduction MDA-MB-231 and 293T cells were obtained from the American Type Culture Collection in 2012 and authenticated using short tandem repeat (STR) analysis by Genewiz Inc. in 2014. STR analysis showed that the submitted samples were in good agreement with the reference cell lines. Hs578T cells were provided by the Cell Bank of Type Culture Collection of the Chinese Academy of Sciences, Shanghai, China in 2016. The pEZ-Lv201 vector was used for overexpressing TRA2A, hnRNPm and RSRC2, and the psi-LVRU6GP was used for hnRNPm or RSRC2 silencing (GeneCopoeia). The shRNA target sequence was (5′-ggtccgagcagacattcttga-3′) for hnRNPM and (5′-ggaagagagcgactaaattca-3′) for RSRC2. Lentiviruses were produced by transient transfection of 293T cells, and the virus suspension was used to infect the target cells. Paclitaxel (PTX) treatment PTX (Selleckchem) was prepared as a 10 μM stock solution in DMSO. Briefly, 1×106 cells were plated in 100-mm culture dishes for 24 hs and then treated with 10 nM PTX. After 3 days, fresh 10 nM PTX-containing media was added for another 2 days, totaling 5 days of PTX treatment. Cells were then rinsed with PBS and maintained in drug-free culture with media replacement every 48 hs until proliferative cell clones established. Cytotoxicity assay and inhibitory concentration 50 (IC50) measurement Cell growth determination kit (MTT based, Sigma) was used and the manufacturer’s instruction was followed. Brifely, cells were separately seeded into 96-well plates at 1 × 103 cells/well and were treated with PTX at different concentrations (a series of

dilutions, each a doubling dilution of the previous one) in 100 μL culture medium. After 48 hs culture, 10 μL MTT (methylthiazol tetrazolium) solution was added and incubated for 4 hs at 37℃. Subsequently the supernatant was removed, and 100 μL MTT solvent was added. Spectrophotometrically measure absorbance at a wavelength of 570 nm was perfomed by using a BioTek ELx800. Dose-response curves were plotted to determine half maximal IC50 for PTX using the GraphPad Prism6 (GraphPad Software, San Diego CA, USA).The assays were performed independently and repeated at least 3 times. Plate clonogenic assay Cells (5 × 104 cells per well) were cultured in 6-well plates overnight and then exposed to 10 nM PTX for 5 days. The cells were further cultured for 10 days in six-well plates containing drug-free medium. Clonogenic cells were determined as those able to form a colony consisting of at least 50 cells. The colonies were fixed with methanol and stained with 0.5% crystal violet. Wound-healing assay Cells were implanted into 6-well plates for 90% confluence, and then a sterilized tip was used to draw a line with the same width on the bottom of the dishes. Images were captured at 72 hours after the wounding. Data shown were representative of three independent repeats. Cell migration and invasion assays Transwell 24 well plates containing permeable polyethylene terephthalate membrane inserts with 8 μM pores (Invitrogen) were used. 1 × 105 cells in 100 μL DMEM without FBS were placed on the upper layer of Transwell inserts. Matrigel (1 mg/mL, BD Biosciences) was additionally coated on the inserts for invasion assay. The lower chamber was filled with 10% FBS-containing medium. The cells were incubated for 48 hours, and subsequently non-migratory or non-invading cells were removed from the upper surface of the membrane. The cells that passed through the membrane were fixed with methanol and stained with 0.5% crystal violet. The number of migratory or invading cells was counted using an inverted light microscope (Nikon). RNA extraction and microarray analysis Total RNA was extracted using Trizol reagent (Tiangen Biotech, Beijing, China), and sent to Oebiotech (Shanghai, China) for Affymetrix GeneChip® Human Transcriptome Array 2.0 analysis. The microarray data have been deposited in

NCBI’s Gene Expression Omnibus (Liu et al., 2016) and are accessible through GEO Series accession number GSE90145 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE90145). Quantitative real-time polymerase chain reaction (qRT-PCR) qRT-PCR was performed as previously described (19), using the primers listed in Table S3. Briefly, 2 μg of total RNA was reverse-transcribed into cDNA using a Reverse Transcription Kit (Takara, RR037A). qRT-PCR analyses were performed with Power SYBR Green (Takara, RR820A) in 7500HT Real-Time PCR System (Applied Biosystems, Foster City, CA). GAPDH was used as an endogenous control. The qRT-PCR data of 47 cases of breast cancer tissue were presented by using the 2–ΔCt method (20). The gene expression of interest was normalized to an internal control (GAPDH) [ΔCt = (Ct gene of interest - Ct GAPDH)], and three PCR replicates per cDNA sample were performed. 2–ΔCt values were calculated and compared between different groups by using Independent-Samples T Test. The qRT-PCR data of breast cancer cells were expressed as 2-∆∆Ct using the equation [ΔΔCT = (Ct gene of interest - Ct GAPDH) treated sample - (Ct gene of interest - Ct GAPDH) control sample] to calculate and were presented as fold change in expression. The fold change > 2 or < -2 was considered as significant. Semiquantitative RT-PCR The assay was performed according to the recommended thermal profile: 95°C for 5 minutes (preincubation), followed by 30 cycles at 95°C for 30 seconds (denaturation), 60°C for 1 minute (annealing), and 72°C for 30 seconds (elongation). The amplified products were subjected to electrophoresis in a 2% agarose gel containing ethidium bromide (Bio-Rad). The expression of GAPDH was used to examine the integrity of the RNA in each sample and to standardize the amount of cDNA added to each PCR tube. The following primers were used: RSRC2: forward: 5′-AGAAAACACAGGAGCCGGAG-3′; reverse: 5′-TGAGTGACTTCTGCCTCTTGA-3′; GAPDH: forward: 5′-CCTGGCCAAGGTCATCCATGAC-3′; reverse: 5′-TGTCATACCAGGAAATGAGCTTG-3′. RSRC2 splicing reporter minigene assays The RSRC2 minigenes’ construct was referred to RSRC2 PubMed ID NR_036434’s sequence. The RSRC2 minigene 1 contains exon3 (44 nucleotide, nt), 237 nt of

flanking intronic sequences upstream from exon4, exon4 (78 nt) and exon5 (191 nt). The RSRC2 minigene 2 contains exon3 (44 nt), exon4 (78 nt), 237 nt of flanking intronic sequences downstream from exon4 and exon5 (191 nt). RSRC2 minigene 3 was constructed by altering minigene 1’s motif AGAAG into AGACG, AGAA into ATAA, GAAG into GACG, AGAAG into AGACG, AGAA into AGTA, AGAAG into AGACG, and GAAG into GATG. These minigenes were assembled into the pEZ-Lv206 vector and were customized from GeneCopoeia, Inc. The sequences of RSRC2 minigenes had been provided in supplementary data. RSRC2 splicing reporter minigene assays were performed in 293T cells. Briefly, 2×105 cells were seeded in 6-well plate 24 hs before transfection using polyethylenimine. Different amounts of TRA2A were cotransfected with the RSRC2 minigenes construct. Cells were collected 48 hs after transfection, and RNA was isolated for subsequent RT-PCR analysis. Tissue specimens The written informed consents from the patients for all the breast caner specimens were obtained. The study was conducted in accordance with Declaration of Helsinki and was approved by the review board of Tianjin Medical University, China. Immunohistochemistry (IHC) and evaluation of IHC staining Information on these staining methods may be referenced to the literatures (21-24). Scoring system was modified and used according to published evaluation standard (25). Antibodies Antibodies used for Western blotting and IHC were as follows: TRA2A (GeneTex), hnRNPm (Novus Biologicals), RSRC2 (Novus Biologicals). Statistical analysis Data analysis was performed with the SPSS16.0 software package (IBM). All p values were two-sided, and statistical significance was measured at the 0.05 level. Supplemental material The online supplemental material included supplemental figures, tables and legends.

Results: 1 Paclitaxel (PTX) treatment induced a higher TRA2A expression in MDA-MB-231 cells

The majority of MDA-MB-231 died in 2 weeks after treated with PTX 10 nM for 5 days (Figure 1A). A small number of cells survived (Figure 1A, black arrowhead), resumed proliferation and established characteristic clones within 3-4 weeks (Figure 1B, black arrow), and such cells were considered to be PTX resistant (PTXR) cells. Next these cells were sent for microarray analysis. Gene Ontology (GO) biological process revealed that the most represented category was mitotic cell cycle, apoptosis, response to drug, angiogenesis and RNA splicing (Figure 1C), etc. There were 8 splicing factors identified by GO analysis, including DHX9, TRA2A, heterogeneous nuclear ribonucleoprotein (hnRNP)m, hnRNPu, hnRNPA2B1, SRSF6, SMC1A, RBM8A (Table S1). By Search Tool for the Retrieval of Interacting Genes/Proteins (STRING)-Known and Predicted Protein-Protein Interactions analysis (Figure 1D), there were stronger interactions among hnRNPM, hnRNPu, hnRNPA2B1, SRSF6 and DHX9 while a weaker interaction between hnRNPm and TRA2A or hnRNPu and TRA2A. To validate these findings qRT-PCR was performed for hnRNPm, hnRNPu, hnRNPA2B1, SRSF6, DHX9 and TRA2A. Remarkably TRA2A and hnRNPm significantly showed higher expression than other splicing factors (Figure 1E). Consistently, western blotting confirmed an increase in relative TRA2A and hnRNPm protein levels under PTX treatment compared to normal conditions (Figure 1F). HnRNPm is a splicing factor that potentiates TGFβ signaling, drives cells to undergo EMT, and promotes breast cancer metastasis in animals (26). Interestingly, TRA2A protein expression neither increased following hnRNPm overexpression nor decreased with hnRNPm knockdown (Figure 1F-G). In addition, when MDA-MB-231 cells with hnRNPm knock down were given PTX treatment, TRA2A protein level still showed a dramatic increase (Figure 1G). These results suggested that TRA2A might play an important role in PTXR, which was independent of hnRNPm. TRA2 has two separated gene paralogs encoding TRA2A and TRA2B proteins. It has been demonstrated that TRA2B regulates splicing patterns which are important to cancer cells splicing (15). Interestingly, TRA2B didn’t show significant increase in contrast to higher TRA2A expression after PTX treatment in our study, suggesting TRA2A, rather than TRA2B, was contributed to cancer cells survival under PTX treatment.

2 TRA2A expression was essential for the survival of PTXR tumor cells and promoted cancer cells migration and invasion MDA-MB-231 and Hs578T cells were transfected with TRA2A overexpression plasmid and they expressed higher levels of TRA2A than the empty vector transfected control cells as analyzed by western blotting (Figure 2A). The MDA-MB-231 or Hs578T cells were sensitive to the cytotoxicity of PTX with an IC50 of 33.7 nM or 20.8 nM. In contrast, the overexpressing TRA2A cells were highly resistant to PTX with an IC50 of 174.5 nM or 80.9 nM (Figure 2B). In line with the MTT data, plate clonogenic assay showed that the TRA2A overexpressing cells had stronger survival ability after PTX treatment and the numbers of cell colony formation were significantly higher in TRA2A overexpressing cells than that of control cells (Figure 2C). Remarkably TRA2A knockdown in PTXR cells impaired the clonogenic survival ability of the resistant cells (Figure 2D) when given PTX treatmen, inducing IC50 decreased from 208.4 nM to 59.6 nM (Figure 2E). Moreover, quantitative analyses of wound healing assay suggested a significant difference in the speed of wound healing between the TRA2A overexpression and the control cells. The TRA2A transfected cells displayed the faster speed of wound healing (Figure 2F). Meantime, the increased migration and invasion ability was observed in TRA2A overexpressing cells by transwell assays (Figure 2G-H).

3 Global Regulation of the Transcriptome by TRA2A in Cancer-Related Genes The microarray data of MDA-MB-231 overexpressing TRA2A and control cells were analyzed by Transcriptome Analysis Console (TAC) software to obtain a list of AS events. Figure 3A showed the gene structure view of one example (MELK gene), there was an exon skipping event in control, but not in the TRA2A overexpressing cells. We identified 3290 TRA2A-regulated AS events, including cassette exon (CE), alternative 5' Donor Site (A5D), alternative 3' Acceptor Site (A3A), intron retention (IR), mutually exclusive exons (MEE) (Figure 3B; Table S2). Subsequent analysis indicated that the AS events could be negatively or positively regulated by TRA2A (decreased or increased splicing index (SI) value by TRA2A expression) (Figure 3C), demonstrating that TRA2A was a general splicing factor that controls different types of AS when specifically binding to pre-mRNAs.

When analyzing cellular functions of TRA2A-regulated AS events using GO, we found that TRA2A affected genes in the RNA processing pathway, including translational initiation, elongation and termination, as well as RNA splicing (Figure 3D). Such functional enrichment is not surprising because TRA2A is RNA binding factor known to regulate splicing and translation. Intriguingly, TRA2A targets were also enriched with the cancer-related functions such as regulation of cell proliferation including mitotic cell cycle, cell division, and programmed cell death and apoptotic process. The TRA2A-regulated splicing targets involved in translation control (Figure 3E) and cell division (Figure 3F) could be functionally connected into well linked interaction networks. Taken together, these results suggested that the biological processes affected by TRA2A were related to proliferation, apoptosis, and tumorigenesis. We also analyzed how TRA2A affected global gene expression. The genes regulated by TRA2A were associated significantly with cancer-related functions, as judged by GO (Figure 3G). The expression of cell proliferation marker, PCNA, CDC25A and CDC6, was elevated following TRA2A overexpression (Figure 3H). Interestingly, the AS factor hnRNPm and SRSF6 showed an increased expression following TRA2A upregulation (Figure 3I).

4 The AS regulated by TRA2A played roles in PTXR of MDA-MB-231 cells MELK, CALU, RSRC2, CEACAM1, LMCD1, PALM and RFWD2 were selected for CE validation (Figure 3A, Figure S1). All the selected genes function as oncogene or tumor suppressor in cell proliferation and migration in tumor progression (27-33). Subsequently we designed qPCR primers to interrogate constitutive exons (con-exons) and AS exons (alt-exons) for each gene (Table S3) and confirmed that TRA2A either positively or negatively controlled all endogenous AS events tested. The fold changes of con-exons and alt-exons were presented in Figure 4A, and the relative changes of SIs obtained from qRT-PCR were highly correlated to those observed by microarray data and TAC analysis (Figure 4B; P = 0.019) (Table S4). We next examined the AS events of the selected 7 genes response to PTX. A striking observation was that CALU, RSRC2, PALM AS events occurred during PTX treatment (Figure 4C-K) while MELK, CEACAM1, LMCD1 and RFWD2 AS events weren’t shown (Figure S2). Upon PTX treatment, CALU or RSRC2 short isoform

(CALUs or RSRC2s) was gradually converted to long isoform (CALUl or RSRC2l) resulting from exon inclusion. The expression of CALUl or RSRC2l was increased 13-fold or 8-fold, while expression of CALUs or RSRC2s was decreased 10-fold or 9-fold respectively (Figure 4C, F). The switch in CALU or RSRC2 isoform was also evidenced by the ratio of CALUl to CALUs or RSRC2l to RSRC2s, which increased approximately 143-fold or 74-fold at day 14 respectively (Figure 4D, G). Importantly, total levels of CALU (CALUt) or RSRC2 (RSRC2t) transcription varied by less than 2-fold throughout the PTX treatment (Figure 4E, H). However, for PALM, exon8 skipping occurred in PTX treatment, thus inducing a strong shift from the long isoform (PALMl) to the short isoform (PALMs) emerged with extended phases of PTX treatment. The PALMl expression was decreased 12-fold and the PALMs expression was increased 11-fold at day 14 (Figure 4I). Correspondingly the ratio of PALMl to PALMs was decreased 130-fold (Figure 4J). However the total level of PALM (PALMt) wasn’t significantly changed (Figure 4K).

5 TRA2A expression was specifically associated with PTXR of TNBC TRA2A protein expression was detected in a cohort of 100 breast cancer patients including 37 TNBC and 63 non-TNBC patients. TRA2A positive signal was located in the nucleus and its expression was higher in TNBC than in non-TNBC (Figure 4L, M, P = 0.035). Moreover, survival analysis showed that TRA2A expression was significantly associated with poorer survival (Figure 4N; Table S5). On multivariate analysis, TRA2A positive expression remained associated with poor survival after correcting for tumour stage (P = 0.000, relative risk = 5.355 for overall survival (OS) and P = 0.000, relative risk = 5.058 for disease specific survival (DSS), respectively) (Table S5), supporting that TRA2A positive expression was a prognostic marker independent of the clinicopathological parameters examined. In this cohort, all the TNBC patients (n = 37) and 69.8% (44/63) non-TNBC patients received PTX chemotherapy. For TNBC patients, elevated TRA2A expression was significantly associated with poor survival (Figure 4O; Table S6) and was strong risk marker (P = 0.001, relative risk = 7.748 for OS and P = 0.001, relative risk = 7.061 for DSS, respectively) (Table S6). However, there was no significant difference in the OS and DSS between TRA2A positive and TRA2A negative of non-TNBC patients with PTX chemotherapy (Figure 4P). These results indicated that the TRA2A expression was

specifically associated with PTX resistance of TNBC but not with that of non-TNBC patients.

6 RSRC2 isoform shift related to TRA2A expression occurred in PTXR of TNBC Forty-seven fresh and paraffin specimen of operable TNBC patients with PTX chemotherapy after surgery were obtained. PTXR was defined as tumor recurrence and metastasis, invasive contralateral breast cancer, or death during the period from the date of PTX assignment to date of follow-up. Remarkably, the expression of both TRA2A protein and mRNA was higher in PTXR (n = 25) than in non-PTXR (n = 22) (Figure 5A-B, Table S7). Moreover, RSRC2l showed dramatically increased expression and RSRC2s showed decreased expression in PTXR than in non-PTXR (Figure 5C-D, Table S7). However, the level of CALUl, CALUs, PALMl and PALMs did not differ significantly between PTXR and non-PTXR (Figure 5E-H) (Table S7). Then TRA2A, CALUl, CALUs, RSRC2l, RSRC2s, PALMl and PALMs mRNA levels were compared between TRA2A+ and TRA2A- cases (Figure 5I, Table S8). The level of TRA2A mRNA was significantly higher in TRA2A+ than in TRA2A- as expected. In addition, we found that RSRC2l showed increased expression while RSRC2s showed decreased expression in TRA2A+ compared with TRA2A-. However no significant differences were found in CALUl, CALUs, PALMl and PALMs levels. Next we obtained receiver-operating characteristic (ROC) curves with an AUC of 0.826 (95% CI: 0.699-0.954) for TRA2A protein, 0.956 (95% CI: 0.906-1.006) for TRA2A mRNA, and 0.904 (95% CI: 0.819-0.989) for RSRC2l expression, whose increased expression showed strong capacity for PTXR prediction. The combination of TRA2A and RSRC2l was superior to individual marker in PTXR prediction (Figure 5J-K). To further demonstrate the role of RSRC2s in PTXR of TNBC cells, RSRC2s mRNA and RSRC2 protein expression was steadily silenced by using lentiviral shRNA RSRC2s (Figure 5L). As shown in Figures 5L and M, the number of colonies and cell survival of MDA-MB-231 cells were significantly increased in shRSRC2s with PTX treatment compared with the control. The downregulation of RSRC2s was also significantly associated to a worse IC50 (33.7 nM in control vs 168.9 nM in shRSRC2s) (Figure 5M).

Now that knockdown of RSRC2s confered PTXR to cancer cells, we asked whether upregulation of RSRC2s may increase PTX sensitivity. The rescue experiments for PTXR of TRA2A overexpressing cells had been performed by upregulation of RSRC2s in MDA-MB-231TRA2A cells. RSRC2s upregulation obviously inhibited cell proliferation and survival of MDA-MB-231TRA2A cells by clonogenic assay when given PTX treatment (Figure 5N). In addition, IC50 dropped from 174.5 nM in MDA-MB-231TRA2A cells to 51.9 nM in cells with RSRC2s forcedly expressed in them (Figure 5O).

7 The decreased RSRC2 protein expression could be a marker for poor survival of TNBC RSRC2 protein expression observed in the nuclei was also examined in the 47 cases of TNBC specimen by IHC. RSRC2 protein was strongly expressed in benign breast epithelium of paracancerous tissues, with a pattern of patchy staining; i.e., positive-stained normal luminal/ductal cells were surrounded by fewer unstained myoepithelial cells (Figure 6A). Moreover a moderate to strong staining of RSRC2 could be observed in most of low grade TNBC (Figure 6A) with the positive rate of 81.8% (9/11) (Figure 6B). In contrast, a decreased staining was observed in high grade TNBC (Figure 6A) and the positive rate of RSRC2 was 44.4% (16/36) (χ2 = 4.727, P = 0.030) (Figure 6B). Importantly, RSRC2 protein expression showed negatively correlation with TRA2A protein expression by Spearman analysis (P = 0.009) (Figure 6C). In addition, the mean expression of RSRC2l mRNA was lower and RSRC2s mRNA was higher in RSRC2+ cases than in RSRC2- cases (Figure 6D-E), suggesting RSRC2 protein expression was related to RSRC2s. Survival analysis showed that the decreased RSRC2 protein expression was significantly associated with poor survival (Figure 6F).

8 A shift from RSRC2s to RSRC2l mRNA regulated by TRA2A induced the decreased RSRC2 protein expression By searching PubMed, we found that RSRC2s’s sequence was in accord with the shortest transcript (NM_023012.5) of RSRC2 gene which could encode a functional protein. Furthermore, RSRC2l’s sequence was in accord with transcripts holding

exon4 which included transcript variant 4 (NR_036435) and variant 5 (NR_036434) of RSRC2 gene. The additional exon4 includes a premature stop codon, so RSRC2l is a nonsense-mediated mRNA decay candidate and does not make a functional protein. As shown in Figure 6G, MDA-MB-231 cells with PTX treatment exhibited a significant shift in RSRC2 AS toward the RSRC2 exon4 inclusion isoform. RT-PCR analysis of RSRC2 mRNA using specific primers showed that expression of RSRC2l containing exon4 was increased, while expression of RSRC2s was decreased. Meantime, we found the forced TRA2A overexpression resulted in an analogous RSRC2 AS pattern as observed with PTX treatment (Figure 6G). Next, RSRC2 protein expression was determined by western blot analyses in cancer cells with PTX treatment or exogenous TRA2A overexpression (Figure 6H). The result showed the decline of a biological active RSRC2 protein following increased RSRC2l and decreased RSRC2s mRNA expression induced by PTX treatment or TRA2A upregulation. Taken together, we hypothesize that TRA2A is a specific regulator of RSRC2 AS by promoting the expression of the RSRC2 exon4 retaining mRNA isoform, and thereby inhibits the generation of biological active RSRC2 protein. Next, we intended to establish a direct link between exon4 inclusion of RSRC2l and TRA2A. We constructed minigene splicing reporters (Figure 6I) according to the literature (34) and cotransfected 293T cells with TRA2A and the RSRC2 minigenes. Cells transfected with the minigene containing upstream intronic sequence (minigene1) showed a remarkable shift in RSRC2 AS pattern toward the RSRC2 exon4-inclusion isoform in a TRA2A dose-dependent manner, resulting in an increase in exon4 inclusion from 8% in control cells to 95% in cells transfected with the highest dosage of TRA2A (Figure 6J). In contrast, cells transfected with the minigene containing downstream intronic sequence (minigene2) displayed no difference in RSRC2 AS pattern compared with control cells (Figure 6J). qRT-PCR analysis of exon-included and -skipped products confirmed these results (Figure 6K). It is well demonstrated that upregulated TRA2A is able to maintain TRA2B target exon inclusion following depletion of TRA2B (12), sugggesting TRA2A could control the splicing targets bound by TRA2B. Furthermore, it is reported that the RAAG or AGAA tetranucleotide is specifically recognized by TRA2B (16). An overview of the 237 bp upstream intronic sequence surrounding exon4 in RSRC2 minigene1 is observed to be particularly rich in these kinds of motifs (Supplementary

data). Therefore we supposed AGAA or RAAG sequence as the potential binding motifs for TRA2A. In order to better demonstrate the interaction, we decided to replace the suspected sequences within this region by constructing mingene3, and remarkably replacing these motifs reduced TRA2A’s splicing inclusion (Figure 6L). Thus, TRA2A is necessary and sufficient to stimulate RSRC2 exon4 inclusion via its interaction with AGAA or RAAG motifs located in introns upstream from exon4.

Discussion TNBC commonly acquires resistance to the first line chemotherapy, PTX, and this is one of major obstacles to the better prognosis. AS and the related splicing factors are de-regulated in cancer. Each of the ‘hallmarks of cancer’ including chemotherapy resistance is associated with a switch in splicing, towards a more aggressive invasive cancer phenotype (35, 36). In this study, we analyzed microarray profiles and identified 8 splicing factors related to PTXR, in which the expression level of TRA2A was the highest. Conformably, we had also found that ectopic expression of TRA2A promoted stronger resistance to PTX in TNBC cells while TRA2A knockdown in the PTXR cells induced the reversion of PTX sensitivity. However, the paralogs product of TRA2A, i.e. TRA2B, didn’t show significant increase after PTX treatment in our study. In addition, STRING analysis show there is interaction between TRA2A and hnRNPm, and hnRNPm showed the second higher level following PTX treatment in our study. However we found that the elevated TRA2A level was independent of hnRNPm expression. Interestingly, hnRNPm and the other AS factor SRSF6 showed an increased expression following TRA2A upregulation. All of these results suggest that individual TRA2A expression has a role in mediating PTXR of TNBC. Functionally, we provided evidence that ectopic expression of TRA2A in TNBC cells was associated with a highly aggressive phenotype, as constitutive expression of TRA2A stimulated breast cancer cell proliferation and survival, migration and invasion. Microarray analysis demonstrated that TRA2A targets were enriched with the cancer-related functions, and proliferative markers of tumor cells such as PCNA, CDC25A and CDC6 showed elevated expression following TRA2A overexpression, suggesting TRA2A’s function in tumor progression. We further demonstrated that TRA2A predominantly functioned as a splicing factor when directly bound to pre-mRNAs and the AS events could be negatively or

positively regulated by TRA2A. By arbitrarily examining the AS events, we found that TRA2A control AS of MELK, CALU, RSRC2, CEACAM1, LMCD1, PALM and RFWD2, which might contribute to aggressive phenotype of TRA2A overexpressing cells. Therefore, it is possible that TRA2A may affect cell growth and survival through controlling AS. Importantly we found that following PTX treatment there was significantly isoform shift of CALU, RSRC2 and PALM in survived cancer cells, hence demonstrating that cells utilized AS controlled by TRA2A to orchestrate a switch in isoform expression of cancer related genes, which in turn promoted PTXR and cancer progression. Our data of breast cancer patient samples showed that TRA2A expression could be a marker for poor survival of breast cancer. Specifically TRA2A expression was associated with PTXR of TNBC, but not with that of non-TNBC patients. Further study in fresh TNBC tissues not only confirmed TRA2A’s role in PTXR, but also demonstrated isoform shift of RSRC2 occurred in PTXR development of TNBC. RSRC2 has been identified as tumor suppressor and it might be associated with chemotherapy sensitivity (37). In our study, in vitro experiment showed that RSRC2 expression obliteration resulted in PTXR in TNBC cells and the forced RSRC2 expression in TRA2A overexpressing cells recovered cancer cells’ PTX sensitivity. Moreover, RSRC2 protein expression was decreased in resected tumor tissues when compared with normal breast acini. The decreased RSRC2 expression levels have been reported to be correlated with tumor metastasis and invasion, as well as shorter post-operative survival in esophageal cancer (29). However, the mechanism of RSRC2 downregulation in tumor aggressiveness was unclear until now. Here we demonstrated that RSRC2’s function could be controlled by TRA2A through splicing. RSRC2 is located at 12q24 and has different isoforms in which only the shortest isoform makes functional protein. At the molecular level, we found that TRA2A directly bound to RSRC2 pre-mRNA at AGAA or RAAG rich sequences and stimulated inclusion of RSRC2 variable exons, which led to RSRC2 isoform switching from RSRC2s to RSRC2l and caused the decreased RSRC2 protein levels during PTXR development of TNBC. Consistently, there was significantly negative correlation between RSRC2 and TRA2A expression in TNBC tissues, further confirming the regulation of RSRC2 by TRA2A in vivo. Crucially, like

TRA2A overexpression, the decreased expression of RSRC2 was associated with poor prognosis in TNBC. In summary, this study represents an important example of how a splicing factor can control critical AS events in chemotherapy resistance and cancer progression of TNBC. These results sugguest a causal role for the TRA2A in TNBC progression and modulating RSRC2 splicing might be a potential therapeutic intervention for TNBC.

Reference 1. Zasadil LM, Andersen KA, Yeum D, Rocque GB, Wilke LG, Tevaarwerk AJ, et al. Cytotoxicity of paclitaxel in breast cancer is due to chromosome missegregation on multipolar spindles. Science translational medicine. 2014;6:229ra43. 2. Jamdade VS, Sethi N, Mundhe NA, Kumar P, Lahkar M, Sinha N. Therapeutic targets of triple-negative breast cancer: a review. British journal of pharmacology. 2015;172:4228-37. 3. Wang Y, Chen D, Qian H, Tsai YS, Shao S, Liu Q, et al. The splicing factor RBM4 controls apoptosis, proliferation, and migration to suppress tumor progression. Cancer cell. 2014;26:374-89. 4. Dehm SM. mRNA splicing variants: exploiting modularity to outwit cancer therapy. Cancer research. 2013;73:5309-14. 5. Warzecha CC, Jiang P, Amirikian K, Dittmar KA, Lu H, Shen S, et al. An ESRP-regulated splicing programme is abrogated during the epithelial-mesenchymal transition. The EMBO journal. 2010;29:3286-300. 6. Biamonti G, Bonomi S, Gallo S, Ghigna C. Making alternative splicing decisions during epithelial-to-mesenchymal transition (EMT). Cellular and molecular life sciences : CMLS. 2012;69:2515-26. 7. Zhang J, Manley JL. Misregulation of pre-mRNA alternative splicing in cancer. Cancer discovery. 2013;3:1228-37. 8. Comiskey DF, Jr., Jacob AG, Singh RK, Tapia-Santos AS, Chandler DS. Splicing factor SRSF1 negatively regulates alternative splicing of MDM2 under damage. Nucleic acids research. 2015;43:4202-18. 9. Bonomi S, Gallo S, Catillo M, Pignataro D, Biamonti G, Ghigna C. Oncogenic alternative splicing switches: role in cancer progression and prospects for therapy. International journal of cell biology. 2013;2013:962038. 10. Wojtuszkiewicz A, Assaraf YG, Maas MJ, Kaspers GJ, Jansen G, Cloos J. Pre-mRNA splicing in cancer: the relevance in oncogenesis, treatment and drug resistance. Expert opinion on drug metabolism & toxicology. 2015;11:673-89. 11. Elliott DJ, Best A, Dalgliesh C, Ehrmann I, Grellscheid S. How does Tra2beta protein regulate tissue-specific RNA splicing? Biochemical Society transactions. 2012;40:784-8. 12. Best A, James K, Dalgliesh C, Hong E, Kheirolahi-Kouhestani M, Curk T, et al. Human Tra2 proteins jointly control a CHEK1 splicing switch among alternative and constitutive target exons. Nature communications. 2014;5:4760. 13. Zhou Z, Fu XD. Regulation of splicing by SR proteins and SR protein-specific kinases. Chromosoma. 2013;122:191-207. 14. Busch A, Hertel KJ. Evolution of SR protein and hnRNP splicing regulatory factors. Wiley interdisciplinary reviews RNA. 2012;3:1-12. 15. Best A, Dagliesh C, Ehrmann I, Kheirollahi-Kouhestani M, Tyson-Capper A, Elliott DJ. Expression of Tra2 beta in Cancer Cells as a Potential Contributory Factor to Neoplasia and Metastasis. International journal of cell biology. 2013;2013:843781.

16. Hirschfeld M, Jaeger M, Buratti E, Stuani C, Grueneisen J, Gitsch G, et al. Expression of tumor-promoting Cyr61 is regulated by hTRA2-beta1 and acidosis. Human molecular genetics. 2011;20:2356-65. 17. Kuwano Y, Nishida K, Kajita K, Satake Y, Akaike Y, Fujita K, et al. Transformer 2beta and miR-204 regulate apoptosis through competitive binding to 3' UTR of BCL2 mRNA. Cell death and differentiation. 2015;22:815-25. 18. Grellscheid S, Dalgliesh C, Storbeck M, Best A, Liu Y, Jakubik M, et al. Identification of evolutionarily conserved exons as regulated targets for the splicing activator tra2beta in development. PLoS genetics. 2011;7:e1002390. 19. Meng J, Sun B, Zhao X, Zhang D, Zhao X, Gu Q, et al. Doxycycline as an inhibitor of the epithelial-to-mesenchymal transition and vasculogenic mimicry in hepatocellular carcinoma. Molecular cancer therapeutics. 2014;13:3107-22. 20. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25:402-8. 21. Liu TJ, Sun BC, Zhao XL, Zhao XM, Sun T, Gu Q, et al. CD133+ cells with cancer stem cell characteristics associates with vasculogenic mimicry in triple-negative breast cancer. Oncogene. 2013;32:544-53. 22. Liu T, Sun B, Zhao X, Li Y, Zhao X, Liu Y, et al. USP44+ Cancer Stem Cell Subclones Contribute to Breast Cancer Aggressiveness by Promoting Vasculogenic Mimicry. Molecular cancer therapeutics. 2015;14:2121-31. 23. Liu T, Sun B, Zhao X, Gu Q, Dong X, Yao Z, et al. HER2/neu expression correlates with vasculogenic mimicry in invasive breast carcinoma. Journal of cellular and molecular medicine. 2013;17:116-22. 24. Liu T, Sun B, Zhao X, Li Y, Gu Q, Dong X, et al. OCT4 expression and vasculogenic mimicry formation positively correlate with poor prognosis in human breast cancer. International journal of molecular sciences. 2014;15:19634-49. 25. Liu T, Niu Y, Yu Y, Liu Y, Zhang F. Increased gamma-tubulin expression and P16INK4A promoter methylation occur together in preinvasive lesions and carcinomas of the breast. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO. 2009;20:441-8. 26. Xu Y, Gao XD, Lee JH, Huang H, Tan H, Ahn J, et al. Cell type-restricted activity of hnRNPM promotes breast cancer metastasis via regulating alternative splicing. Genes & development. 2014;28:1191-203. 27. Kato T, Inoue H, Imoto S, Tamada Y, Miyamoto T, Matsuo Y, et al. Oncogenic roles of TOPK and MELK, and effective growth suppression by small molecular inhibitors in kidney cancer cells. Oncotarget. 2016;7:17652-64. 28. Nagano K, Imai S, Zhao X, Yamashita T, Yoshioka Y, Abe Y, et al. Identification and evaluation of metastasis-related proteins, oxysterol binding protein-like 5 and calumenin, in lung tumors. International journal of oncology. 2015;47:195-203. 29. Kurehara H, Ishiguro H, Kimura M, Mitsui A, Ando T, Sugito N, et al. A novel gene, RSRC2, inhibits cell proliferation and affects survival in esophageal cancer patients. International journal of oncology. 2007;30:421-8. 30. Yang C, He P, Liu Y, He Y, Yang C, Du Y, et al. Assay of serum CEACAM1 as a potential biomarker for breast cancer. Clinica chimica acta; international journal of clinical chemistry. 2015;450:277-81. 31. Chang CY, Lin SC, Su WH, Ho CM, Jou YS. Somatic LMCD1 mutations promoted cell migration and tumor metastasis in hepatocellular carcinoma. Oncogene. 2012;31:2640-52. 32. Albrecht I, Bieri R, Leu A, Granacher P, Hagmann J, Kilimann MW, et al. Paralemmin-1 is expressed in lymphatic endothelial cells and modulates cell migration, cell maturation and tumor lymphangiogenesis. Angiogenesis. 2013;16:795-807. 33. Lu G, Zhang Q, Huang Y, Song J, Tomaino R, Ehrenberger T, et al. Phosphorylation of ETS1 by Src family kinases prevents its recognition by the COP1 tumor suppressor. Cancer cell. 2014;26:222-34.

34. Cooper TA. Use of minigene systems to dissect alternative splicing elements. Methods. 2005;37:331-40. 35. Oltean S, Bates DO. Hallmarks of alternative splicing in cancer. Oncogene. 2014;33:5311-8. 36. David CJ, Manley JL. Alternative pre-mRNA splicing regulation in cancer: pathways and programs unhinged. Genes & development. 2010;24:2343-64. 37. Park JS, Young Yoon S, Kim JM, Yeom YI, Kim YS, Kim NS. Identification of novel genes associated with the response to 5-FU treatment in gastric cancer cell lines using a cDNA microarray. Cancer letters. 2004;214:19-33.

Figure Legends

Figure 1 Paclitaxel (PTX) treatment induced a higher TRA2A expression in MDA-MB-231 cells. (A) After PTX treatment, the majority of these cells died within 2 weeks, black arrowhead indicated few survived cells. (B) The survived cells resumed proliferation and established characteristic clones within 3-4 weeks (black arrow). (C) RNA splicing pathway was activated after PTX treatment. (D) STRING analysis showed there were stronger interactions among hnRNPM, hnRNPu, hnRNPA2B1, SRSF6 and DHX9 while a weaker interaction between hnRNPm and TRA2A or hnRNPu and TRA2A. (E) qPCR validated that the expression TRA2A and hnRNPm was higher than other splicing factors. The product of TRA2A gene paralogs, i.e. TRA2B, didn’t show significant increase in contrast to higher TRA2A expression after PTX treatment. (F) Western blotting showed an increase in TRA2A and hnRNPm protein levels under PTX treatment compared to normal conditions. TRA2A protein level didn’t increase following exogenous hnRNPm expression. (G) TRA2A protein level didn’t decrease following hnRNPm knockdown. TRA2A protein level showed a dramatic increase in hnRNP knockdown MDA-MB-231 cells with PTX treatment.

Figure 2 TRA2A expression promoted PTXR, cancer cells survival and invasion. (A) MDA-MB-231 and Hs578T cells transfected with TRA2A overexpression plasmid expressed higher levels of TRA2A than control cells. (B) The IC50 value for PTX was higher in TRA2A overexpressing cells. (C) TRA2A overexpressing cells formed more colonies than control cells when given PTX treatment. (D) TRA2A silencing in PTXR cells resulted in fewer colonies than control cells when given PTX treatment. (E) TRA2A silencing in PTXR cells induced IC50 decreased from 208.4 nM to 59.6 nM. (F) The TRA2A transfected cells displayed the faster speed of wound healing. (G-H) The increased migration (G) and invasion (H) ability was observed in TRA2A overexpressing cells.

Figure 3 Global splicing and transcriptional regulation by TRA2A. (A) Example of alternative exon affected by TRA2A. (B) Quantification of the different AS events affected by TRA2A. (C) AS events could be negatively or positively regulated by TRA2A. (D) Gene ontology of TRA2A-regulated AS targets. (E-F) Functional association network of TRA2A-regulated AS targets. The genes in (D) were analyzed using the STRING database, and subgroups were marked according to their functions. (G) Gene ontology analyses of TRA2A-regulated gene expression events. (H) Validation of PCNA, CDC25A and CDC6 expression changes by qRT-PCR. (I) hnRNPm and SRSF6 showed an increased expression following TRA2A upregulation. 20

Figure 4 TRA2A expression and the AS events regulated by TRA2A played key roles in PTXR of TNBC. (A) Validation of TRA2A-regulated cassette exon events by qRT-PCR using MDA-MB-231 cells transfected with TRA2A or control vectors (* mean fold change was >2 or <-2). (B) The relative changes of splicing indexes obtained from qRT-PCR were highly correlated to those observed by microarray data and TAC analysis. (C-K) CALU, RSRC2 and PALM isoform shift occurred under PTX treatment. qRT-PCR analysis of levels of CALU (C), RSRC2 (F) and PALM (I) isoforms using primers that specifically detected either short isoform or long isoform containing variable exons. The ratio of CALUl to CALUs (D) and RSRC2l to RSRC2s (G) was increased while the ratio of PALMl to PALMs (J) was decreased. The total levels of CALU (E), RSRC2 (H) and PALM (K) weren’t significantly changed. (L) TRA2A positive signal was located in the nucleus in TNBC tissue. The negative expression of TRA2A presented in non-TNBC tissue (black arrow indicated a normal breast duct). (M) TRA2A expression was higher in TNBC than in non-TNBC. (N) TRA2A expression significantly associated with poorer overall survival and disease-specific survival. (O) For TNBC patients with PTX chemotherapy, elevated TRA2A expression was significantly associated with poor survival. (P) There was no significant difference in the survival between TRA2A positive and TRA2A negative of non-TNBC patients with PTX chemotherapy.

Figure 5 RSRC2s isoform expression played roles in PTXR of TNBC. (A-B) Both TRA2A protein and mRNA expression was higher in PTXR than in non-PTXR of TNBC. (C-D) RSRC2l (C) showed dramatically increased expression while RSRC2s (D) displayed decreased expression in PTXR than in non-PTXR. (E-H) The level of CALUl (E), CALUs (F), PALMl (G) and PALMs (H) did not differ significantly between PTXR and non-PTXR. (I) The level of TRA2A, RSRC2l and RSRC2s mRNA was significantly different between TRA2A protein positive and negative cases. However no significant differences were found in CALUl, CALUs, PALMl and PALMs levels. (J-K) ROC curve analysis of TRA2A protein, TRA2A mRNA and RSRCl for PTXR prediction. (L-M) ShRNA mediated RSRC2 knockdown (mRNA and protein) resulted in PTXR in MDA-MB-231 cells. Compared with control cells, cells with RSRC2 knockdown formed more colonies (L) and showed a worse IC50 (M) when given PTX treatment. (N-O) Forced RSRC2 expression in MDA-MB-231TRA2A cells led to fewer colonies formation (N) and lower IC50 (O) when given PTX treatment.

Figure 6 The decreased RSRC2 protein expression could be induced by the shift from RSRC2s to RSRC2l mRNA which controlled by TRA2A. (A) Normal breast tissue showed strong RSRC2 protein expression and in most of low grade TNBC, a moderate to strong staining of RSRC2 could be observed. In contrast, a decreased staining was observed in high grade TNBC. (B) RSRC2 protein expression was related to TNBC grade. (C) RSRC2 protein expression showed negatively correlation with TRA2A protein expression. (D) The mean expression of RSRC2l mRNA was lower in RSRC2+ cases than in RSRC2- cases. (E) In contrast, the mean expression of RSRC2s mRNA was higher in RSRC2+ cases than in RSRC2- cases. (F) Survival analysis showed that the decreased RSRC2 protein expression was significantly associated with poor survival. (G) MDA-MB-231 cells exhibited a significant shift in RSRC2 AS toward the RSRC2 exon4 inclusion isoform under PTX treatment or TRA2A upregulation. (H) RSRC2 protein expression was decreased under PTX 21

treatment or TRA2A upregulation. (I) Schematic of the RSRC2 minigene construct. The exon4 and its flanking introns were inserted between two constitutive exons. (J) RT-PCR analysis of RNA harvested from 293T cells cotransfected with the RSRC2 minigene containing upstream sequence and the indicated amounts of TRA2A showed that TRA2A promoted RSRC2 exon4 inclusion in a dose-dependent manner. In contrast, cells transfected with the minigene containing downstream sequence displayed no difference in RSRC2 AS pattern compared with control cells. (K) qRT-PCR analysis of RSRC2 exon4 inclusion using RNA samples shown in C. Ratios of exon4 inclusion and skipping were plotted. (L) RT-PCR analysis of cells transfected with 2 μg of TRA2A and RSRC2 minigene constructs that contained upstream sequence (minegene1) displayed exon4 inclusion, while replacing RAAG or AGAA in minegene1 sequence (minigene3) reduced exon4 inclusion.

TRA2A promoted paclitaxel resistance and tumor progression in triple negative breast cancers via regulating alternative splicing

Tieju Liu, Huizhi Sun, Dongwang Zhu, et al.

Mol Cancer Ther Published OnlineFirst April 17, 2017.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-17-0026

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2017/04/15/1535-7163.MCT-17-0026.DC1

Author Author manuscripts have been peer reviewed and accepted for publication but have not yet been Manuscript edited.

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/early/2017/04/15/1535-7163.MCT-17-0026. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.