Isoform switching as a key mechanism in chemotherapy resistance in triple-negative breast cancer

Isoform switching as a key mechanism in chemotherapy resistance in triple-negative breast cancer | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Isoform switching as a key mechanism in chemotherapy resistance in triple-negative breast cancer Katarzyna Nowis, Maria Sąsiadek, Dariusz Martynowski, Dorota Kujawa, and 9 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7041176/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Triple-negative breast cancer (TNBC) is characterized by limited treatment options and high variability in response to neoadjuvant chemotherapy (NAC). While DNA-level alterations have been widely studied, post-transcriptional regulation through alternative splicing remains unexplored in this context. Methods We performed a transcriptome -wide analysis of differential isoform usage in pre-treatment TNBC biopsies from patients stratified by NAC response. Using IsoformSwitchAnalyzer and STRING, we assessed the functional consequences of isoform switching alterations in coding potential, protein domains, and pathway involvement. Structural models of XRCC3 isoforms were generated using AlphaFold and ChimeraX. Results Non-responder exhibited significantly higher rates of isoform switching, particularly involving transcription start/termination site changes and intron retention. Enrichment analyses revealed immune-related pathway signatures in complete responders and DNA repair in both complete and partial responders. Among key genes, the XRCC3 emerged as a notable candidate, with non-responder showing shift toward truncated isoform lacking domains required for interactions with RAD51 and RAD51C. This structural loss likely impairs homologs recombination repair and may contribute to the observed resistance phenotype. Conclusion Isoform switching is a significant regulatory mechanism associated with chemotherapy response in TNBC. Splicing alterations affecting DNA repair and immune may serve as predictive biomarkers. These findings support the integration of isoform-level analysis into clinical transcriptomics for precision oncology. Triple negative breast cancer TNBC alternative splicing transcriptomics therapy response neoadjuvant therapy Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 STATEMENT OF TRANSLATIONAL RELEVANCE Tripple-negative breast cancer (TNBC) remains a therapeutically challenging subtype due to the lack of targeted treatment options and high inter-patient variability in chemotherapy response. In this study, we demonstrate that alternative splicing and isoform switching contribute to differential therapy responses in TNBC. Notably, we identified enrichment of a known, truncated XRCC3 isoform (ENST00000554974) in non-responders. This isoform lacks critical domains required for interaction with RAD51 and RAD51C, likely impairing homologous recombination repair. Importantly, total XRCC3 gene expression did not differ between groups, underscoring the need of isoform-level resolution. Our findings suggest that transcript-based profiling of XRCC3 isoforms could help identify TNBC patients with splicing-driven HRR deficiency, even in the absence of BRCA mutations. This opens the possibility of expanding synthetic lethality-based therapeutic strategies, such as the use of PARPi or ATR/CHK1i, to BRCA wild-type patient exhibiting XRCC3 isoform imbalance. Integrating isoform-level diagnostics may improve patient stratification and guide personalized treatment selection in TNBC. INTRODUCTION Triple-negative breast cancer (TNBC), characterized by the absence of estrogen(ER), progesterone receptors (PR), and human epidermal growth factor receptor 2 (ERBB2/HER2) expression, accounts for approximately 15–20% of all breast cancers (BC) ( 1 – 3 ). TNBC is known for its aggressive clinical behavior, higher rates of recurrence, and limited treatment options due to the lack of targeted therapies ( 3 ). Neoadjuvant chemotherapy (NAC) is the standard systemic treatment for TNBC, aiming to reduce tumor size, eliminate lymph node metastases and micrometastatic disease before surgery ( 4 ). However, response to NAC varies markedly among patients, with some achieving a pathologic complete response (pCR) while others exhibit residual disease, which is associated with a worse prognosis ( 5 , 6 ). Emerging evidence suggests that alternative splicing may also contribute to inter-patient variability in therapeutic response, including sensitivity or resistance to NAC in TNBC ( 7 ). Alternative splicing (AS) is a fundamental post-transcriptional mechanism that enables a single gene to produce multiple mRNA isoforms, thereby contributing to proteomic diversity. Dysregulation of AS has been implicated in various cancers, where it can lead to the production of isoforms that enhance proliferation, metastasis and resistance to therapy ( 8 , 9 ). The regulation of AS is primarily governed by a complex network of splicing factors, whose dysregulation can profoundly affect transcript diversity in cancer ( 10 ) making them potential therapeutic targets. Thus, some of public attention is focused on expression, mutation and even AS of those factors. For instance, pan-cancer studies revealed 167 cancer specific splicing patterns in. splicing factors with consequences across 16 cancer types and 6904 patients ( 11 ). Moreover, splicing factors SRSF1 and SRSF3, has been linked in oncogenesis and poor prognosis ( 12 ). In turn, in TNBC CPSF4 has been shown to regulate AS of HMG20B promoting proliferation and migration ( 13 ). Beyond their impact on tumor biology, AS alterations have also emerged as important determinants of therapeutic response. Increasing evidence suggests that distinct splicing profiles can mediate sensitivity or resistance to chemotherapy, radiotherapy, targeted therapies, or immunotherapy, further underscoring functional significance of splicing regulation in cancer. For instance, SRSF1 improves drug resistance through alternative splicing in MYO1B in BC ( 14 ), BRCA1 exon skipping resulting in ineffective PARPi ( 15 ), or the failure of CAR-T therapy in leukemias due to retention of CD19 intron ( 16 ). These findings highlight the potential of AS events not only as biomarkers of treatment but also targets for overcoming resistant mechanisms. Increasing attention has also been directed to the role of AS in TNBC. Recent studies have demonstrated that transcriptomic reprogramming through AS can drive apoptosis, modulate key oncogenic regulators such as MDM4, or influence prognosis and treatment vulnerabilities through aberrant splicing events ( 17 – 21 ). However, these studies primarily focused on individual genes, specific splicing factors, or targeted modulation., leaving a broader understanding of splicing alterations in clinical context unexplored. To address this gap, we performer a comprehensive, transcriptome-wide analysis of differentia isoform usage in TNBC patients stratified by NAC response. By integrating differential isoform expression with functional enrichment and pathway analyses, we aimed to identify splicing-regulated biological processes associated with treatment outcomes. Our findings provide novel insights into the molecular mechanisms underlying chemotherapy response and resistance, highlighting alternative splicing as a potential source of predictive biomarkers and therapeutic targets in TNBC. MATERIALS AND METHOD Material This study included 50 biopsy samples of breast tumor tissue taken before preoperative chemotherapy from female patients with triple-negative breast cancer (TNBC) diagnosed and treated at the Lower Silesian Oncology, Pulmonology and Hematology Center in Wrocław. Systemic preoperative treatment was based on standard chemotherapy. Immunotherapy was not used in the study group. Response to preoperative systemic treatment was assessed on the basis of postoperative histopathological examination. Detailed information on sample acquisition, patient selection criteria, ethical approvals, and associated clinical data is provided in a previous publication ( 22 ). Data processing The first step was to assess the quality of the raw sequencing data using PRINSEQ-lite v0.20.4 ( 23 )(RRID:SCR_005454). Quality metrics including base quality scores, sequence complexity and GC content are evaluated to identify low quality reads and overrepresented sequences. After the initial quality assessment, read sorting was performed by reordering reads based on their identifiers. Sorting ensures that paired-end reads were consistently aligned during downstream processing. Then the AdapterRemoval v2.3.1 tool was used to remove adapter sequences and low-quality bases (below a score of 30) and to discard fragments shorter than 20 nucleotides ( 24 )(RRID:SCR_011834). A second round of quality assessment was performed using PRINSEQ-lite v0.20.4 to confirm the effectiveness of the trimming and filtering steps (Supplementary Table 1). HISAT2 v2.2.1 was used to map the processed reads to GRCh38 reference genome index, sourced from the HISAT2 website ( http://daehwankimlab.github.io/hisat2/download/ , grch38_tran.tar.gz)(RRID:SCR_015530)( 25 ). This splice-aware aligner efficiently handled large and complex transcriptomes while supporting accurate identification of intronic and splicing regions. The aligned reads were output in BAM format for downstream transcript assembly. The aligned reads were processed using StringTie v2.1.7 for transcript assembly and quantification ( 26 )(RRID:SCR_016323). StringTie reconstructs full-length transcripts by leveraging splice junction information and provides accurate estimations of transcript expression levels. The generated GTF files define the structure and abundance of transcripts for each sample, enabling comprehensive transcriptome profiling. Differential gene expression analysis, isoform switching and functional analysis Differential gene expression was performed using DESeq2 and SARTools R package ( 27 , 28 )(RRID:SCR_00154, SCR_016533). Batch correction was made with ComBat (RRID:SCR_010974) Benjamini and Hochberg padj values for gene expression was used for visualization on isoform switch plots with IsoformSwitchAnalyzeR( 29 ). The final stage of the pipeline involved analyzing isoform switching using IsoformSwitchAnalyzeR. GTF files generated by StringTie and the associated expression matrices were used as input to identify differential isoform expression across experimental conditions. IsoformSwitchAnalyzeR evaluates the functional consequences of isoform switching, including domain analysis and coding potential assessment, offering insights into potential biological impacts. Functional consequences of isoform switching analyzed using IsoformSwitchAnalyzeR was performed based on outputs of several external tools were integrated into the as follows: CPC2, signalP6, DeepLoc2, DeepTMHMM to asses coding potential, identifying functional elements like signal peptides and subcellular localizations, and analyzing transmembrane regions visualized on isoform switch plots ( 30 – 33 ). Alternative splicing-associated functional enrichment and network analysis Functional enrichment analysis was performed using the STRING database (version 12.0). Protein–protein interaction (PPI) networks were generated in STRING with a minimum required interaction score of 0.7 (high confidence) (RRID:SCR_005223). Enrichment analysis was conducted using STRING’s built-in functional enrichment test. Significance was assessed using the false discovery rate (FDR) correction, with an FDR threshold of < 0.05. The resulting enrichment data were visualized using the EnrichmentMap plugin (RRID:SCR_016052). Each node in the enrichment map represents an enriched term, and edges reflect gene overlap between terms. Node color corresponds to the significant normalized enrichment score (NES), indicating the direction and strength of enrichment in the comparison (positive NES for enrichment in responders, negative NES for enrichment in non-responders). Clusters of related terms were annotated using the AutoAnnotate and WordCloud plugins to identify key biological themes enriched in different therapy response groups. All analysis were performed using Cytoscape ( 34 ). Structural Modeling and Superposition of XRCC3 The crystal structure of human RAD51 in complex with single-stranded DNA (ssDNA) and ATP (PDB ID: 7EJC) was used as a reference template for structural superposition. The Alvinella pompejana RAD51/XRCC3 complex (PDB ID: 8GJA) was aligned onto this reference using UCSF ChimeraX ( 35 ) (RRID:SCR_015872). To model the human RAD51/XRCC3/RAD51C complex, AlphaFold3 predictions were generated using the AlphaFold Server ( https://alphafoldserver.com ). The model included three bound ATP molecules and was prepared in two variants: one containing the canonical XRCC3 isoform (ENST00000553264), and the other containing a shorter RAD51 domain (ENST00000554974) identified in isoform switching analysis. Both AlphaFold3-derived complexes were subsequently superimposed onto the hRAD51/apRAD51C/XRCC3 complex using structural alignment tools within ChimeraX. All structural visualizations and superpositions were performed in UCSF ChimeraX (version 1.9) ( 35 ). RESULTS An increase in the number of isoforms switching events has been observed in non-responders. The analysis revealed that the number of isoforms switching events varies depending on the response to treatment used. Volcano plots (Fig. 1 A) show a higher density of significantly switched isoforms in the complete response group compared to the no response group than in the partial response group compared to the no response group. The smallest number of isoform switching was observed when comparing total response to therapy with partial response (Supplementary Table 2). This suggests that non-responders exhibit the most extensive changes in isoform usage, potentially reflecting an adaptive mechanism or dysregulation associated with resistance to therapy compared to partial- and complete-responders. The MA-like plots (Fig. 1 B) illustrate that isoform switching occurs independently of large gene expression changes. Many significantly switched isoforms are observed across a wide range of gene expression levels, highlighting that isoform-level regulation can act independently of transcriptional activation. Moreover, the largest number of significant isoform switching events is observed in the comparisons involving non-responders reinforcing the notion that non-responders experience more pronounced isoform reprogramming. Venn diagrams (Fig. 1 C, D, E) further highlight differences between studied groups. The highest number of switching events is observed in comparison complete vs no response (1138) following with partial vs no response (724) and complete vs partial response ( 42 ). The greatest overlap of switched isoforms is found between no response versus complete response and no response versus partial response (401), indicating that isoform switching events distinguish responders from non-responders (Fig. 1 D). Similar observation is made for differentially used isoforms themselves (Fig. 1 C) (927, 581, 58 and 417 respectively). When considering genes containing differentially used isoforms (Fig. 1 E) and the total number of affected genes, an analogous trend is observed. The highest number of genes with differentially used isoforms is found in the comparison between complete and no response (817), followed by partial versus no response (524), and complete versus partial response ( 51 ). Notably, the largest overlap (390 genes) occurs between the no response versus complete response and no response versus partial response groups, further supporting the concept that isoform usage patterns distinguish responders from non-responders. Differential isoform usage was observed in only five genes consistently across all three comparisons, suggesting a limited but potentially crucial set of genes involved in therapy response modulation (Fig. 1 E). These results underscore the importance of alternative isoform regulation in determining treatment outcomes and highlight key molecular players that may serve as potential biomarkers or therapeutic targets. All these findings demonstrate that the extent of isoform switching increases as the response to therapy diminishes. Significant changes in alternative transcription start/termination sites and intron retention events are associated with poor therapy response The comparison of isoform switching events across therapy response groups reveals distinct patterns of alternative splicing associated with differences in therapy outcomes: nonresponders exhibit a larger frequency of alternative transcription start site (ATSS), alternative transcription termination site (ATTTS) and intron retention (IR) events, suggesting that transcriptional regulation at the initiation and termination levels may play a significant role in therapy resistance. This is supported by the significant enrichment of ATSS, ATTTS, IR in no- vs complete responders and IR in no- vs partial responders (Fig. 2 A, Supplementary Fig. 1A). Furthermore, the analysis shows that these transcriptional events are more pronounced in the non- versus complete responders than in non- versus partial responders (Fig. 2 A, Supplementary Fig. 1A), suggesting that the extent of transcriptional reprogramming correlates with the level of therapeutic resistance. Importantly, comparison of the relative contribution of splicing event categories across response-group contrasts (Fig. 2 B) indicates that ATTTS gains represent the most prominent and discriminative event type, particularly in comparisons involving non-responders. While ATSS and IR events also occur, their differential burden is less pronounced, highlighting alternative transcription termination as a dominant mechanism underlying isoform reprogramming in therapy-resistant tumors. Overall, these results highlight the importance of alternative transcription regulation in non-responders, with ATSS and especially ATTTS events emerging as key contributors to isoform reprogramming. The increased frequency and magnitude of these events in non-responders suggest that transcriptional plasticity may facilitate adaptive mechanisms that promote therapy resistance. Further investigation of genes undergoing ATSS and ATTTS switching could provide better insights into the molecular pathways that underpin poor therapeutic outcomes. Analysis of the transcriptome of tumor cells could be a prognostic marker for response to chemotherapy. Alternative splicing alters protein domains and coding potential To evaluate the impact of isoform switching on protein function, we analyzed the distribution of alternative splicing consequences across therapy response groups. Non-responders exhibited a significantly larger frequency of intron retention and ORF disruptions compared to both partial responders and complete responders (Fig. 2 C, Supplementary Fig. 1B). Complete responders also appeared to be enriched by domain alterations also observed in a larger proportion of transcripts from no response samples. To further evaluate the functional consequences of alternative splicing, we examined shifts in predicted subcellular localization between therapy response groups (Fig. 2 D). The congruence matrix highlights significant changes in localization patterns, with nuclear mRNA being particularly affected in no response samples. These disruptions may contribute to therapy resistance by altering the functional availability of key regulatory proteins involved in apoptosis, metabolism, and transcriptional control. In contrast, comparisons between complete and partial responders did not reveal statistically significant differences in splicing-related localization shifts, suggesting that splicing dysregulation is primarily linked to therapy non-response rather than partial or complete response. Key isoform switching events in treatment response. To identify key splicing-related changes in treatment resistance, we performed an isoform switching analysis comparing non-responders to partial and complete responders. The top three switching events were observed in XRCC3, ABCG1, SNHG14 , and ABCG1, RNF7, BLTP3B (for no versus complete and partial response respectively, padj < = 0.05). Among the five genes common to all comparisons in isoform switching genes (Fig. 1 E), isoform switching was also observed in TP53I3, CALML4 , and STRA6 (padj < 0.01), as well as ATP13A5 and NLRX1 (padj < 0.05)(Supplementary Table 2). XRCC3 : negative impact on BRCA-dependent homologous recombination DNA Repair Pathway in non-responders. XRCC3 encodes a DNA repair protein involved in homologous recombination. Non-responders exhibited a significant shift in dominant isoforms compared to complete responders (Fig. 3 A). The alternative splicing events resulted in the loss of 4 exonic regions from 5’ end (part of Rad51 domain shortening protein from 346 to 141 aa), consistent with isoform switching from the canonical transcript ENST00000553264 to the truncated variant ENST00000554974 (Fig. 3 A, 4 ). Structural studies have shown that the N-terminal domain (NTD) of XRCC3 forms a clamp-like interface with RAD51C, facilitating the formation of the CX3 complex, which plays a critical role in replication fork protection, restart, and RAD51 filament capping ( 36 , 37 ). In the truncated isoform, the NTD is absent, leading to the loss of the structural interface with RAD51C and likely impairing proper assembly of CX3 complex. Furthermore, the missing N-terminal region also includes part of the RAD51-interacting domain and ATP binding motif, suggesting that both RAD51C and RAD51 interactions may be disrupted (Fig. 4 B, C). These defects could impair homologous recombination, limit the ability to stabilize or restart stalled replication forks, and promote genomic instability. ABCG1 : Lipid transport disruptions in non-responders. ABCG1 , a lipid transporter, showed distinct isoform switching in comparisons of both non- versus complete and non- versus partial responders (Fig. 3 B, D). Isoform transitions led to modifications at the beginning of the transcript, resulting in the loss of the signal domain in non-responders, which may lead to altered protein localization, impaired lipid transport efficiency, or disruptions in cellular signaling pathways associated with lipid metabolism and immune response. These changes could contribute to differences in therapeutic outcomes by affecting the functional role of ABCG1 in the tumor microenvironment. SNHG14 , The Small Nucleolar Host Gene 14 SNHG14 , a long non-coding RNA, exhibited significant changes in isoform usage in non-responders comparing with complete responders (Fig. 3 C). As lncRNAs have been implicated in transcriptional regulation and chromatin remodeling, these isoform shifts could play a role in modulating gene expression patterns associated with resistance mechanisms RNF7 , Ring Finger Protein 7 An ubiquitin ligase is associated with oxidative stress response and proteasomal degradation. Isoform switching in non-responders comparing with partial responders may suggests a potential dysregulation in protein homeostasis and stress response pathways (Fig. 3 E). BLTP3B , Bridge-Like Lipid Transfer Protein Family Member 3B BLTP3B is involved in lipid transport and membrane-associated processes. The observed isoform switching in non-responders comparing with partial responders may indicate functional alterations affecting lipid signaling and cellular metabolism, which could influence therapy response (Fig. 3 F). Our findings highlight key isoform switching events that may contribute to treatment resistance by affecting DNA repair ( XRCC3 ), lipid metabolism ( ABCG1, BLTP3B ), proteasomal function ( RNF7 ), and transcriptional regulation ( SNHG14 ). Further functional validation is required to determine the precise impact of these isoform shifts on tumor biology and therapeutic response. Functional enrichment and network analysis of alternatively spliced genes To investigate whether baseline alternative splicing profiles are associated with response to neoadjuvant chemotherapy (NAC) in TNBC, functional enrichment and protein-protein interaction (PPI) network analyses were performed. In the complete versus no response groups comparison several densely connected clusters were enriched, including regulation of cytokine-mediated immune responses, pattern recognition receptor signaling, and DNA damage repair pathways. Notably, pathways such as T cell receptor signaling, Fc gamma receptor activation, MAPK cascade initiation, and TLR4 signaling were significantly enriched in baseline tumors of future complete responders. Another large cluster revealed enrichment in DNA repair pathways including homologous recombination, base excision repair, and nucleotide biosynthesis, suggesting that tumors from future complete responders already express splicing isoforms conducive to efficient DNA damage detection and repair (Fig. 5 , Supplementary Table 3A, Supplementary Fig. 2). In contrast, the partial versus no response groups comparison revealed more limited enrichment with only a few clusters reaching significance — most notably DNA repair pathways and regulation of cellular response (Fig. 6 , Supplementary Table 3B, Supplementary Fig. 3). Immune signaling clusters were largely absent, suggesting that partial responders may share certain repair-related features with complete responders, but lack the immune-related transcriptional readiness that predicts robust therapy outcomes. This implies that cancer cells can avoid immune monitoring and anti-tumor immunity ( 38 ). DISCUSSION Developing an effective anti-cancer therapy remains a significant medical challenge. Despite the increasing number of treatment strategies, a substantial proportion of patients still exhibit poor responses or complete resistance to standard approaches. Until today, exploring DNA-level markers of therapy failure has not revealed the causes underlying this. For instance, up to 40% of patients with TNBC fail to respond effectively to conventional chemotherapy regiments( 39 , 40 ). Previous efforts to identify predictive biomarkers of therapy resistance have focused primarily on DNA-level alterations, such as mutations, copy number variations, and structural rearrangements. However, these approaches often failed to explain full spectrum of resistance mechanisms in clinic. This has led to a growing interest in exploring regulatory layers beyond DNA, particularly involving epigenetic and transcriptomic dysregulation. Notably, epigenetic processes, including DNA methylation, histone modification, and RNA processing, have been shown to play critical roles in cancer development, progression, and therapy response. Among these, the dysregulation of alternative splicing has emerged as a significant contributor to tumor heterogeneity, immune evasion or drug resistance across multiple cancer types ( 8 , 41 ). Importantly, many of these splicing alterations occur in the absence of changes at the genomic level, underscoring the functional relevance of transcriptomic profiling. This study demonstrates that alternative splicing measured through isoform switching is associated with differential responses to neoadjuvant chemotherapy (NAC) in triple-negative breast cancer (TNBC). Isoform switching refers to changes in the dominant transcript isoform of a gene between two conditions—in this case, between tumor tissues from patients with complete response, partial response, or no response. The number and magnitude of isoform switching events varied depending on response level. The largest number of significant events was identified in the comparison between complete responders and non-responders, followed by partial versus non-responders with relatively few events observed between complete and partial responders. This suggests that non-responders exhibit the most pronounced isoform-level differences when compared to treatment-sensitive tumors. Importantly, these differences were detected in pre-treatment biopsy samples, supporting their potential value as predictive molecular features rather than consequences of chemotherapy exposure. In tumors derived from complete responders, isoform switching was associated with functional enrichment of immune-related signaling pathways. This included T cell receptor signaling, cytokine-mediated immune activation, and innate immune processes such as toll-like receptor 4 (TLR4) and MAPK cascades. Similar immune-related enrichment was not observed in partial responders. These findings support prior evidence that transcriptional immune activation and infiltration of immune cells are predictive of chemotherapy sensitivity in TNBC ( 1 , 42 ). Since these differences were detected in pre-treatment biopsies, transcriptomic profiling using RNA sequencing may offer a valuable tool for patient stratification before therapy. In the future, RNA-seq-based assays, such as isoform-level biomarkers could support clinical decision-making and help identify patients most likely to benefit from NAC. Partial responders, when compared to non-responders, showed enrichment of splicing alterations in genes associated with DNA repair mechanisms including homologous recombination and base excision repair. Although these processes were also observed in complete responders, the absence of immune pathway enrichment in partial responders may explain their intermediate outcome. This aligns with previous studies showing that while deficiency DNA repair pathways can enhance chemotherapy efficacy, the presence of a responsive immune environment is often required for complete tumor clearance ( 43 , 44 ), we note that our analysis does not infer the directionality of isoform usage or functional activation. Therefore, while alternative splicing appears to impact genes within these pathways, the consequences on pathway activity, whether gain or loss of pathway, remain to be elucidated. Nonetheless, the simultaneous disruption of DNA repair related transcripts and immune related-components in complete responders suggests the possibility of dual vulnerability, potentially contributing to synthetic lethality mechanisms. Several genes with differential isoform usage contributed to the observed response patterns. Among the most prominent isoform switching events distinguishing non-responders from responders, XRCC3 emerged as a key gene of interest. XRCC3 encodes a protein critical for the homologous recombination repair (HRR) pathway, specifically in resolving DNA double-strand breaks (DSBs). It is a member of the RAD51 paralog family and functions as part of the CX3 complex, which acts downstream of RAD51 recruitment in the BRCA1-BRCA2-dependent HRR cascade. This complex binds to Holliday junctions and stalled replication forks, facilitating the resolution of HR intermediates during the mitotic cycle, often in coordination with GEN1 ( 45 – 49 ). The alternative splicing events observed in non-responders led to the loss of 4 exonic regions from the 5′ end of the transcript, potentially compromising domains critical for RAD51 and RAD51C interaction and complex formation. Structural modeling (Fig. 4 ) indicates that this shortened isoform lacks key interaction surfaces necessary for proper CX3 complex assembly. Such a variant may exert a dominant-negative effect by competing with the full0length protein while failing to mediate functional interactions. Such structural alterations may impair the cell’s ability to resolve DNA damage effectively, leading to genomic instability — an established contributor to chemoresistance. This observation contrasts with previous studies, which reported that XRCC3 overexpression enhances resistance to DNA-damaging agents such as temozolomide, cisplatin or radiotherapy, by promoting homologous recombination repair and limiting apoptosis Notably, in our dataset, overall XRCC3 gene expression did not differ between groups, suggesting that gene-level quantification alone would have failed to detect any regulatory involvement of XRCC3 in therapy outcome. Only through isoform-resolved analysis was it possible to uncover splicing-dependent structural changes that may impair XRCC3 function. This highlights the functional relevance of AS as post-transcriptional regulatory mechanism that can modulate protein activity independently of gene expression level. Our results add a novel isoform-level dimension to this relationship, highlighting the potential of XRCC3 splicing as a predictive biomarker and functional contributor to therapy resistance in TNBC. ABCG1 , a gene involved in cholesterol and lipid transport, also exhibited distinct isoform switching in non-responders compared to both complete and partial responders. The altered transcript lacked key signal peptide regions, which may influence subcellular localization and impair ABCG1’s function in lipid regulation or immune signaling. Beyond its canonical role in intracellular cholesterol efflux, ABCG1 has been shown to enhance tumor-promoting capacity by conferring stem like properties to cancer cells and mediating chemoresistance in multiple malignancies. It may also act as a kinase-like regulator by phosphorylating downstream targets that support tumor growth. Moreover, ABCG1 modulates macrophage polarization in the tumor microenvironment, contributing to the formation of an immunosuppressive niche that favors cancer progression ( 50 ). The long non-coding RNA SNHG14 was differentially spliced between response groups, particularly in non-responders. SNHG14 has been shown to regulate the ERK/MAPK signaling pathway and to promote proliferation and chemoresistance in TNBC models ( 51 ). Interestingly, meta-analyses across various cancer types have demonstrated that elevated SNHG14 expression correlates with poor overall survival, larger tumor size, advanced stage, lymph node involvement and distant metastasis, supporting its role as a general marker of aggressive tumor behavior ( 52 ). Additional isoform switching was observed in genes such as RNF7 and BLTP3B , implicated in proteostasis and lipid transport, respectively. These findings further support the idea that splicing alterations may affect multiple cellular systems critical to chemotherapy sensitivity. RNF7 (SAG or ROC2) encodes a component of the SCF E3 ubiquintin ligase complex and has been implicated in tumor progression and therapy resistance across several cancer types, including prostate, pancreatic and renal carcinomas. Its pro-oncogenic functions have been linked to activation of the PI3K/Akt and JAK/STAT3 pathways, promotion of the cell cycle progression and glycolytic reprograming ( 53 – 55 ). BLTP3B is involved in intracellular lipid trafficking and endosome to Golgi transport, but its role in cancer is not yet defined ( 56 ). Nonetheless, perturbations in lipid homeostasis are increasingly recognized as drivers of tumor growth and drug resistance ( 57 ). In addition to major isoform switching events, we identified five genes— TP53I3 , CALML4 , STRA6 , ATP13A5 , and NLRX1 —that were consistently differentially spliced across all therapy response comparisons. These genes are implicated in apoptosis ( TP53I3 ), calcium signaling ( CALML4 ), immune modulation and vitamin A transport ( STRA6 ), cation transporter ( ATP13A5 ), and mitochondrial immune regulation ( NLRX1 ) ( 58 – 63 ). Their recurring presence suggests a potential core set of isoform-level regulators involved in therapy resistance. Notably, isoform switching in NLRX1 , a gene known to modulate oxidative stress and TNF-α signaling, may reflect mitochondrial reprogramming in non-responders. Similarly, changes in STRA6 splicing may impact JAK/STAT signaling and retinoic acid metabolism, both previously linked to chemoresistance ( 60 , 63 ). These isoform switching events were also associated with structural consequences at the protein level, including loss of coding potential, open reading frame (ORF) disruption, altered domain content, and changes in predicted subcellular localization. Such consequences may affect key aspects of protein function, including signaling, transport, or repair activity, in ways that support therapy resistance. Previous pan-cancer analyses have shown that alternative splicing can significantly impact tumor progression through these mechanisms ( 8 ). For example, mis-splicing can lead to the inclusion or exclusion of critical protein domains, frameshifts, or premature termination codons that alter protein stability and function ( 64 ). Importantly, aberrant RNA splicing is a molecular characteristic present in almost all types of tumors, contributing to tumorigenesis and progression. It plays role as potential target, shedding light on new perspectives for anti-cancer strategies ( 65 ). Because this study focused exclusively on pre-treatment tumor samples, the observed isoform switching events may serve as predictive biomarkers for NAC response. Integrating isoform usage data with mutational profiles, immune infiltration measures, or clinical features may improve patient stratification and guide personalized treatment strategies in TNBC. Further validation and functional characterization of these splicing events will be necessary to confirm their relevance and mechanistic roles in therapy resistance. CONSLUSION This study demonstrates that isoform switching is a prominent and functionally relevant mechanism associated with therapy response in triple-negative breast cancer. Transcriptome-wide analysis of pre-treatment samples revealed extensive isoform reprogramming in non-responders compared to partial and complete responders, implicating alternative splicing events in the development of chemoresistance. Key splicing alterations affected genes involved in DNA repair, lipid metabolism, immune signaling, and transcriptional regulation, with notable disruptions in protein domains, coding potential, and subcellular localization. These findings support the role of isoform-level regulation as a predictive molecular layer, distinct from gene expression alone, and highlight splicing-derived isoforms as potential biomarkers and therapeutic targets in TNBC. However, the assignment of oncogenic or tumor-suppressive roles of specific splicing events remains challenging. While some isoform shifts, such as the truncation of XRCC3 , allowed domain-level inference of functional loss, many events lacked sufficient structural or mechanistic annotation to determine their biological impact. Moreover, the volume and complexity of isoform-level data present substantial challenges for interpretation. Although specific functional categories emerged, the diversity of isoform changes across individual genes makes it difficult to draw overarching conclusions. This underscores the need for focused experimental follow-up, including domain mapping and functional assays to validate the biological relevance of splicing events in therapy response. Conclusively, these findings emphasize the pivotal role of AS in shaping the therapeutic response landscape in TNBC, reinforcing the necessity of an integrative and experimental approach to comprehensively explore and apply isoform-based biomarkers in clinical decision-making. Limitations The main limitation of this study is the relatively small sample size, which may affect statistical power and generalizability. Additionally, the functional impact of identified isoform switching events was inferred from in silico predictions and requires experimental validation. The study also focused on pre-treatment samples only, without capturing dynamic splicing changes during therapy. Declarations ETHICAL APPROVAL AND CONSENT TO PARTICIPATE The project was approved by Bioethics Committee of Wroclaw Medical University (consent no. KB 611/2019). All patients signed an informed consent form before the genetic test, and all procedures performed in this study were in accordance with the principles for medical research of the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. CONSENT FOR PUBLICATION Not applicable AVAILABILITY OF SUPPORTING DATA RNA-seq data are available in the NCBI BioProject database (accession: PRJNA964497): https://www.ncbi.nlm.nih.gov/bioproject/PRJNA964497. COMPETING INTERESTS Authors declare no potential conflicts of interest FUNDING: This research was financed through a statutory subsidies by the Ministry of Health as part of the Department of Oncology Wroclaw Medical University research grant SUBZ.C280.25.016 (record number in the Simple System), statutory subsidies by the Minister of Health as part of the Department of Genetics Wroclaw Medical University research grants SUBK.A290.22.077, SUBZ.A290.23.067, and SUBZ.C280.24.063 (record number in the Simple System) and statutory subsidies from the Genomics and Bioinformatics Laboratory of the Hirszfeld Institute of Immunology and Experimental Therapy PAS (501-26). AUTOR CONTRIBUTIONS K. Nowis: Conceptualization, methodology, bioinformatic analysis (RNA-seq processing, differential gene and isoform expression, isoform switch analysis, functional enrichment), data curation, investigation, writing original draft and preparing figures, writing review and editing, M. Sąsiadek: Funding and revision, D. Martynowski: Structural bioinformatics analysis and visualization D. Kujawa: Genetic material quality control, libraries preparation, libraries quality control, sequencing on Illumina platform, I. Laczmanska: Molecular interpretation insights, clinical methodology, writing review and editing, P. Karpiński: Support in analysis , M. Ekiert, E. Iwaneczko, B. Szynglarewicz, P. Kasprzak: Patient treatment, clinical data collection, result clinical interpretation , M. Abrahamowska: Specifically performing the experiments, data collection, provision of study materials, reagents and laboratory samples , R. Matkowski: Funding and clinical revision, L. Laczmanski: Supervision, review and editing. ACKNOWLEGMENTS Molecular graphics and analyses performed with UCSF ChimeraX, developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco, with support from National Institutes of Health R01-GM129325 and the Office of Cyber Infrastructure and Computational Biology, National Institute of Allergy and Infectious Diseases. References Bianchini G, Balko JM, Mayer IA, Sanders ME, Gianni L. Triple-negative breast cancer: challenges and opportunities of a heterogeneous disease. Nat Rev Clin Oncol. 2016 Nov;13(11):674–90. Gianni L, Colleoni M, Bisagni G, Mansutti M, Zamagni C, Del Mastro L, et al. Effects of neoadjuvant trastuzumab, pertuzumab and palbociclib on Ki67 in HER2 and ER-positive breast cancer. Npj Breast Cancer. 2022 Jan 10;8(1):1. Xiong N, Wu H, Yu Z. Advancements and challenges in triple-negative breast cancer: a comprehensive review of therapeutic and diagnostic strategies. Front Oncol. 2024 May 28;14:1405491. Tufano AM, Teplinsky E, Landry CA. Updates in Neoadjuvant Therapy for Triple Negative Breast Cancer. Clin Breast Cancer. 2021 Feb;21(1):1–9. Antonini M, Mattar A, Pereira TM, Oliveira LL, Teixeira MD, Amorim AG, et al. Pathologic Complete Response and Breast Cancer Survival Post-Neoadjuvant Chemotherapy: A Systematic Review and Meta-Analysis of Real-World Data. Heliyon. 2025 Mar;e43069. Cortazar P, Zhang L, Untch M, Mehta K, Costantino JP, Wolmark N, et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis. The Lancet. 2014 Jul;384(9938):164–72. Martínez-Montiel N, Anaya-Ruiz M, Pérez-Santos M, Martínez-Contreras R. Alternative Splicing in Breast Cancer and the Potential Development of Therapeutic Tools. Genes. 2017 Oct 5;8(10):217. Climente-González H, Porta-Pardo E, Godzik A, Eyras E. The Functional Impact of Alternative Splicing in Cancer. Cell Rep. 2017 Aug;20(9):2215–26. Oltean S, Bates DO. Hallmarks of alternative splicing in cancer. Oncogene. 2014 Nov 13;33(46):5311–8. Anczuków O, Krainer AR. Splicing-factor alterations in cancers. RNA. 2016 Sep;22(9):1285–301. Cheng R, Xiao L, Zhou W, Jin X, Xu Z, Xu C, et al. A pan-cancer analysis of alternative splicing of splicing factors in 6904 patients. Oncogene. 2021 Sep 2;40(35):5441–50. DeLigio JT, Stevens SC, Nazario-Muñoz GS, MacKnight HP, Doe KK, Chalfant CE, et al. Serine/Arginine–Rich Splicing Factor 3 Modulates the Alternative Splicing of Cytoplasmic Polyadenylation Element Binding Protein 2. Mol Cancer Res. 2019 Sep 1;17(9):1920–30. Li Y, others. CPSF4-mediated regulation of alternative splicing of HMG20B facilitates the progression of triple-negative breast cancer. J Transl Med. 2024; Li C, Wang L, Liu Z, Wang X, Sun L, Song X, et al. Cyperotundone promotes chemosensitivity of breast cancer via SRSF1. Front Pharmacol. 2025 Mar 19;16:1510161. Nesic K, Krais JJ, Wang Y, Vandenberg CJ, Patel P, Cai KQ, et al. BRCA1 secondary splice-site mutations drive exon-skipping and PARP inhibitor resistance. Mol Cancer. 2024 Aug 5;23(1):158. Asnani M, Hayer KE, Naqvi AS, Zheng S, Yang SY, Oldridge D, et al. Retention of CD19 intron 2 contributes to CART-19 resistance in leukemias with subclonal frameshift mutations in CD19. Leukemia. 2020 Apr;34(4):1202–7. Sudhakaran M, Navarrete TG, Mejía-Guerra K, Mukundi E, Eubank TD, Grotewold E, et al. Transcriptome reprogramming through alternative splicing triggered by apigenin drives cell death in triple-negative breast cancer. Cell Death Dis. 2023 Dec 13;14(12):824. Yu S, Si Y, Yu J, Jiang C, Cheng F, Xu M, et al. SNRPB2 promotes triple‐negative breast cancer progression by controlling alternative splicing of MDM4 pre‐ mRNA. Cancer Sci. 2024 Dec;115(12):3915–27. Gong S, Song Z, Spezia-Lindner D, Meng F, Ruan T, Ying G, et al. Novel Insights Into Triple-Negative Breast Cancer Prognosis by Comprehensive Characterization of Aberrant Alternative Splicing. Front Genet. 2020 Jun 11;11:534. Wu S, Wang J, Zhu X, Chyr J, Zhou X, Wu X, et al. The Functional Impact of Alternative Splicing on the Survival Prognosis of Triple-Negative Breast Cancer. Front Genet. 2021 Jan 14;11:604262. Caggiano C, Petrera V, Ferri M, Pieraccioli M, Cesari E, Di Leone A, et al. Transient splicing inhibition causes persistent DNA damage and chemotherapy vulnerability in triple-negative breast cancer. Cell Rep. 2024 Sep;43(9):114751. Supplitt S, Karpinski P, Sasiadek M, Laczmanski L, Kujawa D, Matkowski R, et al. The analysis of transcriptomic signature of TNBC—searching for the potential RNA-based predictive biomarkers to determine the chemotherapy sensitivity. J Appl Genet. 2025 Feb;66(1):171–82. Schmieder R, Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011 Mar 15;27(6):863–4. Schubert M, Lindgreen S, Orlando L. AdapterRemoval v2: rapid adapter trimming, identification, and read merging. BMC Res Notes. 2016 Dec;9(1):88. Kim D, Paggi JM, Park C, Bennett C, Salzberg SL. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat Biotechnol. 2019 Aug;37(8):907–15. Kovaka S, Zimin AV, Pertea GM, Razaghi R, Salzberg SL, Pertea M. Transcriptome assembly from long-read RNA-seq alignments with StringTie2. Genome Biol. 2019 Dec;20(1):278. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014 Dec 5;15(12):550. Varet H, Brillet-Guéguen L, Coppée JY, Dillies MA. SARTools: A DESeq2- and EdgeR-Based R Pipeline for Comprehensive Differential Analysis of RNA-Seq Data. Mills K, editor. PLOS ONE. 2016 Jun 9;11(6):e0157022. Vitting-Seerup K, Sandelin A. IsoformSwitchAnalyzeR: analysis of changes in genome-wide patterns of alternative splicing and its functional consequences. Berger B, editor. Bioinformatics. 2019 Nov 1;35(21):4469–71. Kang YJ, Yang DC, Kong L, Hou M, Meng YQ, Wei L, et al. CPC2: a fast and accurate coding potential calculator based on sequence intrinsic features. Nucleic Acids Res. 2017 Jul 3;45(W1):W12–6. Teufel F, Almagro Armenteros JJ, Johansen AR, Gíslason MH, Pihl SI, Tsirigos KD, et al. SignalP 6.0 predicts all five types of signal peptides using protein language models. Nat Biotechnol. 2022 Jul;40(7):1023–5. Thumuluri V, Almagro Armenteros JJ, Johansen AR, Nielsen H, Winther O. DeepLoc 2.0: multi-label subcellular localization prediction using protein language models. Nucleic Acids Res. 2022 Jul 5;50(W1):W228–34. Hallgren J, Tsirigos KD, Pedersen MD, Almagro Armenteros JJ, Marcatili P, Nielsen H, et al. DeepTMHMM predicts alpha and beta transmembrane proteins using deep neural networks [Internet]. 2022 [cited 2025 May 30]. Available from: http://biorxiv.org/lookup/doi/10.1101/2022.04.08.487609 Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: A Software Environment for Integrated Models of Biomolecular Interaction Networks. Genome Res. 2003 Nov;13(11):2498–504. Meng EC, Goddard TD, Pettersen EF, Couch GS, Pearson ZJ, Morris JH, et al. UCSF ChimeraX: Tools for structure building and analysis. Protein Sci. 2023 Nov;32(11). Longo MA, Roy S, Chen Y, Tomaszowski KH, Arvai AS, Pepper JT, et al. RAD51C-XRCC3 structure and cancer patient mutations define DNA replication roles. Nat Commun. 2023 Jul 24;14(1). Szakal B, Branzei D. Hot on RAD51C: structure and functions of RAD51C‐XRCC3. Mol Oncol. 2023 Oct;17(10):1950–2. Dutta S, Ganguly A, Chatterjee K, Spada S, Mukherjee S. Targets of Immune Escape Mechanisms in Cancer: Basis for Development and Evolution of Cancer Immune Checkpoint Inhibitors. Biology. 2023 Jan 30;12(2):218. Baselga J, Bradbury I, Eidtmann H, Di Cosimo S, De Azambuja E, Aura C, et al. Lapatinib with trastuzumab for HER2-positive early breast cancer (NeoALTTO): a randomised, open-label, multicentre, phase 3 trial. The Lancet. 2012 Feb;379(9816):633–40. Resende U, Cabello C, Oliveira Botelho Ramalho S, Zeferino LC. Predictors of Pathological Complete Response in Women with Clinical Complete Response to Neoadjuvant Chemotherapy in Breast Carcinoma. Oncology. 2018;95(4):229–38. Kahles A, Lehmann KV, Toussaint NC, Hüser M, Stark SG, Sachsenberg T, et al. Comprehensive Analysis of Alternative Splicing Across Tumors from 8,705 Patients. Cancer Cell. 2018 Aug;34(2):211-224.e6. Loi S, Drubay D, Adams S, Pruneri G, Francis PA, Lacroix-Triki M, et al. Tumor-Infiltrating Lymphocytes and Prognosis: A Pooled Individual Patient Analysis of Early-Stage Triple-Negative Breast Cancers. J Clin Oncol. 2019 Mar 1;37(7):559–69. Liao G, Jiang Z, Yang Y, Zhang C, Jiang M, Zhu J, et al. Combined homologous recombination repair deficiency and immune activation analysis for predicting intensified responses of anthracycline, cyclophosphamide and taxane chemotherapy in triple-negative breast cancer. BMC Med. 2021 Dec;19(1):190. Oshi M, Patel A, Wu R, Le L, Tokumaru Y, Yamada A, et al. Enhanced immune response outperform aggressive cancer biology and is associated with better survival in triple-negative breast cancer. Npj Breast Cancer. 2022 Aug 9;8(1):92. Chun J, Buechelmaier ES, Powell SN. Rad51 Paralog Complexes BCDX2 and CX3 Act at Different Stages in the BRCA1-BRCA2-Dependent Homologous Recombination Pathway. Mol Cell Biol. 2013 Jan 1;33(2):387–95. Liu Y, Masson JY, Shah R, O’Regan P, West SC. RAD51C Is Required for Holliday Junction Processing in Mammalian Cells. Science. 2004 Jan 9;303(5655):243–6. Rodrigue A, Coulombe Y, Jacquet K, Gagné JP, Roques C, Gobeil S, et al. The RAD51 paralogs ensure cellular protection against mitotic defects and aneuploidy. J Cell Sci. 2013 Jan 1;126(1):348–59. Sage JM, Gildemeister OS, Knight KL. Discovery of a Novel Function for Human Rad51. J Biol Chem. 2010 Jun;285(25):18984–90. SU CH, CHANG WS, HU PS, HSIAO CL, JI HX, LIAO CH, et al. Contribution of DNA Double-strand Break Repair Gene XRCC3 Genotypes to Triple-negative Breast Cancer Risk. Cancer Genomics - Proteomics. 2015 Nov 1;12(6):359. Xinyi X, Gong Y. The role of ATP‐binding cassette subfamily G member 1 in tumor progression. Cancer Med. 2024 Jun;13(12):e7285. Wang B, Xing A, Li G, Liu L, Xing C. SNHG14 promotes triple‐negative breast cancer cell proliferation, invasion, and chemoresistance by regulating the ERK / MAPK signaling pathway. IUBMB Life. 2024 Dec;76(12):1295–308. Liu B, Lu T, Wang Y, Zhang G, Fu L, Yu M, et al. Overexpression of LncRNA SNHG14 as a biomarker of clinicopathological and prognosis value in human cancers: A meta-analysis and bioinformatics analysis. Front Genet. 2022 Oct 6;13:945919. Hua H, Xie H, Zheng J, Lei L, Deng Z, Yu C. RNF7 Facilitated the Tumorigenesis of Pancreatic Cancer by Activating PI3K/Akt Signaling Pathway. Adnan M, editor. Oxid Med Cell Longev. 2023 Jan 4;2023:1–17. Xiao Y, Jiang Y, Song H, Liang T, Li Y, Yan D, et al. RNF7 knockdown inhibits prostate cancer tumorigenesis by inactivation of ERK1/2 pathway. Sci Rep. 2017 Mar 2;7(1):43683. Xiao C, Zhang W, Hua M, Chen H, Yang B, Wang Y, et al. RNF7 inhibits apoptosis and sunitinib sensitivity and promotes glycolysis in renal cell carcinoma via the SOCS1/JAK/STAT3 feedback loop. Cell Mol Biol Lett. 2022 Dec;27(1):36. Braschi B, Bruford EA, Cavanagh AT, Neuman SD, Bashirullah A. The bridge-like lipid transfer protein (BLTP) gene group: introducing new nomenclature based on structural homology indicating shared function. Hum Genomics. 2022 Dec 2;16(1):66. Wang Z, Wang Y, Li Z, Xue W, Hu S, Kong X. Lipid metabolism as a target for cancer drug resistance: progress and prospects. Front Pharmacol. 2023 Sep 28;14:1274335. Choi MS, Graves MJ, Matoo S, Storad ZA, El Sheikh Idris RA, Weck ML, et al. The small EF-hand protein CALML4 functions as a critical myosin light chain within the intermicrovillar adhesion complex. J Biol Chem. 2020 Jul;295(28):9281–96. Guo X, Xia S, Ge T, Lin Y, Hu S, Wu H, et al. Atp13a5 Marker Reveals Pericyte Specification in the Mouse Central Nervous System. J Neurosci. 2024 Oct 23;44(43):e0727242024. He W, Sun Y, Ge J, Wang X, Lin B, Yu S, et al. STRA6 regulates tumor immune microenvironment and is a prognostic marker in BRAF-mutant papillary thyroid carcinoma. Front Endocrinol. 2023 Feb 10;14:1076640. Moore CB, Bergstralh DT, Duncan JA, Lei Y, Morrison TE, Zimmermann AG, et al. NLRX1 is a regulator of mitochondrial antiviral immunity. Nature. 2008 Jan;451(7178):573–7. Porté S, Valencia E, Yakovtseva EA, Borràs E, Shafqat N, Debreczeny JÉ, et al. Three-dimensional Structure and Enzymatic Function of Proapoptotic Human p53-inducible Quinone Oxidoreductase PIG3. J Biol Chem. 2009 Jun;284(25):17194–205. Singh K, Roy M, Prajapati P, Lipatova A, Sripada L, Gohel D, et al. NLRX1 regulates TNF-α-induced mitochondria-lysosomal crosstalk to maintain the invasive and metastatic potential of breast cancer cells. Biochim Biophys Acta BBA - Mol Basis Dis. 2019 Jun;1865(6):1460–76. Wang ET, Sandberg R, Luo S, Khrebtukova I, Zhang L, Mayr C, et al. Alternative isoform regulation in human tissue transcriptomes. Nature. 2008 Nov;456(7221):470–6. Lv X, Sun X, Gao Y, Song X, Hu X, Gong L, et al. Targeting RNA splicing modulation: new perspectives for anticancer strategy? J Exp Clin Cancer Res. 2025 Jan 30;44(1):32. Additional Declarations No competing interests reported. Supplementary Files SupplementaryTables.xlsx Supplementaryfigure1.pdf Supplementary Figure 1 Classification and functional consequences of isoform switching events. A) Distribution of significant isoform switching events (q-value < 0.05) categorized by splicing type: alternative 3'/5' splice sites (A3, A5), alternative transcription start/termination sites (ATSS, ATTS), exon skipping (ES), intron retention (IR), and mutually exclusive exons (MEE, MES). Results are shown for each pairwise comparison between response groups: no vs partial, complete vs partial, and no vs complete. B) Predicted functional consequences of isoform switching for the upregulated isoform, including domain gain/loss/switch, changes in ORF structure (gain/loss, length), intron retention status, subcellular localization, and sensitivity to nonsense-mediated decay (NMD). Supplementaryfigure2.pdf Supplementary Figure 2 Gene ontology and pathway enrichment for isoform switching: no vs complete response. Supplementaryfigure3.pdf Supplementary Figure 3 Gene ontology and pathway enrichment for isoform switching: no vs partial response. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7041176","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":486402885,"identity":"27c49257-c143-467b-a18f-07b9f36308d3","order_by":0,"name":"Katarzyna Nowis","email":"data:image/png;base64,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","orcid":"","institution":"Ludwik Hirszfeld Institute of Immunology and Experimental Therapy","correspondingAuthor":true,"prefix":"","firstName":"Katarzyna","middleName":"","lastName":"Nowis","suffix":""},{"id":486402887,"identity":"2dee1aa9-3ea9-4d08-b76e-f1f2f8f68f4f","order_by":1,"name":"Maria Sąsiadek","email":"","orcid":"","institution":"Wroclaw Medical University","correspondingAuthor":false,"prefix":"","firstName":"Maria","middleName":"","lastName":"Sąsiadek","suffix":""},{"id":486402889,"identity":"28f14883-0d31-4e08-8571-04ffd7df1e4b","order_by":2,"name":"Dariusz Martynowski","email":"","orcid":"","institution":"Ludwik Hirszfeld Institute of Immunology and Experimental Therapy","correspondingAuthor":false,"prefix":"","firstName":"Dariusz","middleName":"","lastName":"Martynowski","suffix":""},{"id":486402890,"identity":"e28b9ab7-55ec-4170-b0b8-1bfbe06f840e","order_by":3,"name":"Dorota Kujawa","email":"","orcid":"","institution":"Ludwik Hirszfeld Institute of Immunology and Experimental Therapy","correspondingAuthor":false,"prefix":"","firstName":"Dorota","middleName":"","lastName":"Kujawa","suffix":""},{"id":486402892,"identity":"60135ac9-7248-45f4-b8ce-12e0794cf13f","order_by":4,"name":"Izabela Laczmanska","email":"","orcid":"","institution":"Lower Silesian Center of Oncology, Pulmonology and Hematology","correspondingAuthor":false,"prefix":"","firstName":"Izabela","middleName":"","lastName":"Laczmanska","suffix":""},{"id":486402894,"identity":"103eac56-5522-4493-8858-a91e0d35937a","order_by":5,"name":"Paweł Karpiński","email":"","orcid":"","institution":"Wroclaw Medical University","correspondingAuthor":false,"prefix":"","firstName":"Paweł","middleName":"","lastName":"Karpiński","suffix":""},{"id":486402896,"identity":"3261527b-d8e1-4758-95d6-868175795829","order_by":6,"name":"Marcin Ekiert","email":"","orcid":"","institution":"Lower Silesian Center of Oncology, Pulmonology and Hematology","correspondingAuthor":false,"prefix":"","firstName":"Marcin","middleName":"","lastName":"Ekiert","suffix":""},{"id":486402897,"identity":"3756c22b-f98d-454a-bb67-0678b5d187de","order_by":7,"name":"Ewelina Iwaneczko","email":"","orcid":"","institution":"Lower Silesian Center of Oncology, Pulmonology and Hematology","correspondingAuthor":false,"prefix":"","firstName":"Ewelina","middleName":"","lastName":"Iwaneczko","suffix":""},{"id":486402899,"identity":"8f7d01b4-642c-451c-a59a-616d5a491006","order_by":8,"name":"Bartłomiej Szynglarewicz","email":"","orcid":"","institution":"Lower Silesian Center of Oncology, Pulmonology and Hematology","correspondingAuthor":false,"prefix":"","firstName":"Bartłomiej","middleName":"","lastName":"Szynglarewicz","suffix":""},{"id":486402900,"identity":"42654873-84f3-4226-8c30-335a99eb046e","order_by":9,"name":"Piotr Kasprzak","email":"","orcid":"","institution":"Lower Silesian Center of Oncology, Pulmonology and Hematology","correspondingAuthor":false,"prefix":"","firstName":"Piotr","middleName":"","lastName":"Kasprzak","suffix":""},{"id":486402901,"identity":"992b8c0d-3cd0-4ff2-9d24-a9a3a77e7ad7","order_by":10,"name":"Mariola Abrahamowska","email":"","orcid":"","institution":"Lower Silesian Center of Oncology, Pulmonology and Hematology","correspondingAuthor":false,"prefix":"","firstName":"Mariola","middleName":"","lastName":"Abrahamowska","suffix":""},{"id":486402903,"identity":"fd2b0a75-7e6c-4201-a97d-fb50f9909e27","order_by":11,"name":"Rafał Matkowski","email":"","orcid":"","institution":"Lower Silesian Center of Oncology, Pulmonology and Hematology","correspondingAuthor":false,"prefix":"","firstName":"Rafał","middleName":"","lastName":"Matkowski","suffix":""},{"id":486402904,"identity":"35abe256-2787-419e-be8d-0766fa5ac872","order_by":12,"name":"Lukasz Laczmanski","email":"","orcid":"","institution":"Ludwik Hirszfeld Institute of Immunology and Experimental Therapy","correspondingAuthor":false,"prefix":"","firstName":"Lukasz","middleName":"","lastName":"Laczmanski","suffix":""}],"badges":[],"createdAt":"2025-07-03 20:08:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7041176/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7041176/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":87380253,"identity":"2a4361a5-ab48-493f-ac05-3beaea2b1adf","added_by":"auto","created_at":"2025-07-23 08:33:30","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":126337,"visible":true,"origin":"","legend":"\u003cp\u003eIsoform switching events and differentially expressed isoforms (dIFs) in relation to therapy response. A: Volcano plots display the distribution of isoform switching events identified in two response groups: no response, partial response and complete response. Each point represents an isoform, with red indicating significant switching events (FDR \u0026lt; 0.05). The x-axis shows the difference in isoform fraction (dIF), and the y-axis shows the negative log10-transformed p-value. B: MA-like plots showing the relationship between gene expression (x-axis: log2 fold change) and isoform switching (y-axis: dIF) for the same response groups. Red points represent isoforms with significant switching events. Most switching events occur within genes with moderate expression changes, though some highly differentially expressed genes also show isoform switching. C: Venn diagrams illustrate the overlap of isoforms undergoing switching between response comparisons (no response vs complete response and complete vs partial response). The left diagram shows overlap between response comparisons in context of switches, and the right in context of switching genes.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/afe256114ec8446802dfadc5.png"},{"id":87384416,"identity":"d1e3ad87-aef8-4fc9-aceb-9ae75fb91f05","added_by":"auto","created_at":"2025-07-23 08:49:30","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":204708,"visible":true,"origin":"","legend":"\u003cp\u003eAlternative splicing (AS) events observed during isoform switching across therapy response groups. A: The dot plot depicts the fraction of genes that primarily undergo specific AS events during isoform switching, with 95% confidence intervals. Results are shown for two comparisons: no vs complete and no vs complete response. Statistically significant differences (FDR \u0026lt; 0.05) are highlighted in red B: Differences in ATSS gain (paired with ATSS loss), ATTTS gain (paired with ATTTS loss), and IR gain (paired with IR loss) between the two primary comparisons: no vs complete and no vs partial response, contrasts the individual condition comparisons in a pairwise manner. This allows for the assessment of whether the ratio of gains is significantly different across the various comparisons, thus providing insight into splicing variations associated with each condition. Each dot represents the fraction of genes undergoing the respective AS event, with point size proportional to the number of genes affected. Red dots indicate statistically significant differences (FDR \u0026lt; 0.05). The impact of alternative splicing on protein function, transcript composition, and subcellular localization across therapy response groups C: The proportion of genes with functional consequences due to isoform switching is compared between response groups. Statistically significant differences (FDR \u0026lt; 0.05, shown in red). D: The impact of alternative splicing on subcellular localization is visualized as a congruence matrix. Gains and losses of localization signals (e.g., nucleus, mitochondrion, cytoplasm) are compared between no vs partial and complete response.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/08bee45edd149b326805ea7a.png"},{"id":87383188,"identity":"c5b8b69e-2ea1-4299-8f3f-fd004a14b816","added_by":"auto","created_at":"2025-07-23 08:41:30","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":209471,"visible":true,"origin":"","legend":"\u003cp\u003eVisualization of the top three isoform switching events based on adjusted p-value in triple-negative breast cancer response groups. A) XRCC3 (no response vs complete response), B) ABCG1 (no response vs complete response), C) SNHG14 (no response vs complete response), D) ABCG1 (no response vs partial response), E) RNF7 (no response vs partial response), F) BLTP3B (no response vs partial response),\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/534b0572e98970abe38a234f.png"},{"id":87383191,"identity":"f2e615fa-37c0-414f-9d11-ff7db065adc6","added_by":"auto","created_at":"2025-07-23 08:41:30","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":271865,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eCartoon representation of: \u003c/strong\u003e\u003c/em\u003e\u003cem\u003eA) Superposition of hRAD51 (PDB: 7EJC) onto the apRAD51C/XRCC3 complex (PDB: 8GJA), B) Structural alignment of the AlphaFold3 model of the human RAD51/XRCC3/RAD51C complex onto the hRAD51 structure (PDB: 7EJC), C) The XRCC3 subunit in complex b) was replaced with a shorter isoform.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/3f85a3df2be124d3d197bc91.png"},{"id":87380261,"identity":"455a2f1c-3d80-4ab2-9224-d71a25092cd6","added_by":"auto","created_at":"2025-07-23 08:33:30","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":182705,"visible":true,"origin":"","legend":"\u003cp\u003eFunctional enrichment and pathway network analysis of alternatively spliced genes in complete versus no response groups.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/67a8a9256fae92b091504b1c.png"},{"id":87383190,"identity":"1a3ef39d-8932-492d-8bf7-d32286c53bd7","added_by":"auto","created_at":"2025-07-23 08:41:30","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":100383,"visible":true,"origin":"","legend":"\u003cp\u003eFunctional enrichment and pathway network analysis of alternatively spliced genes in partial versus no response groups.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/11ace83eddb7af1c58ef9718.png"},{"id":92439472,"identity":"72a5984d-bade-4581-bec9-619336a4197c","added_by":"auto","created_at":"2025-09-29 18:01:38","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1966927,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/82c1b92a-7f67-4bc8-8e71-251f2396ba92.pdf"},{"id":87380258,"identity":"17293c24-7070-4892-a509-73911586bf77","added_by":"auto","created_at":"2025-07-23 08:33:30","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":234794,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/0d469eeb9c6efece25641fa6.xlsx"},{"id":87380251,"identity":"fa2f4244-3d8e-49a8-83ba-7816ede9067b","added_by":"auto","created_at":"2025-07-23 08:33:30","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":378726,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Figure 1 Classification and functional consequences of isoform switching events. A) Distribution of significant isoform switching events (q-value \u0026lt; 0.05) categorized by splicing type: alternative 3'/5' splice sites (A3, A5), alternative transcription start/termination sites (ATSS, ATTS), exon skipping (ES), intron retention (IR), and mutually exclusive exons (MEE, MES). Results are shown for each pairwise comparison between response groups: no vs partial, complete vs partial, and no vs complete. B) Predicted functional consequences of isoform switching for the upregulated isoform, including domain gain/loss/switch, changes in ORF structure (gain/loss, length), intron retention status, subcellular localization, and sensitivity to nonsense-mediated decay (NMD).\u003c/p\u003e","description":"","filename":"Supplementaryfigure1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/7d13162e7e92c3aaf3c5aec0.pdf"},{"id":87380262,"identity":"1d1b390c-905b-43e4-888d-89e805ba01fa","added_by":"auto","created_at":"2025-07-23 08:33:30","extension":"pdf","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":582997,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eSupplementary Figure 2\u003c/em\u003e Gene ontology and pathway enrichment for isoform switching: no vs complete response.\u003c/p\u003e","description":"","filename":"Supplementaryfigure2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/2471ba0292783d9763a68790.pdf"},{"id":87380255,"identity":"c60356b4-0bdc-4a05-aac1-3ab7355dd4a5","added_by":"auto","created_at":"2025-07-23 08:33:30","extension":"pdf","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":97380,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eSupplementary Figure 3 \u003c/em\u003eGene ontology and pathway enrichment for isoform switching: no vs partial response.\u003c/p\u003e","description":"","filename":"Supplementaryfigure3.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7041176/v1/65b775af1506aecbe4524fae.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Isoform switching as a key mechanism in chemotherapy resistance in triple-negative breast cancer","fulltext":[{"header":"STATEMENT OF TRANSLATIONAL RELEVANCE","content":"\u003cp\u003eTripple-negative breast cancer (TNBC) remains a therapeutically challenging subtype due to the lack of targeted treatment options and high inter-patient variability in chemotherapy response. In this study, we demonstrate that alternative splicing and isoform switching contribute to differential therapy responses in TNBC. Notably, we identified enrichment of a known, truncated \u003cem\u003eXRCC3\u0026nbsp;\u003c/em\u003eisoform (ENST00000554974) in non-responders. This isoform lacks critical domains required for interaction with RAD51 and RAD51C, likely impairing homologous recombination repair. Importantly, total \u003cem\u003eXRCC3\u003c/em\u003e gene expression did not differ between groups, underscoring the need of isoform-level resolution.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOur findings suggest that transcript-based profiling of \u003cem\u003eXRCC3\u0026nbsp;\u003c/em\u003eisoforms could help identify TNBC patients with splicing-driven HRR deficiency, even in the absence of BRCA mutations. This opens the possibility of expanding synthetic lethality-based therapeutic strategies, such as the use of PARPi or ATR/CHK1i, to BRCA wild-type patient exhibiting \u003cem\u003eXRCC3\u0026nbsp;\u003c/em\u003eisoform imbalance. Integrating isoform-level diagnostics may improve patient stratification and guide personalized treatment selection in TNBC.\u0026nbsp;\u003c/p\u003e"},{"header":"INTRODUCTION","content":"\u003cp\u003eTriple-negative breast cancer (TNBC), characterized by the absence of estrogen(ER), progesterone receptors (PR), and human epidermal growth factor receptor 2 (ERBB2/HER2) expression, accounts for approximately 15\u0026ndash;20% of all breast cancers (BC) (\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). TNBC is known for its aggressive clinical behavior, higher rates of recurrence, and limited treatment options due to the lack of targeted therapies (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). Neoadjuvant chemotherapy (NAC) is the standard systemic treatment for TNBC, aiming to reduce tumor size, eliminate lymph node metastases and micrometastatic disease before surgery (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e). However, response to NAC varies markedly among patients, with some achieving a pathologic complete response (pCR) while others exhibit residual disease, which is associated with a worse prognosis (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). Emerging evidence suggests that alternative splicing may also contribute to inter-patient variability in therapeutic response, including sensitivity or resistance to NAC in TNBC (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAlternative splicing (AS) is a fundamental post-transcriptional mechanism that enables a single gene to produce multiple mRNA isoforms, thereby contributing to proteomic diversity. Dysregulation of AS has been implicated in various cancers, where it can lead to the production of isoforms that enhance proliferation, metastasis and resistance to therapy (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe regulation of AS is primarily governed by a complex network of splicing factors, whose dysregulation can profoundly affect transcript diversity in cancer (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e) making them potential therapeutic targets. Thus, some of public attention is focused on expression, mutation and even AS of those factors. For instance, pan-cancer studies revealed 167 cancer specific splicing patterns in. splicing factors with consequences across 16 cancer types and 6904 patients (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e). Moreover, splicing factors SRSF1 and SRSF3, has been linked in oncogenesis and poor prognosis (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). In turn, in TNBC CPSF4 has been shown to regulate AS of \u003cem\u003eHMG20B\u003c/em\u003e promoting proliferation and migration (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eBeyond their impact on tumor biology, AS alterations have also emerged as important determinants of therapeutic response. Increasing evidence suggests that distinct splicing profiles can mediate sensitivity or resistance to chemotherapy, radiotherapy, targeted therapies, or immunotherapy, further underscoring functional significance of splicing regulation in cancer. For instance, SRSF1 improves drug resistance through alternative splicing in MYO1B in BC (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e), \u003cem\u003eBRCA1\u003c/em\u003e exon skipping resulting in ineffective PARPi (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e), or the failure of CAR-T therapy in leukemias due to retention of CD19 intron (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e). These findings highlight the potential of AS events not only as biomarkers of treatment but also targets for overcoming resistant mechanisms.\u003c/p\u003e\u003cp\u003eIncreasing attention has also been directed to the role of AS in TNBC. Recent studies have demonstrated that transcriptomic reprogramming through AS can drive apoptosis, modulate key oncogenic regulators such as MDM4, or influence prognosis and treatment vulnerabilities through aberrant splicing events (\u003cspan additionalcitationids=\"CR18 CR19 CR20\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e). However, these studies primarily focused on individual genes, specific splicing factors, or targeted modulation., leaving a broader understanding of splicing alterations in clinical context unexplored.\u003c/p\u003e\u003cp\u003eTo address this gap, we performer a comprehensive, transcriptome-wide analysis of differentia isoform usage in TNBC patients stratified by NAC response. By integrating differential isoform expression with functional enrichment and pathway analyses, we aimed to identify splicing-regulated biological processes associated with treatment outcomes. Our findings provide novel insights into the molecular mechanisms underlying chemotherapy response and resistance, highlighting alternative splicing as a potential source of predictive biomarkers and therapeutic targets in TNBC.\u003c/p\u003e"},{"header":"MATERIALS AND METHOD","content":"\u003cp\u003e\u003cb\u003eMaterial\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThis study included 50 biopsy samples of breast tumor tissue taken before preoperative chemotherapy from female patients with triple-negative breast cancer (TNBC) diagnosed and treated at the Lower Silesian Oncology, Pulmonology and Hematology Center in Wrocław. Systemic preoperative treatment was based on standard chemotherapy. Immunotherapy was not used in the study group. Response to preoperative systemic treatment was assessed on the basis of postoperative histopathological examination. Detailed information on sample acquisition, patient selection criteria, ethical approvals, and associated clinical data is provided in a previous publication (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cb\u003eData processing\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe first step was to assess the quality of the raw sequencing data using PRINSEQ-lite v0.20.4 (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e)(RRID:SCR_005454). Quality metrics including base quality scores, sequence complexity and GC content are evaluated to identify low quality reads and overrepresented sequences.\u003c/p\u003e\u003cp\u003eAfter the initial quality assessment, read sorting was performed by reordering reads based on their identifiers. Sorting ensures that paired-end reads were consistently aligned during downstream processing. Then the AdapterRemoval v2.3.1 tool was used to remove adapter sequences and low-quality bases (below a score of 30) and to discard fragments shorter than 20 nucleotides (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e)(RRID:SCR_011834). A second round of quality assessment was performed using PRINSEQ-lite v0.20.4 to confirm the effectiveness of the trimming and filtering steps (Supplementary Table\u0026nbsp;1).\u003c/p\u003e\u003cp\u003eHISAT2 v2.2.1 was used to map the processed reads to GRCh38 reference genome index, sourced from the HISAT2 website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://daehwankimlab.github.io/hisat2/download/\u003c/span\u003e\u003cspan address=\"http://daehwankimlab.github.io/hisat2/download/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, grch38_tran.tar.gz)(RRID:SCR_015530)(\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e). This splice-aware aligner efficiently handled large and complex transcriptomes while supporting accurate identification of intronic and splicing regions. The aligned reads were output in BAM format for downstream transcript assembly.\u003c/p\u003e\u003cp\u003eThe aligned reads were processed using StringTie v2.1.7 for transcript assembly and quantification (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e)(RRID:SCR_016323). StringTie reconstructs full-length transcripts by leveraging splice junction information and provides accurate estimations of transcript expression levels. The generated GTF files define the structure and abundance of transcripts for each sample, enabling comprehensive transcriptome profiling.\u003c/p\u003e\u003cp\u003e\u003cb\u003eDifferential gene expression analysis, isoform switching and functional analysis\u003c/b\u003e\u003c/p\u003e\u003cp\u003eDifferential gene expression was performed using DESeq2 and SARTools R package (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e)(RRID:SCR_00154, SCR_016533). Batch correction was made with ComBat (RRID:SCR_010974) Benjamini and Hochberg padj values for gene expression was used for visualization on isoform switch plots with IsoformSwitchAnalyzeR(\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe final stage of the pipeline involved analyzing isoform switching using IsoformSwitchAnalyzeR. GTF files generated by StringTie and the associated expression matrices were used as input to identify differential isoform expression across experimental conditions. IsoformSwitchAnalyzeR evaluates the functional consequences of isoform switching, including domain analysis and coding potential assessment, offering insights into potential biological impacts.\u003c/p\u003e\u003cp\u003eFunctional consequences of isoform switching analyzed using IsoformSwitchAnalyzeR was performed based on outputs of several external tools were integrated into the as follows: CPC2, signalP6, DeepLoc2, DeepTMHMM to asses coding potential, identifying functional elements like signal peptides and subcellular localizations, and analyzing transmembrane regions visualized on isoform switch plots (\u003cspan additionalcitationids=\"CR31 CR32\" citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cb\u003eAlternative splicing-associated functional enrichment and network analysis\u003c/b\u003e\u003c/p\u003e\u003cp\u003eFunctional enrichment analysis was performed using the STRING database (version 12.0). Protein\u0026ndash;protein interaction (PPI) networks were generated in STRING with a minimum required interaction score of 0.7 (high confidence) (RRID:SCR_005223). Enrichment analysis was conducted using STRING\u0026rsquo;s built-in functional enrichment test. Significance was assessed using the false discovery rate (FDR) correction, with an FDR threshold of \u0026lt;\u0026thinsp;0.05.\u003c/p\u003e\u003cp\u003eThe resulting enrichment data were visualized using the EnrichmentMap plugin (RRID:SCR_016052). Each node in the enrichment map represents an enriched term, and edges reflect gene overlap between terms. Node color corresponds to the significant normalized enrichment score (NES), indicating the direction and strength of enrichment in the comparison (positive NES for enrichment in responders, negative NES for enrichment in non-responders). Clusters of related terms were annotated using the AutoAnnotate and WordCloud plugins to identify key biological themes enriched in different therapy response groups. All analysis were performed using Cytoscape (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cb\u003eStructural Modeling and Superposition of XRCC3\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe crystal structure of human RAD51 in complex with single-stranded DNA (ssDNA) and ATP (PDB ID: 7EJC) was used as a reference template for structural superposition. The \u003cem\u003eAlvinella pompejana\u003c/em\u003e RAD51/XRCC3 complex (PDB ID: 8GJA) was aligned onto this reference using UCSF ChimeraX (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e) (RRID:SCR_015872).\u003c/p\u003e\u003cp\u003eTo model the human RAD51/XRCC3/RAD51C complex, AlphaFold3 predictions were generated using the AlphaFold Server (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://alphafoldserver.com\u003c/span\u003e\u003cspan address=\"https://alphafoldserver.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e).\u003c/span\u003e The model included three bound ATP molecules and was prepared in two variants: one containing the canonical \u003cem\u003eXRCC3\u003c/em\u003e isoform (ENST00000553264), and the other containing a shorter RAD51 domain (ENST00000554974) identified in isoform switching analysis. Both AlphaFold3-derived complexes were subsequently superimposed onto the hRAD51/apRAD51C/XRCC3 complex using structural alignment tools within ChimeraX. All structural visualizations and superpositions were performed in UCSF ChimeraX (version 1.9) (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e).\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cp\u003e\u003cb\u003eAn increase in the number of isoforms switching events has been observed in non-responders.\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe analysis revealed that the number of isoforms switching events varies depending on the response to treatment used. Volcano plots (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA) show a higher density of significantly switched isoforms in the complete response group compared to the no response group than in the partial response group compared to the no response group. The smallest number of isoform switching was observed when comparing total response to therapy with partial response (Supplementary Table\u0026nbsp;2). This suggests that non-responders exhibit the most extensive changes in isoform usage, potentially reflecting an adaptive mechanism or dysregulation associated with resistance to therapy compared to partial- and complete-responders.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe MA-like plots (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB) illustrate that isoform switching occurs independently of large gene expression changes. Many significantly switched isoforms are observed across a wide range of gene expression levels, highlighting that isoform-level regulation can act independently of transcriptional activation. Moreover, the largest number of significant isoform switching events is observed in the comparisons involving non-responders reinforcing the notion that non-responders experience more pronounced isoform reprogramming.\u003c/p\u003e\u003cp\u003eVenn diagrams (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC, D, E) further highlight differences between studied groups. The highest number of switching events is observed in comparison complete \u003cem\u003evs\u003c/em\u003e no response (1138) following with partial \u003cem\u003evs\u003c/em\u003e no response (724) and complete \u003cem\u003evs\u003c/em\u003e partial response (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e). The greatest overlap of switched isoforms is found between no response \u003cem\u003eversus\u003c/em\u003e complete response and no response \u003cem\u003eversus\u003c/em\u003e partial response (401), indicating that isoform switching events distinguish responders from non-responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD). Similar observation is made for differentially used isoforms themselves (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC) (927, 581, 58 and 417 respectively). When considering genes containing differentially used isoforms (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE) and the total number of affected genes, an analogous trend is observed. The highest number of genes with differentially used isoforms is found in the comparison between complete and no response (817), followed by partial \u003cem\u003eversus\u003c/em\u003e no response (524), and complete versus partial response (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e). Notably, the largest overlap (390 genes) occurs between the no response \u003cem\u003eversus\u003c/em\u003e complete response and no response \u003cem\u003eversus\u003c/em\u003e partial response groups, further supporting the concept that isoform usage patterns distinguish responders from non-responders.\u003c/p\u003e\u003cp\u003eDifferential isoform usage was observed in only five genes consistently across all three comparisons, suggesting a limited but potentially crucial set of genes involved in therapy response modulation (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE). These results underscore the importance of alternative isoform regulation in determining treatment outcomes and highlight key molecular players that may serve as potential biomarkers or therapeutic targets. All these findings demonstrate that the extent of isoform switching increases as the response to therapy diminishes.\u003c/p\u003e\u003cp\u003e\u003cb\u003eSignificant changes in alternative transcription start/termination sites and intron retention events are associated with poor therapy response\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe comparison of isoform switching events across therapy response groups reveals distinct patterns of alternative splicing associated with differences in therapy outcomes: nonresponders exhibit a larger frequency of alternative transcription start site (ATSS), alternative transcription termination site (ATTTS) and intron retention (IR) events, suggesting that transcriptional regulation at the initiation and termination levels may play a significant role in therapy resistance. This is supported by the significant enrichment of ATSS, ATTTS, IR in no- \u003cem\u003evs\u003c/em\u003e complete responders and IR in no- vs partial responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, Supplementary Fig.\u0026nbsp;1A). Furthermore, the analysis shows that these transcriptional events are more pronounced in the non- versus complete responders than in non- versus partial responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, Supplementary Fig.\u0026nbsp;1A), suggesting that the extent of transcriptional reprogramming correlates with the level of therapeutic resistance.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eImportantly, comparison of the relative contribution of splicing event categories across response-group contrasts (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB) indicates that ATTTS gains represent the most prominent and discriminative event type, particularly in comparisons involving non-responders. While ATSS and IR events also occur, their differential burden is less pronounced, highlighting alternative transcription termination as a dominant mechanism underlying isoform reprogramming in therapy-resistant tumors.\u003c/p\u003e\u003cp\u003eOverall, these results highlight the importance of alternative transcription regulation in non-responders, with ATSS and especially ATTTS events emerging as key contributors to isoform reprogramming. The increased frequency and magnitude of these events in non-responders suggest that transcriptional plasticity may facilitate adaptive mechanisms that promote therapy resistance. Further investigation of genes undergoing ATSS and ATTTS switching could provide better insights into the molecular pathways that underpin poor therapeutic outcomes. Analysis of the transcriptome of tumor cells could be a prognostic marker for response to chemotherapy.\u003c/p\u003e\u003cp\u003e\u003cb\u003eAlternative splicing alters protein domains and coding potential\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo evaluate the impact of isoform switching on protein function, we analyzed the distribution of alternative splicing consequences across therapy response groups. Non-responders exhibited a significantly larger frequency of intron retention and ORF disruptions compared to both partial responders and complete responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC, Supplementary Fig.\u0026nbsp;1B). Complete responders also appeared to be enriched by domain alterations also observed in a larger proportion of transcripts from no response samples.\u003c/p\u003e\u003cp\u003eTo further evaluate the functional consequences of alternative splicing, we examined shifts in predicted subcellular localization between therapy response groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD). The congruence matrix highlights significant changes in localization patterns, with nuclear mRNA being particularly affected in no response samples. These disruptions may contribute to therapy resistance by altering the functional availability of key regulatory proteins involved in apoptosis, metabolism, and transcriptional control.\u003c/p\u003e\u003cp\u003eIn contrast, comparisons between complete and partial responders did not reveal statistically significant differences in splicing-related localization shifts, suggesting that splicing dysregulation is primarily linked to therapy non-response rather than partial or complete response.\u003c/p\u003e\u003cp\u003e\u003cb\u003eKey isoform switching events in treatment response.\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo identify key splicing-related changes in treatment resistance, we performed an isoform switching analysis comparing non-responders to partial and complete responders. The top three switching events were observed in \u003cem\u003eXRCC3, ABCG1, SNHG14\u003c/em\u003e, and \u003cem\u003eABCG1, RNF7, BLTP3B\u003c/em\u003e (for no versus complete and partial response respectively, padj\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;0.05). Among the five genes common to all comparisons in isoform switching genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE), isoform switching was also observed in \u003cem\u003eTP53I3, CALML4\u003c/em\u003e, and \u003cem\u003eSTRA6\u003c/em\u003e (padj\u0026thinsp;\u0026lt;\u0026thinsp;0.01), as well as \u003cem\u003eATP13A5\u003c/em\u003e and \u003cem\u003eNLRX1\u003c/em\u003e (padj\u0026thinsp;\u0026lt;\u0026thinsp;0.05)(Supplementary Table\u0026nbsp;2).\u003c/p\u003e\u003cp\u003e\u003cb\u003eXRCC3\u003c/b\u003e: \u003cb\u003enegative impact on BRCA-dependent homologous recombination DNA Repair Pathway in non-responders.\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cem\u003eXRCC3\u003c/em\u003e encodes a DNA repair protein involved in homologous recombination. Non-responders exhibited a significant shift in dominant isoforms compared to complete responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). The alternative splicing events resulted in the loss of 4 exonic regions from 5\u0026rsquo; end (part of Rad51 domain shortening protein from 346 to 141 aa), consistent with isoform switching from the canonical transcript ENST00000553264 to the truncated variant ENST00000554974 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA, \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eStructural studies have shown that the N-terminal domain (NTD) of XRCC3 forms a clamp-like interface with RAD51C, facilitating the formation of the CX3 complex, which plays a critical role in replication fork protection, restart, and RAD51 filament capping (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e). In the truncated isoform, the NTD is absent, leading to the loss of the structural interface with RAD51C and likely impairing proper assembly of CX3 complex. Furthermore, the missing N-terminal region also includes part of the RAD51-interacting domain and ATP binding motif, suggesting that both RAD51C and RAD51 interactions may be disrupted (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB, C). These defects could impair homologous recombination, limit the ability to stabilize or restart stalled replication forks, and promote genomic instability.\u003c/p\u003e\u003cp\u003e\u003cb\u003eABCG1\u003c/b\u003e: \u003cb\u003eLipid transport disruptions in non-responders.\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cem\u003eABCG1\u003c/em\u003e, a lipid transporter, showed distinct isoform switching in comparisons of both non- versus complete and non- versus partial responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB, D). Isoform transitions led to modifications at the beginning of the transcript, resulting in the loss of the signal domain in non-responders, which may lead to altered protein localization, impaired lipid transport efficiency, or disruptions in cellular signaling pathways associated with lipid metabolism and immune response. These changes could contribute to differences in therapeutic outcomes by affecting the functional role of ABCG1 in the tumor microenvironment.\u003c/p\u003e\u003cp\u003e\u003cb\u003eSNHG14\u003c/b\u003e, \u003cb\u003eThe Small Nucleolar Host Gene 14\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cem\u003eSNHG14\u003c/em\u003e, a long non-coding RNA, exhibited significant changes in isoform usage in non-responders comparing with complete responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). As lncRNAs have been implicated in transcriptional regulation and chromatin remodeling, these isoform shifts could play a role in modulating gene expression patterns associated with resistance mechanisms\u003c/p\u003e\u003cp\u003e\u003cb\u003eRNF7\u003c/b\u003e, \u003cb\u003eRing Finger Protein 7\u003c/b\u003e\u003c/p\u003e\u003cp\u003eAn ubiquitin ligase is associated with oxidative stress response and proteasomal degradation. Isoform switching in non-responders comparing with partial responders may suggests a potential dysregulation in protein homeostasis and stress response pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE).\u003c/p\u003e\u003cp\u003e\u003cb\u003eBLTP3B\u003c/b\u003e, \u003cb\u003eBridge-Like Lipid Transfer Protein Family Member 3B\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eBLTP3B is\u003c/b\u003e involved in lipid transport and membrane-associated processes. The observed isoform switching in non-responders comparing with partial responders may indicate functional alterations affecting lipid signaling and cellular metabolism, which could influence therapy response (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF).\u003c/p\u003e\u003cp\u003eOur findings highlight key isoform switching events that may contribute to treatment resistance by affecting DNA repair (\u003cem\u003eXRCC3\u003c/em\u003e), lipid metabolism (\u003cem\u003eABCG1, BLTP3B\u003c/em\u003e), proteasomal function (\u003cem\u003eRNF7\u003c/em\u003e), and transcriptional regulation (\u003cem\u003eSNHG14\u003c/em\u003e). Further functional validation is required to determine the precise impact of these isoform shifts on tumor biology and therapeutic response.\u003c/p\u003e\u003cp\u003e\u003cb\u003eFunctional enrichment and network analysis of alternatively spliced genes\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo investigate whether baseline alternative splicing profiles are associated with response to neoadjuvant chemotherapy (NAC) in TNBC, functional enrichment and protein-protein interaction (PPI) network analyses were performed.\u003c/p\u003e\u003cp\u003eIn the complete versus no response groups comparison several densely connected clusters were enriched, including regulation of cytokine-mediated immune responses, pattern recognition receptor signaling, and DNA damage repair pathways. Notably, pathways such as T cell receptor signaling, Fc gamma receptor activation, MAPK cascade initiation, and TLR4 signaling were significantly enriched in baseline tumors of future complete responders. Another large cluster revealed enrichment in DNA repair pathways including homologous recombination, base excision repair, and nucleotide biosynthesis, suggesting that tumors from future complete responders already express splicing isoforms conducive to efficient DNA damage detection and repair (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, Supplementary Table\u0026nbsp;3A, Supplementary Fig.\u0026nbsp;2).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn contrast, the partial versus no response groups comparison revealed more limited enrichment with only a few clusters reaching significance \u0026mdash; most notably DNA repair pathways and regulation of cellular response (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, Supplementary Table\u0026nbsp;3B, Supplementary Fig.\u0026nbsp;3). Immune signaling clusters were largely absent, suggesting that partial responders may share certain repair-related features with complete responders, but lack the immune-related transcriptional readiness that predicts robust therapy outcomes. This implies that cancer cells can avoid immune monitoring and anti-tumor immunity (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eDeveloping an effective anti-cancer therapy remains a significant medical challenge. Despite the increasing number of treatment strategies, a substantial proportion of patients still exhibit poor responses or complete resistance to standard approaches. Until today, exploring DNA-level markers of therapy failure has not revealed the causes underlying this. For instance, up to 40% of patients with TNBC fail to respond effectively to conventional chemotherapy regiments(\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e).\u003c/p\u003e\u003cp\u003ePrevious efforts to identify predictive biomarkers of therapy resistance have focused primarily on DNA-level alterations, such as mutations, copy number variations, and structural rearrangements. However, these approaches often failed to explain full spectrum of resistance mechanisms in clinic. This has led to a growing interest in exploring regulatory layers beyond DNA, particularly involving epigenetic and transcriptomic dysregulation.\u003c/p\u003e\u003cp\u003eNotably, epigenetic processes, including DNA methylation, histone modification, and RNA processing, have been shown to play critical roles in cancer development, progression, and therapy response. Among these, the dysregulation of alternative splicing has emerged as a significant contributor to tumor heterogeneity, immune evasion or drug resistance across multiple cancer types (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e). Importantly, many of these splicing alterations occur in the absence of changes at the genomic level, underscoring the functional relevance of transcriptomic profiling.\u003c/p\u003e\u003cp\u003eThis study demonstrates that alternative splicing measured through isoform switching is associated with differential responses to neoadjuvant chemotherapy (NAC) in triple-negative breast cancer (TNBC). Isoform switching refers to changes in the dominant transcript isoform of a gene between two conditions\u0026mdash;in this case, between tumor tissues from patients with complete response, partial response, or no response.\u003c/p\u003e\u003cp\u003eThe number and magnitude of isoform switching events varied depending on response level. The largest number of significant events was identified in the comparison between complete responders and non-responders, followed by partial versus non-responders with relatively few events observed between complete and partial responders. This suggests that non-responders exhibit the most pronounced isoform-level differences when compared to treatment-sensitive tumors. Importantly, these differences were detected in pre-treatment biopsy samples, supporting their potential value as predictive molecular features rather than consequences of chemotherapy exposure.\u003c/p\u003e\u003cp\u003eIn tumors derived from complete responders, isoform switching was associated with functional enrichment of immune-related signaling pathways. This included T cell receptor signaling, cytokine-mediated immune activation, and innate immune processes such as toll-like receptor 4 (TLR4) and MAPK cascades. Similar immune-related enrichment was not observed in partial responders. These findings support prior evidence that transcriptional immune activation and infiltration of immune cells are predictive of chemotherapy sensitivity in TNBC (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e). Since these differences were detected in pre-treatment biopsies, transcriptomic profiling using RNA sequencing may offer a valuable tool for patient stratification before therapy. In the future, RNA-seq-based assays, such as isoform-level biomarkers could support clinical decision-making and help identify patients most likely to benefit from NAC.\u003c/p\u003e\u003cp\u003ePartial responders, when compared to non-responders, showed enrichment of splicing alterations in genes associated with DNA repair mechanisms including homologous recombination and base excision repair. Although these processes were also observed in complete responders, the absence of immune pathway enrichment in partial responders may explain their intermediate outcome.\u003c/p\u003e\u003cp\u003eThis aligns with previous studies showing that while deficiency DNA repair pathways can enhance chemotherapy efficacy, the presence of a responsive immune environment is often required for complete tumor clearance (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e), we note that our analysis does not infer the directionality of isoform usage or functional activation.\u003c/p\u003e\u003cp\u003eTherefore, while alternative splicing appears to impact genes within these pathways, the consequences on pathway activity, whether gain or loss of pathway, remain to be elucidated. Nonetheless, the simultaneous disruption of DNA repair related transcripts and immune related-components in complete responders suggests the possibility of dual vulnerability, potentially contributing to synthetic lethality mechanisms.\u003c/p\u003e\u003cp\u003eSeveral genes with differential isoform usage contributed to the observed response patterns. Among the most prominent isoform switching events distinguishing non-responders from responders, \u003cem\u003eXRCC3\u003c/em\u003e emerged as a key gene of interest. \u003cem\u003eXRCC3\u003c/em\u003e encodes a protein critical for the homologous recombination repair (HRR) pathway, specifically in resolving DNA double-strand breaks (DSBs). It is a member of the RAD51 paralog family and functions as part of the CX3 complex, which acts downstream of RAD51 recruitment in the BRCA1-BRCA2-dependent HRR cascade. This complex binds to Holliday junctions and stalled replication forks, facilitating the resolution of HR intermediates during the mitotic cycle, often in coordination with GEN1 (\u003cspan additionalcitationids=\"CR46 CR47 CR48\" citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe alternative splicing events observed in non-responders led to the loss of 4 exonic regions from the 5\u0026prime; end of the transcript, potentially compromising domains critical for RAD51 and RAD51C interaction and complex formation. Structural modeling (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) indicates that this shortened isoform lacks key interaction surfaces necessary for proper CX3 complex assembly. Such a variant may exert a dominant-negative effect by competing with the full0length protein while failing to mediate functional interactions. Such structural alterations may impair the cell\u0026rsquo;s ability to resolve DNA damage effectively, leading to genomic instability \u0026mdash; an established contributor to chemoresistance. This observation contrasts with previous studies, which reported that XRCC3 overexpression enhances resistance to DNA-damaging agents such as temozolomide, cisplatin or radiotherapy, by promoting homologous recombination repair and limiting apoptosis\u003c/p\u003e\u003cp\u003eNotably, in our dataset, overall \u003cem\u003eXRCC3\u003c/em\u003e gene expression did not differ between groups, suggesting that gene-level quantification alone would have failed to detect any regulatory involvement of \u003cem\u003eXRCC3\u003c/em\u003e in therapy outcome. Only through isoform-resolved analysis was it possible to uncover splicing-dependent structural changes that may impair XRCC3 function. This highlights the functional relevance of AS as post-transcriptional regulatory mechanism that can modulate protein activity independently of gene expression level. Our results add a novel isoform-level dimension to this relationship, highlighting the potential of \u003cem\u003eXRCC3\u003c/em\u003e splicing as a predictive biomarker and functional contributor to therapy resistance in TNBC.\u003c/p\u003e\u003cp\u003e\u003cem\u003eABCG1\u003c/em\u003e, a gene involved in cholesterol and lipid transport, also exhibited distinct isoform switching in non-responders compared to both complete and partial responders. The altered transcript lacked key signal peptide regions, which may influence subcellular localization and impair ABCG1\u0026rsquo;s function in lipid regulation or immune signaling. Beyond its canonical role in intracellular cholesterol efflux, \u003cem\u003eABCG1\u003c/em\u003e has been shown to enhance tumor-promoting capacity by conferring stem like properties to cancer cells and mediating chemoresistance in multiple malignancies. It may also act as a kinase-like regulator by phosphorylating downstream targets that support tumor growth. Moreover, \u003cem\u003eABCG1\u003c/em\u003e modulates macrophage polarization in the tumor microenvironment, contributing to the formation of an immunosuppressive niche that favors cancer progression (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe long non-coding RNA \u003cem\u003eSNHG14\u003c/em\u003e was differentially spliced between response groups, particularly in non-responders. \u003cem\u003eSNHG14\u003c/em\u003e has been shown to regulate the ERK/MAPK signaling pathway and to promote proliferation and chemoresistance in TNBC models (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e). Interestingly, meta-analyses across various cancer types have demonstrated that elevated \u003cem\u003eSNHG14\u003c/em\u003e expression correlates with poor overall survival, larger tumor size, advanced stage, lymph node involvement and distant metastasis, supporting its role as a general marker of aggressive tumor behavior (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAdditional isoform switching was observed in genes such as \u003cem\u003eRNF7\u003c/em\u003e and \u003cem\u003eBLTP3B\u003c/em\u003e, implicated in proteostasis and lipid transport, respectively. These findings further support the idea that splicing alterations may affect multiple cellular systems critical to chemotherapy sensitivity. \u003cem\u003eRNF7 (SAG or ROC2)\u003c/em\u003e encodes a component of the SCF E3 ubiquintin ligase complex and has been implicated in tumor progression and therapy resistance across several cancer types, including prostate, pancreatic and renal carcinomas. Its pro-oncogenic functions have been linked to activation of the PI3K/Akt and JAK/STAT3 pathways, promotion of the cell cycle progression and glycolytic reprograming (\u003cspan additionalcitationids=\"CR54\" citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e). \u003cem\u003eBLTP3B\u003c/em\u003e is involved in intracellular lipid trafficking and endosome to Golgi transport, but its role in cancer is not yet defined (\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e). Nonetheless, perturbations in lipid homeostasis are increasingly recognized as drivers of tumor growth and drug resistance (\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn addition to major isoform switching events, we identified five genes\u0026mdash;\u003cem\u003eTP53I3\u003c/em\u003e, \u003cem\u003eCALML4\u003c/em\u003e, \u003cem\u003eSTRA6\u003c/em\u003e, \u003cem\u003eATP13A5\u003c/em\u003e, and \u003cem\u003eNLRX1\u003c/em\u003e\u0026mdash;that were consistently differentially spliced across all therapy response comparisons. These genes are implicated in apoptosis (\u003cem\u003eTP53I3\u003c/em\u003e), calcium signaling (\u003cem\u003eCALML4\u003c/em\u003e), immune modulation and vitamin A transport (\u003cem\u003eSTRA6\u003c/em\u003e), cation transporter (\u003cem\u003eATP13A5\u003c/em\u003e), and mitochondrial immune regulation (\u003cem\u003eNLRX1\u003c/em\u003e) (\u003cspan additionalcitationids=\"CR59 CR60 CR61 CR62\" citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTheir recurring presence suggests a potential core set of isoform-level regulators involved in therapy resistance. Notably, isoform switching in \u003cem\u003eNLRX1\u003c/em\u003e, a gene known to modulate oxidative stress and TNF-α signaling, may reflect mitochondrial reprogramming in non-responders. Similarly, changes in \u003cem\u003eSTRA6\u003c/em\u003e splicing may impact JAK/STAT signaling and retinoic acid metabolism, both previously linked to chemoresistance (\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThese isoform switching events were also associated with structural consequences at the protein level, including loss of coding potential, open reading frame (ORF) disruption, altered domain content, and changes in predicted subcellular localization. Such consequences may affect key aspects of protein function, including signaling, transport, or repair activity, in ways that support therapy resistance. Previous pan-cancer analyses have shown that alternative splicing can significantly impact tumor progression through these mechanisms (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). For example, mis-splicing can lead to the inclusion or exclusion of critical protein domains, frameshifts, or premature termination codons that alter protein stability and function (\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e). Importantly, aberrant RNA splicing is a molecular characteristic present in almost all types of tumors, contributing to tumorigenesis and progression. It plays role as potential target, shedding light on new perspectives for anti-cancer strategies (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eBecause this study focused exclusively on pre-treatment tumor samples, the observed isoform switching events may serve as predictive biomarkers for NAC response. Integrating isoform usage data with mutational profiles, immune infiltration measures, or clinical features may improve patient stratification and guide personalized treatment strategies in TNBC. Further validation and functional characterization of these splicing events will be necessary to confirm their relevance and mechanistic roles in therapy resistance.\u003c/p\u003e"},{"header":"CONSLUSION","content":"\u003cp\u003eThis study demonstrates that isoform switching is a prominent and functionally relevant mechanism associated with therapy response in triple-negative breast cancer. Transcriptome-wide analysis of pre-treatment samples revealed extensive isoform reprogramming in non-responders compared to partial and complete responders, implicating alternative splicing events in the development of chemoresistance. Key splicing alterations affected genes involved in DNA repair, lipid metabolism, immune signaling, and transcriptional regulation, with notable disruptions in protein domains, coding potential, and subcellular localization.\u003c/p\u003e\u003cp\u003eThese findings support the role of isoform-level regulation as a predictive molecular layer, distinct from gene expression alone, and highlight splicing-derived isoforms as potential biomarkers and therapeutic targets in TNBC. However, the assignment of oncogenic or tumor-suppressive roles of specific splicing events remains challenging. While some isoform shifts, such as the truncation of \u003cem\u003eXRCC3\u003c/em\u003e, allowed domain-level inference of functional loss, many events lacked sufficient structural or mechanistic annotation to determine their biological impact. Moreover, the volume and complexity of isoform-level data present substantial challenges for interpretation. Although specific functional categories emerged, the diversity of isoform changes across individual genes makes it difficult to draw overarching conclusions. This underscores the need for focused experimental follow-up, including domain mapping and functional assays to validate the biological relevance of splicing events in therapy response.\u003c/p\u003e\u003cp\u003eConclusively, these findings emphasize the pivotal role of AS in shaping the therapeutic response landscape in TNBC, reinforcing the necessity of an integrative and experimental approach to comprehensively explore and apply isoform-based biomarkers in clinical decision-making.\u003c/p\u003e"},{"header":"Limitations","content":"\u003cp\u003eThe main limitation of this study is the relatively small sample size, which may affect statistical power and generalizability. Additionally, the functional impact of identified isoform switching events was inferred from in silico predictions and requires experimental validation. The study also focused on pre-treatment samples only, without capturing dynamic splicing changes during therapy.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eETHICAL APPROVAL AND CONSENT TO PARTICIPATE\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe project was approved by Bioethics Committee of Wroclaw Medical University (consent no. KB 611/2019). All patients signed an informed consent form before the genetic test, and all procedures performed in this study were in accordance with the principles for medical research of the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCONSENT FOR PUBLICATION\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAVAILABILITY OF SUPPORTING DATA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRNA-seq data are available in the NCBI BioProject database (accession: PRJNA964497):\u003c/strong\u003e\u003cbr\u003e\u0026nbsp;https://www.ncbi.nlm.nih.gov/bioproject/PRJNA964497.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCOMPETING INTERESTS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAuthors declare no potential conflicts of interest\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFUNDING:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research was financed through a statutory subsidies by the Ministry of Health as part of the Department of Oncology Wroclaw Medical University research grant SUBZ.C280.25.016 (record number in the Simple System),\u0026nbsp;statutory subsidies by the Minister of Health as part of the Department of Genetics Wroclaw Medical University research grants SUBK.A290.22.077, SUBZ.A290.23.067, and SUBZ.C280.24.063 (record number in the Simple System) and statutory subsidies from the Genomics and Bioinformatics Laboratory of the Hirszfeld Institute of Immunology and Experimental Therapy PAS (501-26).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAUTOR CONTRIBUTIONS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eK. Nowis:\u003c/strong\u003e Conceptualization, methodology, bioinformatic analysis (RNA-seq processing, differential gene and isoform expression, isoform switch analysis, functional enrichment), data curation, investigation, writing original draft and preparing figures, writing review and editing, \u003cstrong\u003eM. Sąsiadek:\u003c/strong\u003e Funding and revision, \u003cstrong\u003eD. Martynowski:\u003c/strong\u003e Structural bioinformatics analysis and visualization \u003cstrong\u003eD. Kujawa:\u003c/strong\u003e Genetic material quality control, libraries preparation, libraries quality control, sequencing on Illumina platform, \u003cstrong\u003eI. Laczmanska:\u003c/strong\u003e Molecular interpretation insights, clinical methodology, writing review and editing, \u003cstrong\u003eP. Karpiński: Support in analysis\u003c/strong\u003e\u003cstrong\u003e,\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eM. Ekiert, E. Iwaneczko, B. Szynglarewicz, P. Kasprzak:\u0026nbsp;\u003c/strong\u003ePatient treatment, clinical data collection, result clinical interpretation\u003cstrong\u003e, M. Abrahamowska:\u0026nbsp;\u003c/strong\u003eSpecifically performing the experiments, data\u0026nbsp;collection, provision of study materials, reagents and laboratory samples\u003cstrong\u003e, R. Matkowski:\u0026nbsp;\u003c/strong\u003eFunding and clinical revision, \u003cstrong\u003eL. Laczmanski:\u003c/strong\u003e Supervision, review and editing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eACKNOWLEGMENTS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMolecular graphics and analyses performed with UCSF ChimeraX, developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco, with support from National Institutes of Health R01-GM129325 and the Office of Cyber Infrastructure and Computational Biology, National Institute of Allergy and Infectious Diseases.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBianchini G, Balko JM, Mayer IA, Sanders ME, Gianni L. Triple-negative breast cancer: challenges and opportunities of a heterogeneous disease. Nat Rev Clin Oncol. 2016 Nov;13(11):674\u0026ndash;90. \u003c/li\u003e\n\u003cli\u003eGianni L, Colleoni M, Bisagni G, Mansutti M, Zamagni C, Del Mastro L, et al. Effects of neoadjuvant trastuzumab, pertuzumab and palbociclib on Ki67 in HER2 and ER-positive breast cancer. Npj Breast Cancer. 2022 Jan 10;8(1):1. \u003c/li\u003e\n\u003cli\u003eXiong N, Wu H, Yu Z. Advancements and challenges in triple-negative breast cancer: a comprehensive review of therapeutic and diagnostic strategies. Front Oncol. 2024 May 28;14:1405491. \u003c/li\u003e\n\u003cli\u003eTufano AM, Teplinsky E, Landry CA. Updates in Neoadjuvant Therapy for Triple Negative Breast Cancer. Clin Breast Cancer. 2021 Feb;21(1):1\u0026ndash;9. \u003c/li\u003e\n\u003cli\u003eAntonini M, Mattar A, Pereira TM, Oliveira LL, Teixeira MD, Amorim AG, et al. Pathologic Complete Response and Breast Cancer Survival Post-Neoadjuvant Chemotherapy: A Systematic Review and Meta-Analysis of Real-World Data. Heliyon. 2025 Mar;e43069. \u003c/li\u003e\n\u003cli\u003eCortazar P, Zhang L, Untch M, Mehta K, Costantino JP, Wolmark N, et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis. The Lancet. 2014 Jul;384(9938):164\u0026ndash;72. \u003c/li\u003e\n\u003cli\u003eMart\u0026iacute;nez-Montiel N, Anaya-Ruiz M, P\u0026eacute;rez-Santos M, Mart\u0026iacute;nez-Contreras R. Alternative Splicing in Breast Cancer and the Potential Development of Therapeutic Tools. Genes. 2017 Oct 5;8(10):217. \u003c/li\u003e\n\u003cli\u003eClimente-Gonz\u0026aacute;lez H, Porta-Pardo E, Godzik A, Eyras E. The Functional Impact of Alternative Splicing in Cancer. Cell Rep. 2017 Aug;20(9):2215\u0026ndash;26. \u003c/li\u003e\n\u003cli\u003eOltean S, Bates DO. Hallmarks of alternative splicing in cancer. Oncogene. 2014 Nov 13;33(46):5311\u0026ndash;8. \u003c/li\u003e\n\u003cli\u003eAnczuk\u0026oacute;w O, Krainer AR. Splicing-factor alterations in cancers. RNA. 2016 Sep;22(9):1285\u0026ndash;301. \u003c/li\u003e\n\u003cli\u003eCheng R, Xiao L, Zhou W, Jin X, Xu Z, Xu C, et al. A pan-cancer analysis of alternative splicing of splicing factors in 6904 patients. Oncogene. 2021 Sep 2;40(35):5441\u0026ndash;50. \u003c/li\u003e\n\u003cli\u003eDeLigio JT, Stevens SC, Nazario-Mu\u0026ntilde;oz GS, MacKnight HP, Doe KK, Chalfant CE, et al. Serine/Arginine\u0026ndash;Rich Splicing Factor 3 Modulates the Alternative Splicing of Cytoplasmic Polyadenylation Element Binding Protein 2. Mol Cancer Res. 2019 Sep 1;17(9):1920\u0026ndash;30. \u003c/li\u003e\n\u003cli\u003eLi Y, others. CPSF4-mediated regulation of alternative splicing of HMG20B facilitates the progression of triple-negative breast cancer. J Transl Med. 2024; \u003c/li\u003e\n\u003cli\u003eLi C, Wang L, Liu Z, Wang X, Sun L, Song X, et al. Cyperotundone promotes chemosensitivity of breast cancer via SRSF1. Front Pharmacol. 2025 Mar 19;16:1510161. \u003c/li\u003e\n\u003cli\u003eNesic K, Krais JJ, Wang Y, Vandenberg CJ, Patel P, Cai KQ, et al. BRCA1 secondary splice-site mutations drive exon-skipping and PARP inhibitor resistance. Mol Cancer. 2024 Aug 5;23(1):158. \u003c/li\u003e\n\u003cli\u003eAsnani M, Hayer KE, Naqvi AS, Zheng S, Yang SY, Oldridge D, et al. Retention of CD19 intron 2 contributes to CART-19 resistance in leukemias with subclonal frameshift mutations in CD19. Leukemia. 2020 Apr;34(4):1202\u0026ndash;7. \u003c/li\u003e\n\u003cli\u003eSudhakaran M, Navarrete TG, Mej\u0026iacute;a-Guerra K, Mukundi E, Eubank TD, Grotewold E, et al. Transcriptome reprogramming through alternative splicing triggered by apigenin drives cell death in triple-negative breast cancer. Cell Death Dis. 2023 Dec 13;14(12):824. \u003c/li\u003e\n\u003cli\u003eYu S, Si Y, Yu J, Jiang C, Cheng F, Xu M, et al. SNRPB2 promotes triple‐negative breast cancer progression by controlling alternative splicing of MDM4 pre‐ mRNA. Cancer Sci. 2024 Dec;115(12):3915\u0026ndash;27. \u003c/li\u003e\n\u003cli\u003eGong S, Song Z, Spezia-Lindner D, Meng F, Ruan T, Ying G, et al. Novel Insights Into Triple-Negative Breast Cancer Prognosis by Comprehensive Characterization of Aberrant Alternative Splicing. Front Genet. 2020 Jun 11;11:534. \u003c/li\u003e\n\u003cli\u003eWu S, Wang J, Zhu X, Chyr J, Zhou X, Wu X, et al. The Functional Impact of Alternative Splicing on the Survival Prognosis of Triple-Negative Breast Cancer. Front Genet. 2021 Jan 14;11:604262. \u003c/li\u003e\n\u003cli\u003eCaggiano C, Petrera V, Ferri M, Pieraccioli M, Cesari E, Di Leone A, et al. Transient splicing inhibition causes persistent DNA damage and chemotherapy vulnerability in triple-negative breast cancer. Cell Rep. 2024 Sep;43(9):114751. \u003c/li\u003e\n\u003cli\u003eSupplitt S, Karpinski P, Sasiadek M, Laczmanski L, Kujawa D, Matkowski R, et al. The analysis of transcriptomic signature of TNBC\u0026mdash;searching for the potential RNA-based predictive biomarkers to determine the chemotherapy sensitivity. J Appl Genet. 2025 Feb;66(1):171\u0026ndash;82. \u003c/li\u003e\n\u003cli\u003eSchmieder R, Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011 Mar 15;27(6):863\u0026ndash;4. \u003c/li\u003e\n\u003cli\u003eSchubert M, Lindgreen S, Orlando L. AdapterRemoval v2: rapid adapter trimming, identification, and read merging. BMC Res Notes. 2016 Dec;9(1):88. \u003c/li\u003e\n\u003cli\u003eKim D, Paggi JM, Park C, Bennett C, Salzberg SL. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat Biotechnol. 2019 Aug;37(8):907\u0026ndash;15. \u003c/li\u003e\n\u003cli\u003eKovaka S, Zimin AV, Pertea GM, Razaghi R, Salzberg SL, Pertea M. Transcriptome assembly from long-read RNA-seq alignments with StringTie2. Genome Biol. 2019 Dec;20(1):278. \u003c/li\u003e\n\u003cli\u003eLove MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014 Dec 5;15(12):550. \u003c/li\u003e\n\u003cli\u003eVaret H, Brillet-Gu\u0026eacute;guen L, Copp\u0026eacute;e JY, Dillies MA. SARTools: A DESeq2- and EdgeR-Based R Pipeline for Comprehensive Differential Analysis of RNA-Seq Data. Mills K, editor. PLOS ONE. 2016 Jun 9;11(6):e0157022. \u003c/li\u003e\n\u003cli\u003eVitting-Seerup K, Sandelin A. IsoformSwitchAnalyzeR: analysis of changes in genome-wide patterns of alternative splicing and its functional consequences. Berger B, editor. Bioinformatics. 2019 Nov 1;35(21):4469\u0026ndash;71. \u003c/li\u003e\n\u003cli\u003eKang YJ, Yang DC, Kong L, Hou M, Meng YQ, Wei L, et al. CPC2: a fast and accurate coding potential calculator based on sequence intrinsic features. Nucleic Acids Res. 2017 Jul 3;45(W1):W12\u0026ndash;6. \u003c/li\u003e\n\u003cli\u003eTeufel F, Almagro Armenteros JJ, Johansen AR, G\u0026iacute;slason MH, Pihl SI, Tsirigos KD, et al. SignalP 6.0 predicts all five types of signal peptides using protein language models. Nat Biotechnol. 2022 Jul;40(7):1023\u0026ndash;5. \u003c/li\u003e\n\u003cli\u003eThumuluri V, Almagro Armenteros JJ, Johansen AR, Nielsen H, Winther O. DeepLoc 2.0: multi-label subcellular localization prediction using protein language models. Nucleic Acids Res. 2022 Jul 5;50(W1):W228\u0026ndash;34. \u003c/li\u003e\n\u003cli\u003eHallgren J, Tsirigos KD, Pedersen MD, Almagro Armenteros JJ, Marcatili P, Nielsen H, et al. DeepTMHMM predicts alpha and beta transmembrane proteins using deep neural networks [Internet]. 2022 [cited 2025 May 30]. Available from: http://biorxiv.org/lookup/doi/10.1101/2022.04.08.487609\u003c/li\u003e\n\u003cli\u003eShannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: A Software Environment for Integrated Models of Biomolecular Interaction Networks. Genome Res. 2003 Nov;13(11):2498\u0026ndash;504. \u003c/li\u003e\n\u003cli\u003eMeng EC, Goddard TD, Pettersen EF, Couch GS, Pearson ZJ, Morris JH, et al. UCSF ChimeraX: Tools for structure building and analysis. Protein Sci. 2023 Nov;32(11). \u003c/li\u003e\n\u003cli\u003eLongo MA, Roy S, Chen Y, Tomaszowski KH, Arvai AS, Pepper JT, et al. RAD51C-XRCC3 structure and cancer patient mutations define DNA replication roles. Nat Commun. 2023 Jul 24;14(1). \u003c/li\u003e\n\u003cli\u003eSzakal B, Branzei D. Hot on RAD51C: structure and functions of RAD51C‐XRCC3. Mol Oncol. 2023 Oct;17(10):1950\u0026ndash;2. \u003c/li\u003e\n\u003cli\u003eDutta S, Ganguly A, Chatterjee K, Spada S, Mukherjee S. Targets of Immune Escape Mechanisms in Cancer: Basis for Development and Evolution of Cancer Immune Checkpoint Inhibitors. Biology. 2023 Jan 30;12(2):218. \u003c/li\u003e\n\u003cli\u003eBaselga J, Bradbury I, Eidtmann H, Di Cosimo S, De Azambuja E, Aura C, et al. Lapatinib with trastuzumab for HER2-positive early breast cancer (NeoALTTO): a randomised, open-label, multicentre, phase 3 trial. The Lancet. 2012 Feb;379(9816):633\u0026ndash;40. \u003c/li\u003e\n\u003cli\u003eResende U, Cabello C, Oliveira Botelho Ramalho S, Zeferino LC. Predictors of Pathological Complete Response in Women with Clinical Complete Response to Neoadjuvant Chemotherapy in Breast Carcinoma. Oncology. 2018;95(4):229\u0026ndash;38. \u003c/li\u003e\n\u003cli\u003eKahles A, Lehmann KV, Toussaint NC, H\u0026uuml;ser M, Stark SG, Sachsenberg T, et al. Comprehensive Analysis of Alternative Splicing Across Tumors from 8,705 Patients. Cancer Cell. 2018 Aug;34(2):211-224.e6. \u003c/li\u003e\n\u003cli\u003eLoi S, Drubay D, Adams S, Pruneri G, Francis PA, Lacroix-Triki M, et al. Tumor-Infiltrating Lymphocytes and Prognosis: A Pooled Individual Patient Analysis of Early-Stage Triple-Negative Breast Cancers. J Clin Oncol. 2019 Mar 1;37(7):559\u0026ndash;69. \u003c/li\u003e\n\u003cli\u003eLiao G, Jiang Z, Yang Y, Zhang C, Jiang M, Zhu J, et al. Combined homologous recombination repair deficiency and immune activation analysis for predicting intensified responses of anthracycline, cyclophosphamide and taxane chemotherapy in triple-negative breast cancer. BMC Med. 2021 Dec;19(1):190. \u003c/li\u003e\n\u003cli\u003eOshi M, Patel A, Wu R, Le L, Tokumaru Y, Yamada A, et al. Enhanced immune response outperform aggressive cancer biology and is associated with better survival in triple-negative breast cancer. Npj Breast Cancer. 2022 Aug 9;8(1):92. \u003c/li\u003e\n\u003cli\u003eChun J, Buechelmaier ES, Powell SN. Rad51 Paralog Complexes BCDX2 and CX3 Act at Different Stages in the BRCA1-BRCA2-Dependent Homologous Recombination Pathway. Mol Cell Biol. 2013 Jan 1;33(2):387\u0026ndash;95. \u003c/li\u003e\n\u003cli\u003eLiu Y, Masson JY, Shah R, O\u0026rsquo;Regan P, West SC. RAD51C Is Required for Holliday Junction Processing in Mammalian Cells. Science. 2004 Jan 9;303(5655):243\u0026ndash;6. \u003c/li\u003e\n\u003cli\u003eRodrigue A, Coulombe Y, Jacquet K, Gagn\u0026eacute; JP, Roques C, Gobeil S, et al. The RAD51 paralogs ensure cellular protection against mitotic defects and aneuploidy. J Cell Sci. 2013 Jan 1;126(1):348\u0026ndash;59. \u003c/li\u003e\n\u003cli\u003eSage JM, Gildemeister OS, Knight KL. Discovery of a Novel Function for Human Rad51. J Biol Chem. 2010 Jun;285(25):18984\u0026ndash;90. \u003c/li\u003e\n\u003cli\u003eSU CH, CHANG WS, HU PS, HSIAO CL, JI HX, LIAO CH, et al. Contribution of DNA Double-strand Break Repair Gene \u0026lt;em\u0026gt;XRCC3\u0026lt;/em\u0026gt; Genotypes to Triple-negative Breast Cancer Risk. Cancer Genomics - Proteomics. 2015 Nov 1;12(6):359. \u003c/li\u003e\n\u003cli\u003eXinyi X, Gong Y. The role of ATP‐binding cassette subfamily G member 1 in tumor progression. Cancer Med. 2024 Jun;13(12):e7285. \u003c/li\u003e\n\u003cli\u003eWang B, Xing A, Li G, Liu L, Xing C. SNHG14 promotes triple‐negative breast cancer cell proliferation, invasion, and chemoresistance by regulating the ERK / MAPK signaling pathway. IUBMB Life. 2024 Dec;76(12):1295\u0026ndash;308. \u003c/li\u003e\n\u003cli\u003eLiu B, Lu T, Wang Y, Zhang G, Fu L, Yu M, et al. Overexpression of LncRNA SNHG14 as a biomarker of clinicopathological and prognosis value in human cancers: A meta-analysis and bioinformatics analysis. Front Genet. 2022 Oct 6;13:945919. \u003c/li\u003e\n\u003cli\u003eHua H, Xie H, Zheng J, Lei L, Deng Z, Yu C. RNF7 Facilitated the Tumorigenesis of Pancreatic Cancer by Activating PI3K/Akt Signaling Pathway. Adnan M, editor. Oxid Med Cell Longev. 2023 Jan 4;2023:1\u0026ndash;17. \u003c/li\u003e\n\u003cli\u003eXiao Y, Jiang Y, Song H, Liang T, Li Y, Yan D, et al. RNF7 knockdown inhibits prostate cancer tumorigenesis by inactivation of ERK1/2 pathway. Sci Rep. 2017 Mar 2;7(1):43683. \u003c/li\u003e\n\u003cli\u003eXiao C, Zhang W, Hua M, Chen H, Yang B, Wang Y, et al. RNF7 inhibits apoptosis and sunitinib sensitivity and promotes glycolysis in renal cell carcinoma via the SOCS1/JAK/STAT3 feedback loop. Cell Mol Biol Lett. 2022 Dec;27(1):36. \u003c/li\u003e\n\u003cli\u003eBraschi B, Bruford EA, Cavanagh AT, Neuman SD, Bashirullah A. The bridge-like lipid transfer protein (BLTP) gene group: introducing new nomenclature based on structural homology indicating shared function. Hum Genomics. 2022 Dec 2;16(1):66. \u003c/li\u003e\n\u003cli\u003eWang Z, Wang Y, Li Z, Xue W, Hu S, Kong X. Lipid metabolism as a target for cancer drug resistance: progress and prospects. Front Pharmacol. 2023 Sep 28;14:1274335. \u003c/li\u003e\n\u003cli\u003eChoi MS, Graves MJ, Matoo S, Storad ZA, El Sheikh Idris RA, Weck ML, et al. The small EF-hand protein CALML4 functions as a critical myosin light chain within the intermicrovillar adhesion complex. J Biol Chem. 2020 Jul;295(28):9281\u0026ndash;96. \u003c/li\u003e\n\u003cli\u003eGuo X, Xia S, Ge T, Lin Y, Hu S, Wu H, et al. \u003cem\u003eAtp13a5\u003c/em\u003e Marker Reveals Pericyte Specification in the Mouse Central Nervous System. J Neurosci. 2024 Oct 23;44(43):e0727242024. \u003c/li\u003e\n\u003cli\u003eHe W, Sun Y, Ge J, Wang X, Lin B, Yu S, et al. STRA6 regulates tumor immune microenvironment and is a prognostic marker in BRAF-mutant papillary thyroid carcinoma. Front Endocrinol. 2023 Feb 10;14:1076640. \u003c/li\u003e\n\u003cli\u003eMoore CB, Bergstralh DT, Duncan JA, Lei Y, Morrison TE, Zimmermann AG, et al. NLRX1 is a regulator of mitochondrial antiviral immunity. Nature. 2008 Jan;451(7178):573\u0026ndash;7. \u003c/li\u003e\n\u003cli\u003ePort\u0026eacute; S, Valencia E, Yakovtseva EA, Borr\u0026agrave;s E, Shafqat N, Debreczeny J\u0026Eacute;, et al. Three-dimensional Structure and Enzymatic Function of Proapoptotic Human p53-inducible Quinone Oxidoreductase PIG3. J Biol Chem. 2009 Jun;284(25):17194\u0026ndash;205. \u003c/li\u003e\n\u003cli\u003eSingh K, Roy M, Prajapati P, Lipatova A, Sripada L, Gohel D, et al. NLRX1 regulates TNF-\u0026alpha;-induced mitochondria-lysosomal crosstalk to maintain the invasive and metastatic potential of breast cancer cells. Biochim Biophys Acta BBA - Mol Basis Dis. 2019 Jun;1865(6):1460\u0026ndash;76. \u003c/li\u003e\n\u003cli\u003eWang ET, Sandberg R, Luo S, Khrebtukova I, Zhang L, Mayr C, et al. Alternative isoform regulation in human tissue transcriptomes. Nature. 2008 Nov;456(7221):470\u0026ndash;6. \u003c/li\u003e\n\u003cli\u003eLv X, Sun X, Gao Y, Song X, Hu X, Gong L, et al. Targeting RNA splicing modulation: new perspectives for anticancer strategy? J Exp Clin Cancer Res. 2025 Jan 30;44(1):32. \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Triple negative breast cancer, TNBC, alternative splicing, transcriptomics, therapy response, neoadjuvant therapy","lastPublishedDoi":"10.21203/rs.3.rs-7041176/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7041176/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eTriple-negative breast cancer (TNBC) is characterized by limited treatment options and high variability in response to neoadjuvant chemotherapy (NAC). While DNA-level alterations have been widely studied, post-transcriptional regulation through alternative splicing remains unexplored in this context.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eWe performed a transcriptome -wide analysis of differential isoform usage in pre-treatment TNBC biopsies from patients stratified by NAC response. Using IsoformSwitchAnalyzer and STRING, we assessed the functional consequences of isoform switching alterations in coding potential, protein domains, and pathway involvement. Structural models of XRCC3 isoforms were generated using AlphaFold and ChimeraX.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eNon-responder exhibited significantly higher rates of isoform switching, particularly involving transcription start/termination site changes and intron retention. Enrichment analyses revealed immune-related pathway signatures in complete responders and DNA repair in both complete and partial responders. Among key genes, the \u003cem\u003eXRCC3\u003c/em\u003e emerged as a notable candidate, with non-responder showing shift toward truncated isoform lacking domains required for interactions with RAD51 and RAD51C. This structural loss likely impairs homologs recombination repair and may contribute to the observed resistance phenotype.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eIsoform switching is a significant regulatory mechanism associated with chemotherapy response in TNBC. Splicing alterations affecting DNA repair and immune may serve as predictive biomarkers. These findings support the integration of isoform-level analysis into clinical transcriptomics for precision oncology.\u003c/p\u003e","manuscriptTitle":"Isoform switching as a key mechanism in chemotherapy resistance in triple-negative breast cancer","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-23 08:33:25","doi":"10.21203/rs.3.rs-7041176/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"013177be-c38e-4241-9a09-255496e67410","owner":[],"postedDate":"July 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-26T12:16:13+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-23 08:33:25","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7041176","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7041176","identity":"rs-7041176","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}