Chemical Space and Biological Target Network of Anti-Inflammatory Natural Products

Natural products (NPs) are a promising source of anti-inflammatory molecules for the development of drugs. Despite there being an abundance of reports of large numbers of NPs having bioactivity in preliminary cell-based assays of anti-inflammatory potential, their further optimization and exploration are limited by the lack of a comprehensive understanding of their effective scaffold structure or biological targets. To facilitate target-based studies of anti-inflammatory NPs, the details of 665 NPs reported to have anti-inflammatory activity were extracted from the literature and compiled into a data set we termed InflamNat. The physicochemical properties of the NPs were analyzed, and the distribution of their structures and scaffolds is presented. A compound-target network was constructed from data in the PubChem Bioassay database. The results demonstrated that, compared to natural anticancer compounds in the NPACT database, compounds from the InflamNat data set contained a comparable distribution of compound types but with a higher proportion satisfying Lipinski’s rule. The all-atom structures and scaffold of the compounds were diverse and barely convergent, with flavonoids and triterpenoids being the groups with the greatest abundance. The biological targets of the InflamNat compounds were identified as belonging to a variety of protein families that had varied function. Seventy-two percent of InflamNat compounds involved in the network were identified as having more than one biological target, highlighting the potential for multitarget anti-inflammatory drug development. In conclusion, anti-inflammatory NPs provide a good library for the screening of target-based leads or fragment-based drug design. Thus, elucidation of their biological targets is fundamental for either a specific single-target or multitarget drug development strategy. Meanwhile, a large proportion of the chemical space of anti-inflammatory NPs is still unexplored, with novel active scaffolds remaining to be discovered.

天然产物（Natural products, NPs）是药物开发领域极具潜力的抗炎分子来源。尽管已有大量研究表明，诸多NPs在抗炎潜力的初步细胞实验中展现出生物活性，但由于对其有效骨架结构与生物学靶点缺乏全面认知，其后续优化与探索仍受到极大限制。为推进抗炎NPs的靶点导向研究，我们从文献中提取了665种已被报道具有抗炎活性的NPs的详细信息，并将其汇编为一个名为InflamNat的数据集。我们对这些NPs的理化性质进行了分析，并展示了其结构与骨架的分布特征。基于PubChem生物测定数据库（PubChem Bioassay）的数据，我们构建了化合物-靶点网络。研究结果显示，与NPACT数据库中的天然抗癌化合物相比，InflamNat数据集中的化合物在化合物类型分布上较为相似，但符合Lipinski规则（Lipinski’s rule）的占比更高。这些化合物的全原子结构与骨架类型丰富多样，且几乎不存在趋同性，其中黄酮类与三萜类化合物的占比最高。InflamNat数据集中化合物的生物学靶点被归类为多个功能各异的蛋白质家族。该网络中72%的InflamNat化合物被证实具有多个生物学靶点，这凸显了多靶点抗炎药物的开发潜力。综上，抗炎NPs可为靶点导向的先导化合物筛选或基于片段的药物设计提供优质的化合物库。因此，阐明其生物学靶点对于单一靶点或多靶点的药物开发策略均具有基础性意义。与此同时，抗炎NPs的化学空间仍有大量区域未被探索，新型活性骨架有待进一步发掘。