NF-κB is a central regulator of immunity and inflammation. It is suggested that the inflammatory response mediated by SARS-CoV-2 is predominated by NF-κB activation. Thus, NF-κB inhibition is considered a potential therapeutic strategy for COVID-19. The aim of this study was to identify potential anti-inflammation lead molecules that target NF-κB using a quantitative structure-activity relationships (QSAR) model of currently used and investigated anti-inflammatory drugs as the basis for screening. We applied an integrated approach by starting with the inflammation-based QSAR model to screen three libraries containing more than 220,000 drug-like molecules for the purpose of finding potential drugs that target the NF-κB/IκBα p50/p65 (RelA) complex. We also used QSAR models to rule out molecules that were predicted to be toxic. Among screening libraries, 382 molecules were selected as potentially nontoxic and were analyzed further by short and long molecular dynamics (MD) simulations and free energy calculations. We have discovered five hit ligands with highly predicted anti-inflammation activity and nearly no predicted toxicities which had strongly favorable protein-ligand interactions and conformational stability at the binding pocket compared to a known NF-κB inhibitor (procyanidin B2). We propose these hit molecules as potential NF-κB inhibitors which can be further investigated in pre-clinical studies against SARS-CoV-2 and may be used as a scaffold for chemical optimization and drug development efforts.

NF-κB 是免疫和炎症的中枢调节因子。这表明 SARS-CoV-2 介导的炎症反应以 NF-κB 激活为主。因此，抑制 NF-κB 被认为是 COVID-19 的潜在治疗策略。本研究的目的是使用当前使用和研究的抗炎药物的定量构效关系 (QSAR) 模型作为筛选基础，确定靶向 NF-κB 的潜在抗炎先导分子。我们采用综合方法，从基于炎症的 QSAR 模型开始筛选三个包含超过 220,000 种药物样分子的文库，以寻找靶向 NF-κB/IκBα p50/p65 (RelA) 复合物的潜在药物。我们还使用 QSAR 模型排除了预测有毒的分子。在筛选库中，382 个分子被选为潜在无毒分子，并通过短分子动力学 (MD) 模拟和自由能计算进一步分析。与已知的 NF-κB 抑制剂（原花青素 B2）相比，我们发现了五种具有高度预测抗炎活性且几乎没有预测毒性的命中配体，它们在结合口袋处具有非常有利的蛋白质-配体相互作用和构象稳定性。我们建议将这些命中分子作为潜在的 NF-κB 抑制剂，可以在针对 SARS-CoV-2 的临床前研究中进一步研究，并可用作化学优化和药物开发工作的支架。与已知的 NF-κB 抑制剂（原花青素 B2）相比，我们发现了五种具有高度预测抗炎活性且几乎没有预测毒性的命中配体，它们在结合口袋处具有非常有利的蛋白质-配体相互作用和构象稳定性。我们建议将这些命中分子作为潜在的 NF-κB 抑制剂，可以在针对 SARS-CoV-2 的临床前研究中进一步研究，并可用作化学优化和药物开发工作的支架。与已知的 NF-κB 抑制剂（原花青素 B2）相比，我们发现了五种具有高度预测抗炎活性且几乎没有预测毒性的命中配体，它们在结合口袋处具有非常有利的蛋白质-配体相互作用和构象稳定性。我们建议将这些命中分子作为潜在的 NF-κB 抑制剂，可以在针对 SARS-CoV-2 的临床前研究中进一步研究，并可用作化学优化和药物开发工作的支架。