End-to-end sequence-structure-function meta-learning predicts genome-wide chemical-protein interactions for dark proteins.

Systematically discovering protein-ligand interactions across the entire human and pathogen genomes is critical in chemical genomics, protein function prediction, drug discovery, and many other areas. However, more than 90% of gene families remain "dark"-i.e., their small-molecule ligands are undiscovered due to experimental limitations or human/historical biases. Existing computational approaches typically fail when the dark protein differs from those with known ligands. To address this challenge, we have developed a deep learning framework, called PortalCG, which consists of four novel components: (i) a 3-dimensional ligand binding site enhanced sequence pre-training strategy to encode the evolutionary links between ligand-binding sites across gene families; (ii) an end-to-end pretraining-fine-tuning strategy to reduce the impact of inaccuracy of predicted structures on function predictions by recognizing the sequence-structure-function paradigm; (iii) a new out-of-cluster meta-learning algorithm that extracts and accumulates information learned from predicting ligands of distinct gene families (meta-data) and applies the meta-data to a dark gene family; and (iv) a stress model selection step, using different gene families in the test data from those in the training and development data sets to facilitate model deployment in a real-world scenario. In extensive and rigorous benchmark experiments, PortalCG considerably outperformed state-of-the-art techniques of machine learning and protein-ligand docking when applied to dark gene families, and demonstrated its generalization power for target identifications and compound screenings under out-of-distribution (OOD) scenarios. Furthermore, in an external validation for the multi-target compound screening, the performance of PortalCG surpassed the rational design from medicinal chemists. Our results also suggest that a differentiable sequence-structure-function deep learning framework, where protein structural information serves as an intermediate layer, could be superior to conventional methodology where predicted protein structures were used for the compound screening. We applied PortalCG to two case studies to exemplify its potential in drug discovery: designing selective dual-antagonists of dopamine receptors for the treatment of opioid use disorder (OUD), and illuminating the understudied human genome for target diseases that do not yet have effective and safe therapeutics. Our results suggested that PortalCG is a viable solution to the OOD problem in exploring understudied regions of protein functional space.

Fluoroalkoxylated C-3 and C-9 (S)-12-bromostepholidine analogues with D1R antagonist activity.

To illuminate the tolerance of fluoroalkoxylated groups at the C-3 and C-9 positions of tetrahydroprotoberberines (THPBs) on D1R activity, C-3 and C-9 fluoroalkoxylated analogues of (S)-12-bromostepholidine were prepared and evaluated. All compounds examined were D1R antagonists as measured by a cAMP assay. Our structure-activity studies herein indicate that the C-3 position tolerates a 1,1-difluoroethoxy substituent for D1R antagonist activity. Compound 13a was the most potent cAMP-based D1R antagonist identified and was also found to antagonize ß-arrestin translocation in a TANGO assay. Affinity assessments at other dopamine receptors revealed that 13a is selective for D1R and unlike other naturally-occurring THPBs such as (S)-stepholidine, lacks D2R affinity. In preliminary biopharmaceutical assays, excellent BBB permeation was observed for 13a. Further pharmacological studies are warranted on (S)-stepholidine congeners to harvest their potential as a source of novel, druggable D1R-targeted agents.

Natural Product-Inspired Dopamine Receptor Ligands.

Due to their evolutionary bias as ligands for biologically relevant drug targets, natural products offer a unique opportunity as lead compounds in drug discovery. Given the involvement of dopamine receptors in various physiological and behavioral functions, they are linked to numerous diseases and disorders such as Parkinson's disease, schizophrenia, and substance use disorders. Consequently, ligands targeting dopamine receptors hold considerable therapeutic and investigative promise. As this perspective will highlight, dopamine receptor targeting natural products play a pivotal role as scaffolds with unique and beneficial pharmacological properties, allowing for natural product-inspired drug design and lead optimization. As such, dopamine receptor targeting natural products still have untapped potential to aid in the treatment of disorders and diseases related to central nervous system (CNS) and peripheral nervous system (PNS) dysfunction.

Discovery of Selective Dopamine Receptor Ligands Derived from (-)-Stepholidine via C-3 Alkoxylation and C-3/C-9 Dialkoxylation.

We evaluated C-3 alkoxylated and C-3/C-9 dialkoxylated (-)-stepholidine analogues to probe the tolerance at the C-3 and C-9 positions of the tetrahydroprotoberberine (THPB) template toward affinity for dopamine receptors. A C-9 ethoxyl substituent appears optimal for D1R affinity since high D1R affinities were observed for compounds that contain an ethyl group at C-9, with larger C-9 substituents tending to decrease D1R affinity. A number of novel ligands were identified, such as compounds 12a and 12b, with nanomolar affinities for D1R and no affinity for either D2R or D3R, with compound 12a being identified as a D1R antagonist for both G-protein- and ß-arrestin-based signaling. Compound 23b was identified as the most potent and selective D3R ligand containing a THPB template to date and functions as an antagonist for both G-protein- and ß-arrestin-based signaling. Molecular docking and molecular dynamics studies validated the D1R and D3R affinity and selectivity of 12a, 12b, and 23b.