AKR1C3 是前列腺癌和其他癌症中上调的酶；除了调节激素合成外，这种酶还被认为在肿瘤的侵袭性和药物防御中发挥作用。我们在这里使用一种公正的方法来发现新的有效的 AKR1C3 抑制剂：通过基于人工智能的虚拟药物筛选，化合物 4 被确定为一种有效的选择性酶抑制剂，能够在 22RV1 前列腺癌细胞模型中将此活性转化为显着的抗增殖作用。与其他已知的 AKR1C3 抑制剂一样，化合物 4 可以显着提高阿比特龙（一种被批准用于治疗晚期前列腺癌的药物）的活性。化合物4在骨肉瘤细胞系中也显示出与阿霉素的协同作用；具体来说，该效果与 AKR1C3 表达相关。因此，在这项研究工作中，人工智能的使用使得能够鉴定出 AKR1C3 抑制剂的新结构及其增强化疗效果的潜力。© 2024 美国化学会。

AKR1C3 is an upregulated enzyme in prostate and other cancers; in addition to regulating hormone synthesis, this enzyme is thought to play a role in the aggressiveness of tumors and in the defense against drugs. We here used an unbiased method to discover new potent AKR1C3 inhibitors: through an AI-based virtual drug screen, compound 4 was identified as a potent and selective enzymatic inhibitor able to translate this activity into a pronounced antiproliferative effect in the 22RV1 prostate cancer cell model. As other known AKR1C3 inhibitors, compound 4 determined a significantly increased activity of abiraterone, a drug approved for advanced prostate cancer. Compound 4 also showed a synergic effect with doxorubicin in osteosarcoma cell lines; specifically, the effect is correlated with AKR1C3 expression. In this research work, therefore, the use of AI allowed the identification of a new structure as an AKR1C3 inhibitor and its potential to enhance the effect of chemotherapeutics.© 2024 American Chemical Society.