PO.ET03.06 · 实验与分子治疗
开发ADC耐药肿瘤模型用于新一代抗癌疗法的疗效评估
Developing ADC-resistant tumor models for efficacy evaluation of next-generation anticancer therapies
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
引言：抗体-药物偶联物（ADC）将单克隆抗体与细胞毒性有效载荷偶联。Trastuzumab deruxtecan（DS8201）靶向HER2，datopotamab deruxtecan（DS1062）靶向TROP2。二者均通过可切割连接子使用拓扑异构酶I抑制剂（Dxd）。在靶点介导的内化后，蛋白酶释放有效载荷，导致DNA损伤、凋亡和旁观者杀伤。然而，耐药可能源于抗原丢失、内化改变、有效载荷释放受损、外排泵或凋亡调控。我们通过慢性给药、慢病毒转导和Cas9介导的敲除构建了ADC耐药细胞和类器官模型。
方法：首先，通过慢病毒转导将外排泵蛋白（ABCB1、ABCG2）导入HER2+ HCC1954细胞（HCC1954-ABCB1、HCC1954-ABCG2）；Western blot证实了过表达。通过CellTiterGlo（CTG）实验评估对Dxd/ADC的细胞毒性。其次，体外慢性暴露产生了HCC1954-DS8201R和HCC1954-DxdR。通过流式细胞术（FACS）评估HER2表达，通过IncuCyte成像评估ADC结合/内化。RNA-seq研究耐药机制。最后，在两个DS1062应答的类器官中使用CRISPR敲除TROP2。通过免疫荧光证实TROP2丢失，使用CTG实验检测ADC敏感性。ADC耐药细胞的异种移植模型正在开发中以进行进一步的体内研究。
结果：HCC1954-ABCB1对MMAE表现出显著耐药（IC50：85.8 nM vs 亲本0.3 nM），但对Dxd无变化。而HCC1954-ABCG2对Dxd耐药（IC50：>10 μM vs 亲本0.509 μM），对MMAE影响极小。RC48（携带MMAE的Her2靶向ADC）和DS8201被确认为基于有效载荷的耐药。HCC1954-ABCB1对RC48耐药（IC50：>100 nM vs 亲本0.288 nM），而HCC1954-ABCG2对DS8201耐药（IC50：>100 nM vs 亲本5.2434 nM）。这些发现表明外排泵蛋白优先转运不同的细胞毒性有效载荷。对Dxd的慢性暴露产生的HCC1954-DxdR细胞对Dxd（IC50：>10 μM vs 亲本0.2023 μM）和DS8201（IC50：>100 nM vs 2.1695 nM）的敏感性均降低。同样，对DS8201的慢性暴露产生了HCC1954-DS8201R，表现出显著耐药（IC50：>70 nM vs 2.1695 nM）。在HER2降低的HCC1954-DS8201R中（经FACS和RNA-seq证实），显示HER2结合/内化减少，而HCC1954-DxdR表现出ABCG2扩增。TROP2敲除效率为30%（LU11786B）和5%（BR9465B）。TROP2敲除模型的细胞毒性对DS1062耐药，伴随DS1062摄取丢失以及DNA损伤标志物yH2AX缺失。ADC耐药细胞的异种移植模型能够在体内生长。
结论：我们开发了一组ADC耐药肿瘤模型以研究获得性耐药机制。这些模型为推进ADC治疗提供了宝贵的临床前平台。
查看英文原文 English abstract
Introduction: Antibody-drug conjugates (ADCs) pair monoclonal antibodies with cytotoxic payloads. Trastuzumab deruxtecan (DS8201) targets HER2, and datopotamab deruxtecan (DS1062) targets TROP2. Both use a topoisomerase I inhibitor (Dxd) via a cleavable linker. After target-mediated internalization, proteases release the payload, causing DNA damage, apoptosis, and bystander killing. However, resistance may stem from antigen loss, altered internalization, impaired payload release, efflux pumps, or apoptosis regulation. We developed ADC-resistant cell and organoid models via chronic dosing, lentivirus transduction, and Cas9 mediated knock out.
Methods: First, efflux pump proteins (ABCB1, ABCG2) were introduced into HER2+ HCC1954 cells via lentiviral transduction (HCC1954-ABCB1, HCC1954-ABCG2); overexpression was confirmed by Western blot. Cytotoxicity to Dxd/ADC was evaluated by CellTiterGlo (CTG)-‑assay. Second, chronic in vitro exposure generated HCC1954-DS8201R and HCC1954-DxdR. HER2 expression was assessed by flow cytometry (FACS), and ADC binding/internalization by IncuCyte imaging. RNA-seq investigates the resistance mechanisms. Finally, TROP2 was knocked out using CRISPR in two DS1062 responsive organoids. Loss of TROP2 was confirmed using immunofluorescence, and ADC sensitivity was tested using CTG assay. Xenograft models of ADC-resistant cells are in development for further in vivo studies.
Results: HCC1954-ABCB1 showed remarkable resistance to MMAE (IC₅₀: 85.8 nM vs 0.3 nM parental) but showed no change to Dxd. While, HCC1954-ABCG2 is resistant to Dxd (IC₅₀: >10 μM vs 0.509 μM parental), with minimal MMAE impact. The RC48 (Her2-targeting ADC with MMAE) and DS8201 are confirmed payload-based resistances. HCC1954-ABCB1 is resistant to RC48 (IC₅₀: >100 nM vs 0.288 nM parental), while HCC1954-ABCG2 is resistant to DS8201 (IC₅₀: >100 nM vs 5.2434 nM parental). These findings suggest that efflux pump proteins preferentially transport distinct cytotoxic payloads. Chronic exposure to Dxd generated HCC1954‑DxdR cells reduced sensitivity to both Dxd (IC₅₀: >10 μM vs 0.2023 μM parental) and DS8201 (IC₅₀: >100 nM vs 2.1695 nM). Similarly, chronic exposure to DS8201 produced the HCC1954‑DS8201R, showing significant resistance (IC₅₀: >70 nM vs 2.1695 nM). In HCC1954‑DS8201R with reduced HER2, confirmed by FACS and RNA seq, showed a decrease in HER2 binding/internalization, whereas HCC1954‑DxdR exhibited ABCG2 amplification. TROP2 KO efficiency was 30% (LU11786B) and 5% (BR9465B). Cytoxicity of TROP2 KO models resistant to DS1062. with loss of DS1062 uptake and absence of the DNA damage marker yH2AX. Xenograft models of ADC-resistant cells are able to grow in vivo .
Conclusion: We developed a panel of ADC-resistant tumor models to study mechanisms of acquired resistance. These models provide a valuable preclinical platform for advancing ADC therapy.
利益披露 Disclosure
J. Feng, None..
F. Zhao, None..
X. Dong, None..
A. Hua, None..
C. Nie, None..
D. Verstegen, None..
C. Hulsebosch, None..
W. Qian, None..
S. Guo, None..
J. Wang, None..
L. Krenning, None..
M. Putker, None..
L. Bourre, None..
P. Wang, None..
J. Zhou, None.