Biosynthesis—a process utilizing biological systems to synthesize chemical compounds—has emerged as a revolutionary solution to 21st-century challenges due to its environmental sustainability, scalability, and high stereoselectivity and regioselectivity. Recent advancements in artificial intelligence (AI) are accelerating biosynthesis by enabling intelligent design, construction, and optimization of enzymatic reactions and biological systems. We first introduce the molecular retrosynthesis route planning in biochemical pathway design, including single-step retrosynthesis algorithms and AI-based chemical retrosynthesis route design tools. We highlight the advantages and challenges of large language models in addressing the sparsity of chemical data. Furthermore, we review enzyme discovery methods based on sequence and structure alignment techniques. Breakthroughs in AI-based structural prediction methods are expected to significantly improve the accuracy of enzyme discovery. We also summarize methods for de novo enzyme generation for nonnatural or orphan reactions, focusing on AI-based enzyme functional annotation and enzyme discovery techniques based on reaction or small molecule similarity. Turning to enzyme engineering, we discuss strategies to improve enzyme thermostability, solubility, and activity, as well as the applications of AI in these fields. The shift from traditional experiment-driven models to data-driven and computationally driven intelligent models is already underway. Finally, we present potential challenges and provide a perspective on future research directions. We envision expanded applications of biocatalysis in drug development, green chemistry, and complex molecule synthesis.

Transmembrane alkane monooxygenase (AlkB)-type monooxygenases, especially AlkB2 monooxygenases, are crucial for aerobic degradation of the medium-to-long-chain n-alkanes in hydrocarbon-utilizing microorganisms. In this study, we identified a GntR/VanR transcription regulator AlkR of Pseudomonas aeruginosa SJTD-1 involved in the negative regulation of AlkB2 and deciphered its nature of DNA binding and ligand release. The deletion of alkR enhanced the transcription levels of the alkB2 gene and the utilization efficiency of the medium-to-long-chain n-alkanes by strain SJTD-1. The dimer of AlkR recognizes and binds to a conserved palindromic motif in the promoter of the alkB2 gene, and structural symmetry is vital for DNA binding and transcription repression. The long-chain fatty acyl coenzyme A compounds can release AlkR and stimulate transcription of alkB2, reflecting the effect of alkane catabolic metabolites. Structural insights unveiled that the arginine residues and scaffold residues of AlkR are critical for DNA binding. Further bioinformatics analysis of AlkR revealed the widespread VanR–AlkB couples distributed in Pseudomonadaceae with high conservation in the sequences of functional genes and intergenic regions, highlighting a conserved regulatory pattern for n-alkane utilization across this family. These findings demonstrate the regulatory mechanism and structural basis of GntR/VanR transcription regulators in modulating n-alkane biodegradation and provide valuable insights in improving the bioremediation efficiency of hydrocarbon pollution.

The escalating antibiotic resistance crisis poses a major global health threat. Bacteriophage therapy offers a promising alternative for combating multidrug-resistant infections. However, bacterial resistance to phages remains a significant hurdle. Innovative strategies are needed to overcome this challenge. In this study, we developed a phage cocktail based on our phage library, consisting of three phages that suppressed phage resistance of carbapenem-resistant hypervirulent Klebsiella pneumoniae (CR-hvKp). This cocktail capitalized on dual instances of collateral sensitivity, thereby constraining the evolution of phage resistance. The first-layered collateral sensitivity arose from overlapping coverage between capsular polysaccharide (CPS) and lipopolysaccharide (LPS), rendering the bacteria resistant to CPS-binding phages but more susceptible to LPS-binding phages. The second-layered collateral sensitivity resulted from an O serotype switch (from O1 to O2), causing resistance to O1 antigen-binding phages but increasing susceptibility to phages that target the O2 antigen. This dual-layered collateral sensitivity phage cocktail effectively mitigated infection caused by CR-hvKp in mice. Our research highlights the importance of the collateral sensitivity mechanism in counteracting the evolution of phage resistance and offers a sophisticated strategy for configuring phage cocktails to eliminate bacterial resistance.

The caseinolytic protease complex ClpP1P2 is crucial for protein homeostasis in mycobacteria and stress response and virulence of the pathogens. Its role as a potential drug target for combating tuberculosis (TB) has just begun to be substantiated in drug discovery research. We conducted a biochemical screening targeting the ClpP1P2 using a library of compounds phenotypically active against Mycobacterium tuberculosis (Mtb). The screening identified a phenyl ester compound GDI-5755, inhibiting the growth of Mtb and M. bovis BCG, the model organism of mycobacteria. GDI-5755 covalently modified the active-site serine residue of ClpP1, rendering the peptidase inactive, which was delineated through protein mass spectrometry and kinetic analyses. GDI-5755 exerted antibacterial activity by inhibiting ClpP1P2 in the bacteria, which could be demonstrated through a minimum inhibitory concentration (MIC) shift assay with a clpP1 CRISPRi knockdown (clpP1-KD) mutant GH189. The knockdown also remarkably heightened the mutant's sensitivity to ethionamide and meropenem, but not to many other TB drugs. On the other hand, a comparative proteomic analysis of wild-type cells exposed to GDI-5755 revealed the dysregulated proteome, specifically showing changes in the expression levels of multiple TB drug targets, including EthA, Ldt_Mt2, and PanD. Subsequent evaluation confirmed the synergistic activity of GDI-5755 when combined with the TB drugs to inhibit mycobacterial growth. Our findings indicate that small-molecule inhibitors targeting ClpP1P2, when used alongside existing TB medications, could represent novel therapeutic strategies.

Mycobacterium abscessus, a fast-growing, non-tuberculous mycobacterium resistant to most antimicrobial drugs, causes a wide range of serious infections in humans, posing a significant public health challenge. The development of effective genetic manipulation tools for M. abscessus is still in progress, limiting both research and therapeutic advancements. However, the clustered regularly interspaced short palindromic repeats (CRISPR)-associated protein (Cas) systems have emerged as promising tools for generating highly specific double-strand breaks (DSBs) in its genome. One of the mechanisms that repair these DSBs is the error-prone nonhomologous end-joining (NHEJ) pathway, which facilitates targeted gene editing. In this study, we introduced a novel application of the CRISPR-NHEJ approach in M. abscessus. We demonstrated that NrgA from M. marinum plays a crucial role in repairing DSBs induced by the CRISPR-Cas system in M. abscessus. Contrary to previous findings, our study also revealed that inhibiting or overexpressing components of homologous recombination/single-strand annealing significantly reduces the efficiency of NHEJ repair in M. abscessus. This discovery challenges current perspectives and suggests that NHEJ repair in M. abscessus may involve components from both homologous recombination and single-strand annealing pathways, highlighting the complex interactions among the three DSB repair mechanisms in M. abscessus.

Klebsiella pneumoniae is a Gram-negative opportunistic pathogenic bacterium that occasionally inhabits the human gastrointestinal tracts. Gut-colonized K. pneumoniae may then metastasize to other organs and tissues, thus causing severe infections. In this study, we identified three cpxA mutations in K. pneumoniae that experimentally evolved to show reduced adhesive ability. CpxA is a sensor histidine kinase that rendered reduced surface adhesion and gut colonization ability in K. pneumoniae. Interestingly, one experimentally gained CpxA mutant (L168del) also commonly occurs in nature. K. pneumoniae containing CpxA variants showed different colonization potentials through altered type 3 fimbriae expression. Lastly, we demonstrated that CpxA contributes to amino acid sensing, thus regulating the colonization of K. pneumoniae both on solid surfaces and in mouse intestines. The polymorphism of CpxA may help to broaden the environmental adaptation of the bacterium. These findings together reveal a Cpx-mediated regulation to diversify intestinal colonization in K. pneumoniae.

Eukaryotes are hypothesized to be archaeal–bacterial chimeras. Given the different chemical structures of membrane phospholipids in archaea and bacteria, transformations of membranes during eukaryogenesis that led to the bacterial-type membranes of eukaryotic cells remain a major conundrum. One of the possible intermediates of eukaryogenesis could involve an archaeal–bacterial hybrid membrane. So far, organisms with hybrid membranes have not been discovered, and experimentation on such membranes has been limited. To generate mixed membranes, we reconstructed the archaeal membrane lipid biosynthesis pathway in Escherichia coli, creating three strains that individually produced archaeal lipids ranging from simple, such as DGGGOH (digeranylgeranylglycerol) and archaeol, to complex, such as GDGT (glycerol dialkyl glycerol tetraether). The physiological responses became more pronounced as the hybrid membrane incorporated more complex archaeal membrane lipids. In particular, biosynthesis of GDGT induced a pronounced SOS response, accompanied by cellular filamentation, explosive cell lysis, and ATP accumulation. Thus, bacteria seem to be able to incorporate simple archaeal membrane lipids, such as DGGGOH and archaeol, without major fitness costs, compatible with the involvement of hybrid membranes at the early stages of cell evolution and in eukaryogenesis. By contrast, the acquisition of more complex, structurally diverse membrane lipids, such as GDGT, appears to be strongly deleterious to bacteria, suggesting that this type of lipid is an archaeal innovation.

Microorganisms dominate marine environments in the polar oceans and are known to harbor greater diversity and abundance than was once thought, and yet, little is known about their biogeographic distribution patterns in marine sediments at a broad spatial scale. In this study, we conducted extensive sampling of marine sediments along a latitudinal transect spanning 2500 km from the Bering Sea to the Arctic Ocean to investigate the geographical distribution patterns of bacteria, archaea, and fungi. Our findings revealed that the community similarities of bacteria and fungi decay at similar rates with increasing geographical distance (slope: −0.005 and −0.002), which are much lower than the decay rate of archaeal communities (slope: −0.012). Notably, microbial richness and community composition showed significant differences in the region of 75−80°N compared to other regions in 60−75°N. Salinity, temperature, pH, ammonium nitrogen, and total organic carbon are key factors that significantly affect microbial community variations. Furthermore, bacterial co-occurrence networks showed more complex interactions but lower modularity than fungal counterparts. This study provides crucial insights into the spatial distribution patterns of bacteria, archaea, and fungi in the Arctic marine sediments and will be critical for a better understanding of microbial global distribution and ecological functions.