PO.TB10.13 · 肿瘤生物学
整合多组学分析确定神经元重编程和血管重塑为复发性高级别浆液性卵巢癌的标志
Integrative multi-omic analysis identifies neuronal reprogramming and vascular remodeling as hallmarks of recurrent high-grade serous ovarian carcinoma
该海报暂无可下载的资料
AACR 官方页面
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
背景:高级别浆液性卵巢癌(HGSOC)是上皮性卵巢癌最常见和最具侵袭性的亚型,超过80%的患者在初始治疗后经历复发。复发性疾病通常耐化疗,仍是妇科癌症死亡的主要原因,然而驱动复发的分子变化,尤其是在细胞类型特异性水平上,仍知之甚少。
目的:定义区分晚期原发与复发性疾病的细胞类型解析的转录程序,将这些程序置于从良性组织到复发的分期对齐的细胞背景中,并提名具有外部重复验证和基因组一致性的稳健生物标志物。
方法:对配对的晚期原发和复发HGSOC标本进行snRNAseq分析,并纳入额外的良性和早期肿瘤用于分期背景化。使用DESeq2和limma voom进行晚期原发与复发之间的配对pseudobulk差异表达分析,采用患者分区、低表达过滤、修剪均值标准化和错误发现率控制。在四个外部队列中评估46个显著基因的重复验证。此外,将来自全外显子组测序生成的晚期原发肿瘤的体细胞变异和拷贝数谱与panel位点的转录效应进行比较。使用精确秩统计和bootstrap稳健性评估将转录特征与复发时间相关联。
结果:配对晚期原发和复发HGSOC的snRNAseq揭示了一个复发程序,其映射至肿瘤相关内皮和恶性上皮中的神经元谱系状态。FLT1、STC1和CA2以及内皮长非编码RNA MEF2C AS1的选择性抑制;FLT1和CA2与MEF2C AS1共同构成该疾病中新的复发相关内皮标志物。恶性细胞表现出一种由NTM和KCNQ5诱导及PEG3相互下调定义的新型神经元重编程特征,基质区室显示ABCA8减少和新标志物PLCXD3。晚期原发肿瘤的全外显子组拷贝数分析提供了与转录组的方向性一致,包括KCNQ5的深度缺失、FLT1、ESM1和NTM的晚期拷贝数减少与其在复发时的诱导一致,以及C7和PLCXD3的晚期获得与其在复发时的抑制一致。
结论:我们的整合多组学分析揭示,复发性HGSOC以肿瘤相关内皮的协调血管重塑和恶性上皮的神经元重编程为特征,并伴随基质重构。这些发现提名细胞状态特异性生物标志物和通路,用于复发性疾病的前瞻性验证和治疗靶向。
查看英文原文 English abstract
Background: High-grade serous ovarian carcinoma (HGSOC) is the most common and aggressive subtype of epithelial ovarian cancer, with over 80% of patients experiencing recurrence after initial therapy. Recurrent disease is typically chemoresistant and remains the leading cause of gynecologic cancer mortality, yet the molecular changes driving recurrence, particularly at the cell type-specific level, are poorly understood.
Objective: To define cell type resolved transcriptional programs that distinguish late primary from recurrent disease, to place these programs into a stage aligned cellular context from benign tissue through recurrence, and to nominate robust biomarkers with external replication and genomic concordance.
Methods: Matched late primary and recurrent HGSOC specimens were profiled by snRNAseq with additional benign and early tumors included for stage contextualization. Paired pseudobulk differential expression between late primary and recurrence was performed using DESeq2 and limma voom with patient blocking, low expression filtering, trimmed mean normalization, and false discovery rate control. Replication of 46 significant genes was assessed in four external cohorts. In addition, somatic variant and copy number profiles from late primary tumors generated by whole exome sequencing were compared with transcriptional effects at panel loci. A transcriptional signature was related to time to recurrence using exact rank statistics with bootstrap assessment of robustness.
Results: snRNAseq of matched late primary and recurrent HGSOC revealed a recurrence program that maps to tumor associated endothelium and to a neuronal lineage state in malignant epithelium. FLT1 STC1 and CA2 and selective suppression of the endothelial long noncoding RNA MEF2C AS1; FLT1 and CA2 together with MEF2C AS1 constitute novel recurrence associated endothelial markers in this disease. Malignant cells exhibited a novel neuronal reprogramming signature defined by induction of NTM and KCNQ5 with reciprocal downregulation of PEG3 , and stromal compartments showed reduced ABCA8 and the novel marker PLCXD3 . Whole exome copy number profiling of late primary tumors provided directional concordance with the transcriptome, including deep loss at KCNQ5 , late stage copy number reduction at FLT1 , ESM1, and NTM aligned with their induction at recurrence, and late stage gain at C7 and PLCXD3 aligned with its suppression at recurrence.
Conclusion: Our integrative multi‑omic analysis reveals that recurrent HGSOC is characterized by coordinated vascular remodeling in tumor‑associated endothelium and neuronal reprogramming in malignant epithelium, accompanied by stromal reconfiguration. These findings nominate cell‑state-specific biomarkers and pathways for prospective validation and therapeutic targeting in recurrent disease.
利益披露 Disclosure
I. Huseynli, None..
G. Gibson, None..
B. B. Benigno, None..
S. N. Housley, None.