Search for: All records

Creators/Authors contains: "Oh, Tae-Jin"

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. Accurate prediction of enzyme commission (EC) numbers remains a significant challenge in bioinformatics, limiting our understanding of enzyme functions and their roles in biological processes. This paper presents the integration and evaluation of Kolmogorov-Arnold networks (KANs), a new deep learning paradigm, in state-of-the-art models for EC prediction. KAN modules are incorporated into existing models to assess their impact on predictive performance. Additionally, we introduce a novel interpretation method designed for KANs to identify relevant input features, addressing a key limitation of these networks. Our evaluation demonstrates that KAN integration substantially improves predictive accuracy, with up to a 15.7% increase in micro-averaged F1 score and a 34.2% increase in macro-averaged F1 score. Moreover, our interpretation method enhances the trustworthiness of predictions and facilitates the discovery of motif sites within enzyme sequences. This approach provides insight into enzyme functionality and highlights potential new targets for research. The code is available at: https://github.com/datax-lab/kan_ecnumber. 
    more » « less
    Free, publicly-accessible full text available December 1, 2027
  2. Antarctic marine environments are extreme cold ecosystem and the microbes inhabiting them play a crucial role in ecosystem function. In this study, genomic and phenotypic analysis of Cryobacterium sp. PAMC25264, isolated from seawater of Barton, South Shetland Island, Antarctica Peninsula was performed. Comparative genomic analysis with related strains focused on nitrogen metabolic enzymes, carbohydrate-active enzymes (CAZymes) and stress-related proteins to elucidate the functional diversity and survival strategy of the strains in these harsh environments. Type Strain Genome Server (TYGS) analysis suggest that this strain may represent a distinct phylogenetic lineage within the genus Cryobacterium. Nitrogen metabolic enzymes including respiratory nitrate reductase (α, β, γ, and δ subunits), assimilatory nitrate reductase (NasA), nitrate/nitrite transporter (NarK/U), assimilatory nitrite reductase (NirB and NirD), and glutamate synthase (GltB and GltD) were identified from genomic analysis. dbCAN3 revealed the presence of a total of 125 CAZymes. Among them 66 glycoside hydrolases (GHs), 36 glycosyltransferases (GTs), 6 carbohydrate esterases (CEs), 15 carbohydrate-binding modules (CBMs), and 2 polysaccharide lyase (PL). In addition, proteins related to oxidative stress (superoxide dismutase, catalase, and thiol peroxidase), temperature stress (Csp and Hsp), and heavy metal(loid)s (HMMs) resistance were identified. Furthermore, the strain exhibited nitrate reduction in aerobic condition, resulting in nitrite accumulation, starch hydrolysis, and cellulose hydrolysis ability at different temperatures (4 °C, 10 °C, 15 °C, and 25 °C). Overall, these genomic and phenotypic traits indicate metabolic flexibility and functional diversity, supporting its survival and potential ecological role in extreme cold marine ecosystems. 
    more » « less
    Free, publicly-accessible full text available December 1, 2027
  3. Accurate and trustworthy prediction of Enzyme Commission (EC) numbers is critical for understanding enzyme functions and their roles in biological processes. Despite the success of recently proposed deep learning-based models, there remain limitations, such as low performance in underrepresented EC numbers, lack of learning strategy with incomplete annotations, and limited interpretability. To address these challenges, we propose a hierarchical interpretable transformer model, HIT-EC, for trustworthy EC number prediction. HIT-EC employs a four-level transformer architecture that aligns with the hierarchical structure of EC numbers, and leverages both local and global dependencies within protein sequences for this multi-label classification task. We also propose a learning strategy to handle samples associated with incomplete EC numbers. HIT-EC, as an evidential deep learning model, produces trustworthy predictions by providing domain-specific evidence through a biologically meaningful interpretation scheme. The predictive performance of HIT-EC is assessed by multiple experiments: a cross-validation with a large dataset, a validation with external data, and a species-based performance evaluation. HIT-EC shows statistically significant improvement in predictive performance when compared to the current state-of-the-art benchmark models. HIT-EC’s robust interpretability is further validated by identifying well-known conserved motifs and functional regions. HIT-EC is a robust, interpretable, and reliable solution for EC number prediction, with significant implications for enzymology, drug discovery, and metabolic engineering. 
    more » « less
    Free, publicly-accessible full text available December 1, 2027
  4. Bidirectional acetylation and deacetylation of small molecules by acetyltransferases (ATs) remain poorly understood. In this study, we present the structural, functional, and computational characterization of a chloramphenicol O-AT from Bacillus sp. PAMC22265 (AT65) that exhibits bidirectional catalytic activity toward the regiospecific C21-acetylation and deacetylation of steroid substrates. The crystal structure of AT65 was determined at 2.40 Å resolution (PDB ID: 24ZY), revealing a trimeric architecture in which the putative active-sites located at the interfaces between adjacent subunits. In vitro kinetic analyses of both acetylation and deacetylation reactions demonstrated substrate-dependent catalytic efficiencies, with comparable activity under the experimental conditions. Molecular docking, 100 ns molecular dynamics simulations, hydrogen-bond analysis, and MM-PBSA free-energy calculations supported stable substrate binding within the active site and suggested that His187 and Asp191 may contribute to substrate recognition and catalysis. Consistent with this observation, substitution of His187 with alanine abolished detectable enzymatic activity, highlighting its functional importance. Whole-cell biotransformation further demonstrated the selective production of C21-acetylated steroid derivatives, some of which exhibited preliminary antiproliferative activity against murine breast cancer cell lines. Collectively, these findings provide structural, biochemical, and computational insights into the bidirectional catalytic properties of AT65 and establish a foundation for further mechanistic studies and the development of selective enzymatic strategies for steroid modification. 
    more » « less
    Free, publicly-accessible full text available November 1, 2027
  5. Cytochrome P450 (CYPs) catalyze regioselective oxidation reactions, yet the structural determinants governing regioselectivity and catalytic efficiency remain poorly understood. Here, we characterize CYP154C10 as a versatile steroid hydroxylase exhibiting predominant C2 regioselectivity and expanded substrate scope relative to previously described CYP154C enzymes. Notably, CYP154C10 supports steroid conversion using both chemical (PIDA) and NADPH-dependent (Fdx/FdR) systems, while maintaining consistent product distribution across both oxidation systems. Structure-guided mutagenesis, kinetic analysis, and molecular dynamics (MD) simulations together provide mechanistic insights into the functional separation between active-site and access-channel residues. Mutation of the active-site residue L93 to phenylalanine (L93F) induces substrate-dependent changes in regioselectivity by increasing conformational flexibility and enabling multiple catalytically competent substrate orientations. In contrast, mutations at the access-channel position (M199F/Y) preserve regioselectivity but modulate catalytic efficiency by perturbing substrate positioning and local hydration dynamics. Comparative structural and dynamic analyses further suggest that active-site architecture governs substrate orientation, whereas access-channel residues influence substrate delivery and catalytic efficiency in CYP154C10. Together, these findings provide a mechanistic framework for the rational engineering of selective P450 biocatalysts with tunable regioselectivity and catalytic efficiency 
    more » « less
    Free, publicly-accessible full text available September 1, 2027
  6. Laccases are multicopper oxidases that offer environmentally friendly routes for polymer synthesis and modification under mild aqueous conditions; however, their application in alkaline systems and in the development of functional material remains limited. In this study, a recombinant bacterial laccase (rBCLac) from Antarctic Bacillus sp. PAMC28748 was employed as a green biocatalyst for the oxidative polymerization of catechol and its simultaneous attachment onto chitosan in fully aqueous media, without the use of organic solvents or harsh reagents. The enzymatic oxidation generated reactive o-quinone intermediates, which subsequently underwent non-enzymatic coupling to form catechol-functionalized chitosan. The modified biopolymer exhibited enhanced swelling behavior, controlled methylene blue release, and increased iron- chelation capacity. Optimal polymerization was achieved at pH 8–9 and 20 °C. Kinetic analysis and molecular dynamics simulations confirmed that the enzyme maintained sufficient catalytic activity and structural stability under alkaline aque- ous conditions. The enzymatically generated poly(catechol) was further utilized for surface functionalization of Fe₃O₄ nanoparticles, yielding well-dispersed composites with enhanced surface functionality. FTIR analysis revealed broadening of the O-H band and the formation of C-O-C bonds, confirming catechol polymerization, while Fe-O vibrations verified preservation of the Fe₃O₄ core after modification. These materials exhibited rapid adsorption behavior and pronounced selectivity toward congo red over methylene blue, which can be attributed to synergistic π–π interactions, hydrogen bond- ing, and electrostatic interactions. The poly(catechol)-Fe3O4 composite showed a high adsorption capacity for Congo red (qm = 384.62 mg g− 1), whereas adsorption of methylene blue remined limited. Reusability studies demonstrated partial retention of adsorption performance over multiple cycles. Overall, this work presents a sustainable, enzyme-mediated strategy for polysaccharide modification and functional polymer development under environmentally relevant aqueous conditions, with potential applications in selective dye remediation. 
    more » « less
    Free, publicly-accessible full text available July 1, 2027
  7. Flavonoids and stilbenes are phenylpropanoid-derived secondary metabolites found in plants, fungi, and microorganisms. They serve important physiological functions and offer a range of pharmacological benefits, including antioxidant, anti-inflammatory, anticancer, and antimicrobial effects. However, many native flavonoids and stilbenes have limited pharmacology and industrial use due to poor bioavailability and low metabolic stability. O-methylation is a key post-biosynthetic modification that improves the physicochemical and pharmacokinetic properties of these compounds. This process, catalyzed by S-adenosyl-L-methionine [SAM]-dependent O-methyltransferases [OMTs], increases lipophilicity, chemical stability, and membrane permeability, often resulting in greater bioactivity. As a result, methylated derivatives such as sakuranetin, acacetin, isorhamnetin, and pterostilbene frequently show enhanced therapeutic efficacy, even though they are less abundant in nature. Advances in metabolic engineering and synthetic biology have enabled efficient biosynthesis of methylated flavonoids and stilbenes in microbial hosts such as Escherichia coli, Saccharomyces cerevisiae, and Streptomyces species. Strategies including pathway optimization, SAM regeneration, enzyme engineering, and CRISPR-based genome editing have significantly improved production yields and regioselectivity, with some systems surpassing native producers. Despite this progress, challenges remain in enzyme specificity, metabolic bottlenecks, and downstream processing. This review summarizes recent developments in understanding methylation biosynthetic pathways, OMT diversity, and microbial engineering strategies. It also highlights the pharmacological significance of methylation and the potential of engineered microbial platforms for sustainable, scalable production of methylated flavonoids and stilbenes for pharmaceutical and nutraceutical use. 
    more » « less
    Free, publicly-accessible full text available July 1, 2027
  8. Newton, Irene L G (Ed.)
    ABSTRACT Here, we report the complete genome sequence ofSphingobacteriumsp. PAMC25397, isolated from Ny-Ålesund, Spitsbergen, Norway. The genome is 5.2 Mb in length with a GC content of 37%. This genome harbors biodegradation-related genes involved in benzoate degradation. 
    more » « less
    Free, publicly-accessible full text available October 8, 2027
  9. Flavonoid O-methyltransferases (OMTs) catalyze the methylation of flavonoid hydroxyl groups, enhancing structural diversity and biological activity. In this study, we identified and characterized a novel Class I flavonoid OMT from the Antarctic bacterium Pedobacter sp. PAMC26386. Despite originating from a cold-adapted organism, the enzyme exhibited high catalytic activity at 55 °C and a strong preference for Co²⁺ as a cofactor. Sequence and phylogenetic analyses confirmed its classification as a flavonoid-specific OMT and revealed conserved motifs for S-adenosyl-L-methionine (SAM) binding and metal coordination. The enzyme accepted a wide range of flavonoid substrates, with quercetin and fisetin showing the highest activities. Kinetic analysis indicated greater substrate affinity for quercetin (Km = 36.55 µM) than for fisetin (Km = 49.41 µM). Whole-cell biotransformation using recombinant Escherichia coli C41 co-expressing the OMT and metK enabled efficient intracellular methylation, yielding 62.5 mg L⁻¹ of methylated quercetin and 55.5 mg L⁻¹ of 3′-O-methyl fisetin. The predicted methylation site at the 3′-hydroxyl of fisetin, determined by molecular docking, was confirmed by NMR spectroscopy. Notably, the previously reported 3′-O-methyl fisetin showed enhanced in vitro anticancer activity against mouse breast cancer cells and selectively improved antimycobacterial activity against Mycobacterium tuberculosis compared to the parent compound. To our knowledge, this study is among the first to report the enzymatic production of 3′-O-methyl fisetin using a polar microbial OMT that is optimally active at elevated temperatures in the presence of Co²⁺. The increased bioactivity compared to the parent compounds highlights the potential of this OMT as a biocatalyst for the sustainable production of pharmaceutically relevant methylated flavonoids. 
    more » « less
    Free, publicly-accessible full text available September 1, 2027
  10. Newton, Irene L G (Ed.)
    ABSTRACT The present study presents the complete genomic sequence of a novelBacillus zhangzhouensisPAMC22242 isolated from biofilm at Kongsfjorden dock, Svalbard, Norway. The genome size is 3,507,368, with a 41.5% GC content, including genes putatively associated with non-ribosomal peptide synthetase biosynthesis and metal-related transport systems. 
    more » « less
    Free, publicly-accessible full text available September 10, 2027