Abstract
Cytochrome P450 3A4 (CYP3A4) is a key mediator in xenobiotic metabolism and drug-drug interactions (DDI), developing orally active fluorogenic substrates for sensing and imaging of a target enzyme in biological systems remains challenging. Here, an artificial intelligence (AI)-driven strategy is used to construct a highly specific and orally active fluorogenic substrate for imaging CYP3A4 in complex biological systems. After the fusion of an AI-selected drug-like fragment with a CYP3A4-preferred fluorophore, three candidates are designed and synthesized. Among all evaluated candidates, NFa exhibits excellent isoform-specificity, ultra-high sensitivity, outstanding spatial resolution, favorable safety profiles, and acceptable oral bioavailability. Specifically, NFa excels at functional in situ imaging of CYP3A4 in living systems with exceptional endoplasmic reticulum (ER)-colocalization performance and high imaging resolution, while this agent can also replace hCYP3A4 drug-substrates for high-throughput screening of CYP3A4 inhibitors and for assessing DDI potential in vivo. With the help of NFa, a novel CYP3A4 inhibitor (D13) was discovered, and its anti-CYP3A4 effects are assessed in live cells, ex vivo and in vivo. Collectively, an AI-powered strategy is adapted for developing highly-specific and drug-like fluorogenic substrates, resulting in the first orally available tool (NFa) for sensing and imaging CYP3A4 activities, which facilitates CYP3A4-associated fundamental investigations and the drug discovery process.