Predictive Modeling of PROTAC Cell Permeability with Machine Learning

Poongavanam, Vasanthanathan; Kölling, Florian; Giese, Anja; Göller, Andreas H.; Lehmann, Lutz; Meibom, Daniel; Kihlberg, Jan

doi:10.1021/acsomega.2c07717

Cited by 16 publications

(9 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It was likely that a PROTAC with PEG-type linkers that evoked folded conformation would be transported through passive diffusion to a greater extent than a PROTAC with alkyl linkers that evoked elongated anti-conformation in in vitro permeation assay calculating passive permeability using Caco-2 cell monolayers. Nevertheless, PROTACs with PEG-type linkers and PROTACs with alkyl linkers demonstrated permeation through the artificial membrane, to the same degree, in a parallel artificial membrane permeability assay (PAMPA) [ 16 ]. The prediction of the permeability will be improved by machine learning methods through physical property factors such as molecular weight, hetero atoms, H-bond donors, H-bond acceptors, the number of rotatable bonds (NRotB), logD, and the polar surface area (PSA).…”

Section: Discussionmentioning

confidence: 99%

Proteolysis-Targeting Chimera (PROTAC) Delivery into the Brain across the Blood-Brain Barrier

Tashima

2023

Antibodies

View full text Add to dashboard Cite

Drug development for neurodegenerative diseases such as Alzheimer’s disease, Parkinson’s disease, and Huntington’s disease has challenging difficulties due to the pharmacokinetic impermeability based on the blood-brain barrier (BBB) as well as the blurriness of pharmacodynamic targets based on their unclarified pathogenesis and complicated progression mechanisms. Thus, in order to produce innovative central nervous system (CNS) agents for patients suffering from CNS diseases, effective, selective delivery of CNS agents into the brain across the BBB should be developed. Currently, proteolysis-targeting chimeras (PROTACs) attract rising attention as a new modality to degrade arbitrary intracellular proteins by the ubiquitin-proteasome system. The internalizations of peptide-based PROTACs by cell-penetrating peptides and that of small molecule-based PROTACs through passive diffusion lack cell selectivity. Therefore, these approaches may bring off-target side effects due to wrong distribution. Furthermore, efflux transporters such as multiple drug resistance 1 (MDR1) expressed at the BBB might interrupt the entry of small molecule-based PROTACs into the brain. Nonetheless, intelligent delivery using machinery systems to absorb the nutrition into the brain for homeostasis, such as carrier-mediated transport (CMT) or receptor-mediated transcytosis (RMT), can be established. PROTACs with N-containing groups that are recognized by the proton-coupled organic cation antiporter might cross the BBB through CMT. PROTAC-antibody conjugates (PACs) might cross the BBB through RMT. Subsequently, such small molecule-based PROTACs released in the brain interstitial fluid would be transported into cells such as neurons through passive diffusion and then demonstrate arbitrary protein degradation. In this review, I introduce the potential and advantages of PROTAC delivery into the brain across the BBB through CMT or RMT using PACs in a non-invasive way.

show abstract

Section: Discussionmentioning

confidence: 99%

Proteolysis-Targeting Chimera (PROTAC) Delivery into the Brain across the Blood-Brain Barrier

Tashima

2023

Antibodies

View full text Add to dashboard Cite

show abstract

“…[125] Moreover, machine-learning algorithms have been used to assess the intrinsic degradability of POIs, [137] and PROTAC permeability. [138][139]…”

Section: Machine Learning-based Pipelinesmentioning

confidence: 99%

“…Deep learning has also been applied in methods such as DeepPROTACs to detect the capacity of PROTACs to induce degradation (Figure 4). [125] Moreover, machine‐learning algorithms have been used to assess the intrinsic degradability of POIs, [137] and PROTAC permeability [138–139] …”

Section: Computational Modelingmentioning

confidence: 99%

Biophysical and Computational Approaches to Study Ternary Complexes: A ‘Cooperative Relationship’ to Rationalize Targeted Protein Degradation

2023

View full text Add to dashboard Cite

Degraders have illustrated that compound-induced proximity to E3 ubiquitin ligases can prompt the ubiquitination and degradation of disease-relevant proteins. Hence, this pharmacology is becoming a promising alternative and complement to available therapeutic interventions (e. g., inhibitors). Degraders rely on protein binding instead of inhibition and, hence, they hold the promise to broaden the druggable proteome. Biophysical and structural biology approaches have been the cornerstone of understanding and rationalizing degraderinduced ternary complex formation. Computational models have now started to harness the experimental data from these approaches with the aim to identify and rationally help design new degraders. This review outlines the current experimental and computational strategies used to study ternary complex formation and degradation and highlights the importance of effective crosstalk between these approaches in the advancement of the targeted protein degradation (TPD) field. As our understanding of the molecular features that govern druginduced interactions grows, faster optimizations and superior therapeutic innovations for TPD and other proximity-inducing modalities are sure to follow.

show abstract

“…The application of ML methods for the prediction of membrane permeability has also sharply increased in recent years. Some examples include neural networks, support-vector machines, and random forest classifiers trained on molecular structures and physicochemical properties of small and large compounds. − To predict degrader permeability, Poongavanam et al have tested several binary classifiers using descriptors that represent molecular size, shape, and chemical functionalities. While predictions were good (accuracy >80%) in some cases (e.g., VHL-recruiting degraders), the classifiers performed poorly on cereblon-recruiting degraders due to imbalances in the corresponding training data.…”

Section: Modeling the Tpd Processmentioning

confidence: 99%

Targeted Protein Degradation: Advances, Challenges, and Prospects for Computational Methods

Mostofian,

Martin,

Razavi

et al. 2023

J. Chem. Inf. Model.

View full text Add to dashboard Cite

The therapeutic approach of targeted protein degradation (TPD) is gaining momentum due to its potentially superior effects compared with protein inhibition. Recent advancements in the biotech and pharmaceutical sectors have led to the development of compounds that are currently in human trials, with some showing promising clinical results. However, the use of computational tools in TPD is still limited, as it has distinct characteristics compared with traditional computational drug design methods. TPD involves creating a ternary structure (protein−degrader−ligase) responsible for the biological function, such as ubiquitination and subsequent proteasomal degradation, which depends on the spatial orientation of the protein of interest (POI) relative to E2-loaded ubiquitin. Modeling this structure necessitates a unique blend of tools initially developed for small molecules (e.g., docking) and biologics (e.g., protein−protein interaction modeling). Additionally, degrader molecules, particularly heterobifunctional degraders, are generally larger than conventional small molecule drugs, leading to challenges in determining drug-like properties like solubility and permeability. Furthermore, the catalytic nature of TPD makes occupancy-based modeling insufficient. TPD consists of multiple interconnected yet distinct steps, such as POI binding, E3 ligase binding, ternary structure interactions, ubiquitination, and degradation, along with traditional small molecule properties. A comprehensive set of tools is needed to address the dynamic nature of the induced proximity ternary complex and its implications for ubiquitination. In this Perspective, we discuss the current state of computational tools for TPD. We start by describing the series of steps involved in the degradation process and the experimental methods used to characterize them. Then, we delve into a detailed analysis of the computational tools employed in TPD. We also present an integrative approach that has proven successful for degrader design and its impact on project decisions. Finally, we examine the future prospects of computational methods in TPD and the areas with the greatest potential for impact.

show abstract

Predictive Modeling of PROTAC Cell Permeability with Machine Learning

Cited by 16 publications

References 41 publications

Proteolysis-Targeting Chimera (PROTAC) Delivery into the Brain across the Blood-Brain Barrier

Proteolysis-Targeting Chimera (PROTAC) Delivery into the Brain across the Blood-Brain Barrier

Biophysical and Computational Approaches to Study Ternary Complexes: A ‘Cooperative Relationship’ to Rationalize Targeted Protein Degradation

Targeted Protein Degradation: Advances, Challenges, and Prospects for Computational Methods

Contact Info

Product

Resources

About