生物信息学
计算生物学
基因组
对接(动物)
同源建模
异源表达
生物
序列比对
活动站点
化学
异源的
同源(生物学)
位阻效应
生物发生
结构母题
聚合物
基因
甲烷单加氧酶
立体化学
结构线形
作者
Oksana P. Kovalenko,Gulsima Dilek Usluer,İlker Karacan,Kinga Plasa,Anna Kasprzycka,Michał Chyski,Selda Murat Hocaoğlu,Artur Góra,Agata Raczyńska
摘要
ABSTRACT Marine metagenomes represent a vast reservoir of biocatalysts with untapped industrial potential. In this study, we employed a sequence‐structure‐guided workflow to mine a metagenomic library of over 15.3 million sequences from the Marmara Sea for novel polyurethane (PUR) degrading enzymes. Using Hidden Markov Model (HMM) profiles based on known polymer‐degrading cutinases (HiC, Tf Cut2, LCC) and urethanases (UMG‐SP1–3), coupled with AlphaFold2 structural filtration, 13 candidate hydrolases were prioritized. Following heterologous expression in E. coli , six candidates were successfully produced and screened for activity toward selected small substrates and Impranil DLN. Five candidates demonstrated robust esterolytic activity toward 4‐nitrophenyl butyrate, but lacked activity against urethane bonds or the polyester‐polyurethane (PUR) dispersion Impranil DLN. Strikingly, a retrospective in silico docking analysis using a library of 414 PUR oligomers nevertheless strongly predicted productive polymer binding. Structural analysis indicates that this limitation stems from surface topography, including prominent gating loops and sterically restricted active sites that likely restrict bulky polymer access. Ultimately, this study introduces five novel functional marine esterases, including highly divergent archaeal homologs, and highlights the critical limitations of using in silico docking and simple sequence homology to predict macroscopic polymer degradation.
科研通智能强力驱动
Strongly Powered by AbleSci AI