PO.CH01.02 · 化学
高内涵形态学分析鉴定自噬调节化合物及其脱靶活性
High-content morphology profiling identifies autophagy-modulating compounds and their off-target activities
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
背景：自噬为许多肿瘤提供了应对应激、调整代谢和抵御治疗的途径。因此，干扰这一过程的化合物引起了广泛关注。为了高效发现此类化合物，我们依赖能够检测细胞结构细微变化的筛选方法。Cell Painting（细胞染色）分析提供了这种可能性：通过捕捉丰富的形态学快照，它使我们能够观察不同化合物如何改变参与自噬的细胞器和通路。
方法：我们将HepG2细胞暴露于几种特征明确的自噬抑制剂——氯喹、巴弗洛霉素A1、蝙蝠葛碱、蝙蝠葛苏林碱和多索吗啡——以及两种已确立的激活剂雷帕霉素和PP242中24小时。MCOPPB及其衍生物MS107和MS108也被纳入同一组样本中，同时还有JUMP靶标对照。染色后，我们在Yokogawa CV8000显微镜上对细胞进行成像，每孔记录跨九个视野的四个通道。从这些图像中，我们使用pycytominer提取并归一化了1,065个单细胞特征。在R中进行了降维（UMAP）和相关性分析，并通过免疫荧光评估了自噬标志物（LAMP2B、p62/SQSTM1和LC3）。
结果：形态学特征以明确的界限将抑制剂与激活剂区分开来，每组形成紧密的聚类。抑制剂产生了预期的特征：高尔基体碎裂、线粒体肿胀和内质网结构扩张，均表明自噬流受损。具有已知机制的化合物表现符合预期。引人注目的是，通常被归类为NOP受体激动剂的MCOPPB直接落入抑制剂聚类中。后续染色证实了这一模式：自噬体-溶酶体融合受损、p62积累和LC3变化均指向自噬抑制。
结论：Cell Painting被证明能够通过化合物特征性的细胞器水平效应有效区分自噬调节化合物。MCOPPB的意外行为凸显了一种此前被忽视的机制。将形态学分析与基于标志物的验证相结合，为识别具有潜在肿瘤学相关性的新型自噬调节剂提供了一个实用框架。
本工作得到捷克共和国技术署项目PERMED：T2BA（TN02000109）的支持。我们还感谢基础设施项目CZ-OPENSCREEN（LM2023052）、EATRIS-CZ（LM2023053）、IGA_LF_2025_21以及由OP JAK支持的SALVAGE项目（CZ.02.01.01/00/22_008/0004644）的贡献，并由欧盟和国家预算共同资助。
查看英文原文 English abstract
Background: Autophagy gives many tumors a way to cope with stress, adjust their metabolism, and weather therapy. Compounds that interfere with this process, therefore, attract considerable interest. To efficiently spot such compounds, we rely on screening methods that detect subtle shifts in cell structure. The Cell Painting assay offers this possibility: by capturing rich morphological snapshots, it lets us see how different compounds alter organelles and pathways involved in autophagy.
Methods: We exposed HepG2 cells for 24 hours to several well-characterized autophagy inhibitors-chloroquine, bafilomycin A1, dauricine, daurisoline, and dorsomorphine-as well as two established activators, rapamycin and PP242. MCOPPB and its derivatives MS107 and MS108 were included in the same panels, alongside JUMP-target controls. After staining, we imaged the cells on a Yokogawa CV8000 microscope, recording four channels across nine fields per well. From these images, we extracted and normalized 1,065 single-cell features using pycytominer. Dimensionality reduction (UMAP) and correlation analyses were performed in R, and autophagy markers (LAMP2B, p62/SQSTM1, and LC3) were assessed by immunofluorescence.
Results: The morphological profiles separated inhibitors from activators with a clear margin, each group forming tight clusters. Inhibitors produced the expected signatures: a fragmented Golgi apparatus, swollen mitochondria, and expanded ER structures, all indicative of impaired autophagic flux. Compounds with known mechanisms behaved as anticipated. Strikingly, MCOPPB-usually classified as a NOP receptor agonist-fell directly into the inhibitor cluster. Follow-up staining corroborated this pattern: disrupted autophagosome-lysosome fusion, p62 accumulation, and LC3 changes all pointed toward autophagy inhibition.
Conclusion: Cell Painting proved effective at distinguishing autophagy-modulating compounds through their characteristic organelle-level effects. The unexpected behaviour of MCOPPB highlights a previously overlooked mechanism. Combining morphological profiling with marker-based validation provides a practical framework for identifying new modulators of autophagy with potential oncological relevance.
This work was supported by the Technology Agency of the Czech Republic project PERMED: T2BA (TN02000109). We also acknowledge the contributions from infrastructural projects CZ-OPENSCREEN (LM2023052), EATRIS-CZ (LM2023053), IGA_LF_2025_21, and the SALVAGE project (CZ.02.01.01/00/22_008/0004644) supported by OP JAK, with cofinancing from the EU and the State Budget.
利益披露 Disclosure
P. Dzubak, None..
A. Srovnalova, None..
P. Polishchuk, None..
A. Siskova, None..
M. Mistrik, None..
M. Lacey, None..
M. Hajduch, None.