Reverse‐prenylated phenolic compounds are an abundant class of bioactive plant natural products. Hyperixanthone A, an inhibitor of multidrug‐resistant Staphylococcus aureus, is a polyprenylated xanthone carrying two forward geminal and one reverse prenyl group. Although prenyltransferases responsible for the forward prenylations were identified, the final reverse prenylation reaction remained elusive. No plant enzyme catalyzing reverse prenylation of an aromatic carbon has been described so far. Here we use metabolic profiling and transcriptomic information from Hypericum perforatum and H. sampsonii to identify homologous enzymes involved in the formation of reverse‐prenylated xanthones and characterize their functions using in vitro, in vivo, and in silico approaches. The identified enzymes are non‐canonical UbiA‐type prenyltransferases, which surprisingly catalyze both forward and reverse prenylations with different regioselectivities. Reconstruction of the enzyme cascade in Saccharomyces cerevisiae and Nicotiana benthamiana confirmed reverse‐prenylated hyperixanthone A as the major product. Molecular modeling and docking simulations supported by site‐directed mutagenesis suggest two distinct binding modes, which enable forward and reverse prenylations and provide a rationale for the preferred catalysis of the reverse prenyl transfer reaction. The identification of reverse prenylation augments the repertoire of reactions catalyzed by membrane‐bound UbiA‐type plant aromatic prenyltransferases. The insights also provide a new tool for the biotechnological modification of pharmaceutically valuable natural products.

Three geranylgeranyl pyrophosphate derivatives carrying an ether group at different positions within geranylgeranyl pyrophosphate were employed in biotransformations with five diterpene synthases (CotB2, PvHVS, PaFS, Bnd4 and TXS) derived from plants, bacteria and fungi. A total of six new oxygen-containing diterpenoids were isolated and characterized, deepening our knowledge on the substrate promiscuity of diterpene synthases. In addition, the diterpene synthase PvHVS also accepts an ether derivative of farnesyl pyrophosphate and converts it to the same tetrahydrofuran core as found for the analogous extended GGPP substrate. This result further demonstrates that diterpene synthases also exhibit promiscuity toward truncated unnatural substrates.

Protein fitness prediction plays a crucial role in the advancement of protein engineering endeavours. However, the combinatorial complexity of the protein sequence space and the limited availability of assay-labelled data hinder the efficient optimization of protein properties. Data-driven strategies utilizing machine learning methods have emerged as a promising solution, yet their dependence on labelled training datasets poses a significant obstacle. To overcome this challenge, in this work, we explore various ways of introducing the latent information present in evolutionarily related sequences (homologous sequences) into the training process. To do so, we establish several strategies based on semi-supervised learning (unsupervised pre-processing and wrapper methods) and perform a comprehensive comparison using 19 datasets containing protein-fitness pairs. Our findings reveal that using the information present in the homologous sequences can improve the performance of the models, especially when the number of available labelled sequences is considerably low. Specifically, the combination of a sequence encoding method based on Direct Coupling Analysis (DCA), with MERGE (a hybrid regression framework that combines evolutionary information with supervised learning) and an SVM regressor, outperforms other encodings (PAM250, UniRep, eUniRep) and other semi-supervised wrapper methods (Tri-Training Regressor, Co-Training Regressor). In summary, the demonstrated performance gains of this strategy mark a substantial leap towards more robust and reliable predictive models for protein engineering tasks. This advancement holds the potential to streamline the design and optimisation of proteins for diverse applications in biotechnology and therapeutics.

Poly(ADP-ribose) polymerases (PARP) are a family of enzymes that were proven to play an essential role in the initiation and activation of DNA repair processes in the case of DNA single-strand breaks. The inhibition of PARP enzymes might be a promising option for the treatment of several challenging types of cancers, including triple-negative breast cancer (TNBC) and non-small cell lung carcinoma (NSCLC). This study utilizes several techniques to screen the compound collection of the Leibniz Institute of Plant Biochemistry (IPB) to identify novel hPARP-1 inhibitors. First, an in silico pharmacophore-based docking study was conducted to virtually screen compounds with potential inhibitory effects. To evaluate these compounds in vitro, a cell-free enzyme assay was developed, optimized, and employed to identify hPARP-1 inhibitors, resulting in the discovery of two novel scaffolds represented by compounds 54 and 57, with the latter being the most active one from the compound library. Furthermore, fluorescence microscopy and synergism assays were performed to investigate the cellular and nuclear pathways of hPARP-1 inhibitor 57 and its potential synergistic effect with the DNA-damaging agent temozolomide. The findings suggest that the compound requires further lead optimization to enhance its ability to target the nuclear PARP enzyme effectively. Nonetheless, this new scaffold demonstrated a five-fold higher PARP inhibitory activity at the enzyme level compared to the core structure of olaparib (OLP), phthalazin-1(2H)-one.

Protein engineering through directed evolution and (semi)rational design has become a powerful approach for optimizing and enhancing proteins with desired properties. The integration of artificial intelligence methods has further accelerated protein engineering process by enabling the development of predictive models based on datadriven strategies. However, the lack of interpretability and transparency in these models limits their trustworthiness and applicability in real-world scenarios. Explainable Artificial Intelligence addresses these challenges by providing insights into the decision-making processes of machine learning models, enhancing their reliability and interpretability. Explainable strategies has been successfully applied in various biotechnology fields, including drug discovery, genomics, and medicine, yet its application in protein engineering remains underexplored. The incorporation of explainable strategies in protein engineering holds significant potential, as it can guide protein design by revealing how predictive models function, benefiting approaches such as machine learning-assisted directed evolution. This perspective work explores the principles and methodologies of explainable artificial intelligence, highlighting its relevance in biotechnology and its potential to enhance protein design. Additionally, three theoretical pipelines integrating predictive models with explainable strategies are proposed, focusing on their advantages, disadvantages, and technical requirements. Finally, the remaining challenges of explainable artificial intelligence in protein engineering and future directions for its development as a support tool for traditional protein engineering methodologies are discussed.

Protein engineering is a powerful tool to enhance the catalytic performance of enzymes. State-of-the-art methodologies usually focus on identifying positions in localized regions, such as the binding pockets and substrate access tunnels. Here, we presented a global dynamic design strategy to evolve cytochrome P450s for improved catalytic performance by identifying distant residues. Three potential positions (D164, A195, and E405) were identified and subjected to site saturation mutagenesis (SSM), resulting in the beneficial variant 8G8/D164N/A195P, which exhibited 7- and 148-fold improved kcat values compared to variants 8G8 and WT, respectively. Dynamic tunnel and conformational dynamic analyses revealed that substitutions distant from substrate access tunnels increased substrate translocation through the substrate access tunnels of CYP116B3, improving the catalytic performance. General applicability was demonstrated by transferring the global dynamic design strategy to an additional P450 enzyme, namely P450 BM3. The proposed global design strategy advances state-of-the-art P450 engineering by improving the catalytic performance of P450 by identifying distant correlated positions that enhance substrate translocation through access tunnels.

Enhancing the performance of cellulases at high temperatures is crucial for efficient biomass hydrolysis—a fundamental process in biorefineries. Traditional protein engineering methods, while effective, are time-consuming and labour-intensive, limiting rapid advancements. To streamline the engineering process, we tested two distinct in silico methods for predicting thermally resistant and highly active variants of Penicillium verruculosum endoglucanase II. Specifically, we used FoldX to pinpoint structure-stabilizing substitutions (ΔΔG < 0) and applied the sequence-based method EVmutation to identify evolutionarily favorable substitutions (ΔE > 0). Experimental validation of the top 20 ranked single-substituted variants from both methods showed that EVmutation outperformed FoldX, identifying variants with enhanced enzyme activity after one-hour incubation at 75 1C (up to 3.6-fold increase), increased melting temperature (ΔTm of 2.8 °C), and longer half-lives at 75 1C (up to 104 minutes vs. 40 minutes for the wild type). Building upon these results, EVmutation was used to predict variants with two amino acid substitutions. These double-substituted endoglucanase variants showed further improvements—up to a 4.4-fold increase in activity, ΔTm gains of 3.7 °C, and half-life extensions up to 82 minutes. This study highlights EVmutation’s potential for accelerating protein engineering campaigns and enhancing enzyme properties while reducing experimental efforts, thereby contributing to more efficient and sustainable bioprocesses.

Hornstedtia scyphifera (J.Koenig) Steud. represents a lesser-known member of the ginger family (Zingiberaceae) that is used in Malaysia as spice and traditional medicine. The phytochemical investigation of leaves from this species utilizing diverse analytical methods has provided comprehensive insights into its chemical profile for the first time. Headspace gas chromatography-mass spectrometry (HS-GCMS) and GCMS analyses of essential oil and nonpolar extracts verified α-pinene, camphene, p-cymene, and camphor as main volatile compounds. Metabolite profiling of the crude extract by ultra-high-performance-liquid chromatography-high resolution mass spectrometry (UHPLC-HRMS) unveiled terpenoids, flavonoids and other phenolics as major compound classes. Isolation and follow-up structure elucidation, involving 1D and 2D NMR, HRMS, UV and CD analysis, yielded two new sesquiterpenoids, (1R,5S,6S,7R,10R)-mustak-14-oic acid (1) and (1R,6S,7S,10R)-6-hydroxy-anhuienosol (2), along with 24 known compounds (seven terpenoids, seven flavonoids, ten phenolics), 21 of these never reported for H. scyphifera. Additionally, the crude extract and fractions from the purification process were screened for antibacterial and antifungal activity. This is supplemented by an extensive literature research for described bioactivities of all isolated compounds. Our results support and explain previously detected antimicrobial, antifungal and neuroprotective effects of H. scyphifera extracts and provide evidence for its potential pharmacological importance.