Abstract:

The natural products terpenoids are widely distributed in animals (marine invertebrates), plants, microorganisms, with diverse molecular structures for bioactivities. A large number of terpenoids have been extracted directly from plants and microorganisms. However, traditional methods based on natural screening face challenges in discovering new terpenes due to the increasing number of known compounds at large quantities. The advent of next-generation sequencing and synthetic biology technologies marks the onset of the era of genome mining-driven natural product discovery, particularly in the exploration of new terpenoids. However, challenges persist in this regard, such as low efficiencies, interference of known compounds, and limited data throughput. In this review, we focus on recent advances in terpenoid discovery via microbial genome mining strategies, including the use of the precursor supplying microbial chassis (Escherichia coli, Saccharomyces cerevisiae, Aspergillus oryzae, Streptomyces albus, etc.), the microbial resources from extreme geographical environments, deep genome mining, and terpene mining platforms driven by artificial intelligence and automation techniques. To produce more terpenoids using heterologous hosts, multiple microbial chassis with enhanced precursor supply have been developed to improve their production yields and thus facilitate the discovery of structurally unique terpenoids. With the growing understanding of terpene biosynthesis machinery, the deep mining of terpenoid biosynthetic gene clusters and terpene synthases can effectively address issues related to repeated and irrelevant discoveries. Furthermore, the integration of artificial intelligence and automation platform with synthetic biology has ushered in the high-throughput intelligent discovery of terpenoids, which significantly improves the research and enables the discovery of numerous terpenoids with new structures. Finally, we address challenges and future directions for genome mining based terpenoid discovery. Driven by synthetic biology and artificial intelligence, a new chapter for the discovery of terpenoids and other natural products will open. We are looking forward to seeing more terpenoids to be developed as drugs and valuable chemicals in the future.

Key words: terpenoid natural products, terpene synthases, microbial chassis, genome mining, artificial intelligence, automated high-throughput platform

摘要：

萜类天然产物广泛分布于动物（包括海洋无脊椎动物）、植物、微生物中，具有复杂的化学结构和丰富的生物活性。人们通过从植物和微生物中直接分离提取的方式获得了大量萜类天然产物，然而随着越来越多化合物被发现，使用基于自然筛选的传统挖掘方式很难获得新的萜类天然产物。随着基因组测序技术和合成生物学使能技术的不断发展，我们进入了基因组挖掘驱动天然产物发现的时代，萜类天然产物的挖掘也进入了“井喷式”发现新阶段。针对基因组挖掘在微生物萜类天然产物发现方面的应用，本文综述了近年来使用的主要研究策略和最新研究进展，介绍了多种高效微生物底盘、基因组深度挖掘策略、人工智能与自动化平台等驱动的萜类化合物挖掘的最新研究进展，讨论了基因组挖掘萜类天然产物面临的挑战，展望了未来萜类化合物创新发现的发展趋势。通过在多种微生物中强化前体供应途径，人们打造了多个萜类化合物合成底盘，突破了异源合成萜类天然产物时“产量低”和“产物难获取”的瓶颈；针对萜类天然产物生物合成基因簇或萜类合酶进行深度挖掘，可以有效地解决“重复发现”和“集中度低”的难题；随着人工智能和自动化技术在合成生物学领域的发展和应用，萜类化合物的发现也进入了高通量智能发现时期，显著地改善了“研究通量低”的现状，高效获得了大量新结构萜类天然产物。在未来，更多萜类化合物将开发成药物、进入工业化生产应用，更多萜类“暗物质”会走进我们视野。

关键词: 萜类天然产物, 萜类合酶, 微生物底盘, 基因组挖掘, 人工智能, 自动化高通量平台