In silico identification of putative druggable pockets in PRL3, a significant oncology target

Abstract

Protein tyrosine phosphatases (PTP) have emerged as targets in diseases characterized by aberrant phosphorylations such as cancers. The activity of the phosphatase of regenerating liver 3, PRL3, has been linked to several oncogenic and metastatic pathways, particularly in breast, ovarian, colorectal, and blood cancers. Development of small molecules that directly target PRL3, however, has been challenging. This is partly due to the lack of structural information on how PRL3 interacts with its inhibitors. Here, computational methods are used to bridge this gap by evaluating the druggability of PRL3. In particular, web-based pocket prediction tools, DoGSite3 and FTMap, were used to identify binding pockets using structures of PRL3 currently available in the Protein Data Bank. Druggability assessment by molecular dynamics simulations with probes was also performed to validate these results and to predict the strength of binding in the identified pockets. While several druggable pockets were identified, those in the closed conformation show more promise given their volume and depth. These two pockets flank the active site loops and roughly correspond to pockets predicted by molecular docking in previous papers. Notably, druggability simulations predict the possibility of low nanomolar affinity inhibitors in these sites implying the potential to identify highly potent small molecule inhibitors for PRL3. Putative pockets identified here can be leveraged for high-throughput virtual screening to further accelerate the drug discovery against PRL3 and development of PRL3-directed therapeutics.