PO.ET02.06 · 实验与分子治疗
通过临床蛋白酶谱分析和基于结构的计算机模拟，对用于EGFR阳性肿瘤的PSM101 Lock-抗体进行整合设计
Integrative design of PSM101 Lock-antibody for EGFR-positive tumors through clinical protease profiling and structure-based computer simulation
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
目的：靶点相关毒性仍然是抗EGFR抗体的一个主要局限，例如cetuximab治疗的患者因皮肤毒性而出现9.8%的停药率。蛋白酶激活的前抗体可以提高肿瘤选择性；然而，其性能受到临床蛋白酶活性异质性和抗体特异性CDR结构限制的制约。我们旨在开发一种优化的EGFR Lock-抗体（PSM101），具有更高的阻断效率、高肿瘤特异性可切割性以及最小的正常组织结合。
方法：PSM101通过将Antibody Lock™屏蔽结构域、肿瘤相关蛋白酶底物和优化的连接子融合到一个抗EGFR抗体上而构建。对临床肿瘤组织进行谱分析，以识别在多种癌症类型中高活性的蛋白酶底物。建立了一个自动化的基于结构的模拟平台（MSCS 2.0），以预测各变体设计的覆盖率和蛋白酶可切割性。对先导PSM101分子进行以下评估：（1）与人正常细胞的结合（流式细胞术），（2）在临床肿瘤组织中的蛋白酶激活（IHZ检测），以及（3）在EGFR阳性CDX和PDX小鼠模型中的抗肿瘤疗效。
结果：来自头颈、肺、结肠和胃肿瘤的临床蛋白酶谱分析识别出具有高且一致的肿瘤相关切割活性的底物。MSCS 2.0成功预测了具有最佳屏蔽覆盖率和可切割性的变体，从而实现对最终PSM101设计的理性选择。与cetuximab相比，PSM101与人正常细胞的结合降低了100-200倍，并在头颈肿瘤组织中表现出强效激活。在体内，PSM101在CDX和PDX模型中均实现了显著的抗肿瘤活性。
结论：我们开发了一种蛋白酶激活的EGFR Lock-抗体PSM101，具有显著改善的肿瘤选择性和强大的治疗疗效。MSCS 2.0为前抗体的早期优化提供了一个预测框架，可能加速下一代抗体药物开发并降低临床毒性。正在进行的研究包括在食蟹猴中对PSM101进行安全性评估。
查看英文原文 English abstract
Objective: On-target toxicity remains a major limitation of anti-EGFR antibodies, as exemplified by a 9.8% discontinuation rate in cetuximab-treated patients due to skin toxicity. Protease-activated pro-antibodies can improve tumor selectivity; however, their performance is limited by heterogenous clinical protease activity and antibody-specific CDR structural constraints. We aimed to develop an optimized EGFR Lock-antibody (PSM101) with improved blocking efficiency, high tumor-specific cleavability, and minimal normal-tissue binding.
Methods: PSM101 was engineered by fusing the Antibody Lock™ masking domain, tumor-associated protease substrates, and optimized linkers to an anti-EGFR antibody. Clinical tumor tissues were profiled to identify highly active protease substrates across multiple cancer types. An automated structure-based simulation platform (MSCS 2.0) was established to predict cover rate and protease cleavability for variant designs. Lead PSM101 molecules were evaluated for (1) binding to human normal cells (flow cytometry), (2) protease activation in clinical tumor tissues (IHZ assay), and (3) antitumor efficacy in EGFR-positive CDX and PDX mouse models.
Results: Clinical protease profiling from head and neck, lung, colon, and gastric tumors identified substrates with high and consistent tumor-associated cleavage activity. MSCS 2.0 successfully predicted variants with optimal masking coverage and cleavability, enabling rational selection of the final PSM101 design. PSM101 demonstrated a 100-200-fold reduction in binding to human normal cells compared with cetuximab, and showed robust activation in head and neck tumor tissues. In vivo, PSM101 achieved significant antitumor activity in both CDX and PDX models.
Conclusion: We developed a protease-activated EGFR Lock-antibody, PSM101, with markedly improved tumor selectivity and strong therapeutic efficacy. MSCS 2.0 provides a predictive framework for early-stage optimization of pro-antibodies, which may accelerate next-generation antibody drug development and reduce clinical toxicity. Ongoing studies include safety evaluation of PSM101 in cynomolgus monkeys.
利益披露 Disclosure
Y. Cheng, None..
Y. Lu, None..
Y. Lee, None..
C. Hung, None.