Tweet
Sci Rep
. 2026 Jul 2.
doi: 10.1038/s41598-026-60085-w. Online ahead of print.
Computer-assisted design of arylethylbenzamides as predicted nanomolar inhibitors of papain-like cysteine protease of SARS-CoV-2
Lukas Kerti 1 , Vladimir Frecer 2
Affiliations
The pandemic of the new coronavirus SARS-CoV-2, which causes the severe acute respiratory syndrome COVID-19, represents a long-term threat to the health of the human population. Therefore, the continuous development of new small-molecule antivirals remains essential to effectively address current infections and ensure preparedness for future pandemic threats. The objective of the study was to computationally design, optimize, and prioritize (R)-arylethylbenzamide (AREB) analogs as potential inhibitors of SARS-CoV-2 papain-like protease (PLpro), a conserved antiviral target involved in viral polyprotein cleavage and host immune evasion signaling. We hypothesize that receptor-structure-guided expansion of the AREB scaffold in four distinct regions - reaching towards the BL2-groove, the Cys111 catalytic site, the electrostatic Glu167/Asp164 region, and the Val70Ub pocket adjacent to ubiquitin - can produce candidates with predicted low-nanomolar inhibitory potency while maintaining the drug-like properties of the new analogs. Multiple published PLpro-inhibitor crystal structures1-3 were used as templates after molecular mechanics refinement. A QSAR model was built from 51 published1,3 AREB inhibitors by correlating the calculated relative enzyme-inhibitor interaction energies (ΔΔEint,MM) with the experimentally determined pIC50 values using linear regression and extensive 5-fold cross-validation. Four virtual combinatorial libraries were designed and enumerated for substitution sites (R1 - R4) of a common scaffold and screened by extra-precision docking followed by MM-GB/SA rescoring. The predicted IC50pre values were used for the potency ranking of the new AREB analogs. Predicted ADME-related descriptors guided iterative virtual library focusing and filtering. For selected leads, QM/MM interaction energies (ΔΔEint,QM/MM,aq) and TIP4P explicit-solvent 200 ns MD simulations were used to assess the stability of the binding mode of promising new PLpro inhibitors. The QSAR model demonstrated strong internal validity and predictability, enabling the prioritization of designed analogs. After ADME-based filtering, three final drug candidates with predicted IC50pre = 5.2-5.3 nM and an estimated 18-fold increase in potency compared to the most potent reference inhibitor considered (IC50exp = 94 nM) were prioritized. The top candidates preserved the hallmark BL2-loop closure binding mode while extending stabilizing interaction networks to additional subsites targeted by the individual R-group design strategy. MD trajectories supported stable pocket occupancy over a 200 ns simulation and sustained key hydrogen bonds and hydrophobic contacts. An integrated QSAR and structure-guided computational workflow prioritized synthetically available drug-like AREB analogs as putative SARS-CoV-2 PLpro inhibitors with predicted low nanomolar potency. Although the presented findings are mostly computational and require experimental validation, the identified lead compounds represent promising candidates for further antiviral agent development.
Keywords: Computer-aided drug design; Molecular dynamics; Papain-like protease of SARS-CoV-2; QM/MM calculations; QSAR.
. 2026 Jul 2.
doi: 10.1038/s41598-026-60085-w. Online ahead of print.
Computer-assisted design of arylethylbenzamides as predicted nanomolar inhibitors of papain-like cysteine protease of SARS-CoV-2
Lukas Kerti 1 , Vladimir Frecer 2
Affiliations
- PMID: 42393217
- DOI: 10.1038/s41598-026-60085-w
The pandemic of the new coronavirus SARS-CoV-2, which causes the severe acute respiratory syndrome COVID-19, represents a long-term threat to the health of the human population. Therefore, the continuous development of new small-molecule antivirals remains essential to effectively address current infections and ensure preparedness for future pandemic threats. The objective of the study was to computationally design, optimize, and prioritize (R)-arylethylbenzamide (AREB) analogs as potential inhibitors of SARS-CoV-2 papain-like protease (PLpro), a conserved antiviral target involved in viral polyprotein cleavage and host immune evasion signaling. We hypothesize that receptor-structure-guided expansion of the AREB scaffold in four distinct regions - reaching towards the BL2-groove, the Cys111 catalytic site, the electrostatic Glu167/Asp164 region, and the Val70Ub pocket adjacent to ubiquitin - can produce candidates with predicted low-nanomolar inhibitory potency while maintaining the drug-like properties of the new analogs. Multiple published PLpro-inhibitor crystal structures1-3 were used as templates after molecular mechanics refinement. A QSAR model was built from 51 published1,3 AREB inhibitors by correlating the calculated relative enzyme-inhibitor interaction energies (ΔΔEint,MM) with the experimentally determined pIC50 values using linear regression and extensive 5-fold cross-validation. Four virtual combinatorial libraries were designed and enumerated for substitution sites (R1 - R4) of a common scaffold and screened by extra-precision docking followed by MM-GB/SA rescoring. The predicted IC50pre values were used for the potency ranking of the new AREB analogs. Predicted ADME-related descriptors guided iterative virtual library focusing and filtering. For selected leads, QM/MM interaction energies (ΔΔEint,QM/MM,aq) and TIP4P explicit-solvent 200 ns MD simulations were used to assess the stability of the binding mode of promising new PLpro inhibitors. The QSAR model demonstrated strong internal validity and predictability, enabling the prioritization of designed analogs. After ADME-based filtering, three final drug candidates with predicted IC50pre = 5.2-5.3 nM and an estimated 18-fold increase in potency compared to the most potent reference inhibitor considered (IC50exp = 94 nM) were prioritized. The top candidates preserved the hallmark BL2-loop closure binding mode while extending stabilizing interaction networks to additional subsites targeted by the individual R-group design strategy. MD trajectories supported stable pocket occupancy over a 200 ns simulation and sustained key hydrogen bonds and hydrophobic contacts. An integrated QSAR and structure-guided computational workflow prioritized synthetically available drug-like AREB analogs as putative SARS-CoV-2 PLpro inhibitors with predicted low nanomolar potency. Although the presented findings are mostly computational and require experimental validation, the identified lead compounds represent promising candidates for further antiviral agent development.
Keywords: Computer-aided drug design; Molecular dynamics; Papain-like protease of SARS-CoV-2; QM/MM calculations; QSAR.