Alternative Splicing-Driven Drug Response Heterogeneity Across CNS, Lung, and Skin Cancers Cell Lines = CNS, 폐암, 피부암 전반에 걸친 대체 스플라이싱 기반 약물 반응의 이질성

다국어 초록 (Multilingual Abstract)

Alternative splicing (AS) is a fundamental mechanism that contributes to transcriptomic diversity and cellular plasticity, and is often dysregulated in various cancer types. AS plays a pivotal role in shaping tumor heterogeneity, giving rise to distinct cellular phenotypes within the same tumor and influencing divergent therapeutic outcomes. Despite that, the extent to which AS affects drug response remains poorly understood. This study was systematically investigated the association between AS and treatment sensitivity in cancer, focusing on cell lines derived from central nervous system (CNS), lung, and skin tumors. Using RNA-seq data from 292 cancer cell lines and drug response profiles from the Genomics of Drug Sensitivity in Cancer (GDSC) database, splicing events were identified by GESS and EventPointer and integrated with differential gene expression analysis, pathway enrichment, and drug correlation analysis. A total of 2,668,455 AS events were detected, with Exon Skipping being the most prevalent. Lung cancer cell lines were shown the highest AS frequency, followed by CNS and skin cancer cell lines, mirroring their varying degrees of genomic instability and lineage-spec if ic diversity across cancer types. Differential Gene Expression (DEG) and pathway analysis * A thesis submitted to the Council of the Graduate School Kyungpook National University in partial fulfillment of the requirements for the degree of Master of Science in December 2025 revealed cancer-type-specific transcriptomic programs influenced by AS, including the MYC, p53, and EMT pathways. Notably, 411 significant gene–drug associations were identified, such as SHCBP1-WNT-C59 in CNS cancers, AUTS2-drug associations in lung cancers, and SCP2-HDAC inhibitor sensitivity in skin cancers. These findings underscore how AS contributes to functional heterogeneity by altering gene expression and protein isoform repertoires in a context-dependent manner, ultimately impacting therapeutic response. In conclusion, this study demonstrates that AS is a key contributor to both the molecular and functional heterogeneity of tumors. The AS landscape not only reflects the complexity of tumor biology but also holds promise as a predictive biomarker for drug sensitivity, offering new opportunities for splicing-informed precision oncology.

국문 초록 (Abstract)

Alternative splicing(AS)은 전사체 다양성과 세포 가소성을 형성하는 데 기여하는 핵심적인 기전이며, 다양한 암 유형에서 자주 조절 이상을 보임. AS는 종양 이질성을 형성하는 데 중추적인 역할을 하여, 동일한 종양 내에서도 서로 다른 세포 표현형을 만들어내고 치료 반응의 차이를 유발함. 그럼에도 불구하고, AS가 약물 반응에 어떤 영향을 미치는지에 대해서는 아직 충분히 밝혀져 있지 않음. 본 연구는 중추신경계(CNS), 폐, 피부 종양에서 유래한 암 세포주를 중심으로 AS와 치료 민감성 간의 연관성을 체계적으로 조사하였음. 292개의 암 세포주 RNA-seq 데이터와 Genomics of Drug Sensitivity in Cancer(GDSC) 데이터베이스의 약물 반응 정보를 활용하여, GESS와 EventPointer 기반으로 스플라이싱 이벤트를 식별하고 이를 차등 유전자 발현 분석, 경로 분석, 약물 상관 분석과 통합하였음. 총 2,668,455개의 AS 이벤트가 검출되었으며, 그중 Exon Skipping이 가장 많이 관찰되었음. 폐암 세포주가 가장 높은 AS 빈도를 보였고, 그 다음으로 CNS 및 피부 암 세포주가 뒤를 이었으며, 이는 암 유형별로 존재하는 게놈 불안정성과 계통 특이적 다양성의 정도를 반영함. 차등 유전자 발현(DEG) 및 경로 분석은 AS의 영향을 받는 암종 특이적 전사체 프로그램을 밝혀냈으며, 그중에는 MYC, p53, EMT 경로가 포함되었음. 특히 SHCBP1–WNT-C59(CNS), AUTS2–관련 약물(폐암), SCP2–HDAC 억제제 민감성(피부암) 등 총 411개의 유의한 유전자–약물 연관성이 확인되었음. 이러한 결과는 AS가 유전자 발현과 단백질 아이소폼을 변화시켜 맥락 의존적으로 기능적 이질성에 기여하며, 궁극적으로 치료 반응에 영향을 미친다는 점을 강조함. 결론적으로, 본 연구는 AS가 종양의 분자적·기능적 이질성의 주요 기여자임을 보여줌. AS 랜드스케이프는 종양 생물학의 복잡성을 반영할 뿐 아니라 약물 민감성을 예측하는 바이오마커로써 잠재력을 지니며, 스플라이싱 기반 정밀의학의 새로운 가능성을 제시함.

목차 (Table of Contents)

• I. Introduction 1
• II. Material and Methods 5
• 1. Sample Collection and RNA-Seq Data 5
• 2. RNA-Seq Alignment and Quantification 5
• 3. Alternative Splicing Event Detection 5
• 4. Drug Response Data Processing 6
• 5. Differential Gene Expression Analysis 6
• 6. Pathway Enrichment-Based Methods 6
• 7. Pathway Activity-Based Methods 7
• 8. Data Visualization 7
• III. Results 8
• 1. Overview of Alternative Splicing Events in Cancer Cell Lines 8
• 2. Identification of Differentially Expressed Genes 15
• 3. Pathway Enrichment Analysis 17
• 4. Correlation Between Alternative Splicing and Drug Response 21
• 5. Cancer Type–Specific DEG Analysis 25
• 6. Pathway Enrichment Analysis by Cancer Type 29
• 7. Cancer Type–Specific Analysis of Drug Response Associated with Alternative Splicing 41
• IV. Discussion 53
• V. References 63
• Abstract 67
• 국문 초록 69