超过 75% 的上皮性卵巢癌 (EOC) 患者在初次治疗后会出现疾病复发，这凸显了我们对化疗后 EOC 进展过程中化疗耐药人群如何演变的不完全了解。在这里，我们展示了两种紫杉醇 (PTX) 治疗方法（单次高剂量和每周节拍剂量，持续四个星期）如何产生独特的化疗耐药群体。使用机械相关的藻酸盐微球并结合转录谱分析和异质性分析，我们发现这些 PTX 治疗方案产生了独特且有弹性的亚群，这些亚群在代谢重编程特征、对 PTX 和失巢凋亡的抵抗性的获得以及癌症干细胞的富集方面有所不同（CSC）和多倍体巨型癌细胞（PGCC）能够补充大量细胞群。我们使用非靶向代谢组学研究了这些代谢重编程事件的寿命，发现与干性和治疗诱导的衰老相关的代谢物在富含 CSC 和 PGCC 的人群中特别丰富。预测网络分析表明，抗氧化机制的活性可能因时间和接触 PTX 的不同而有差异。我们的结果说明了当前的标准化疗如何通过选择本质耐药亚群或促进耐药机制的进化来促进化学耐药 EOC 亚群的发展。此外，我们的工作描述了每个不同耐药亚群的独特表型特征，从而突出了可用于更有效治疗的潜在漏洞。© 2023。作者。

More than 75% of epithelial ovarian cancer (EOC) patients experience disease recurrence after initial treatment, highlighting our incomplete understanding of how chemoresistant populations evolve over the course of EOC progression post chemotherapy treatment. Here, we show how two paclitaxel (PTX) treatment methods- a single high dose and a weekly metronomic dose for four weeks, generate unique chemoresistant populations. Using mechanically relevant alginate microspheres and a combination of transcript profiling and heterogeneity analyses, we found that these PTX-treatment regimens produce distinct and resilient subpopulations that differ in metabolic reprogramming signatures, acquisition of resistance to PTX and anoikis, and the enrichment for cancer stem cells (CSCs) and polyploid giant cancer cells (PGCCs) with the ability to replenish bulk populations. We investigated the longevity of these metabolic reprogramming events using untargeted metabolomics and found that metabolites associated with stemness and therapy-induced senescence were uniquely abundant in populations enriched for CSCs and PGCCs. Predictive network analysis revealed that antioxidative mechanisms were likely to be differentially active dependent on both time and exposure to PTX. Our results illustrate how current standard chemotherapies contribute to the development of chemoresistant EOC subpopulations by either selecting for intrinsically resistant subpopulations or promoting the evolution of resistance mechanisms. Additionally, our work describes the unique phenotypic signatures in each of these distinct resistant subpopulations and thus highlights potential vulnerabilities that can be exploited for more effective treatment.© 2023. The Author(s).