Characterization and heterologous reconstitution of Taxus biosynthetic enzymes leading to baccatin III.

Paclitaxel is a well known anticancer compound. Its biosynthesis involves the formation of a highly functionalized diterpenoid core skeleton (baccatin III) and the subsequent assembly of a phenylisoserinoyl side chain. Despite intensive investigation for half a century, the complete biosynthetic pathway of baccatin III remains unknown. In this work, we identified a bifunctional cytochrome P450 enzyme [taxane oxetanase 1 (TOT1)] in Taxus mairei that catalyzes an oxidative rearrangement in paclitaxel oxetane formation, which represents a previously unknown enzyme mechanism for oxetane ring formation. We created a screening strategy based on the taxusin biosynthesis pathway and uncovered the enzyme responsible for the taxane oxidation of the C9 position (T9αH1). Finally, we artificially reconstituted a biosynthetic pathway for the production of baccatin III in tobacco.

Shaping the Water-Harvesting Behavior of Metal-Organic Frameworks Aided by Fine-Tuned GPT Models.

We construct a data set of metal-organic framework (MOF) linkers and employ a fine-tuned GPT assistant to propose MOF linker designs by mutating and modifying the existing linker structures. This strategy allows the GPT model to learn the intricate language of chemistry in molecular representations, thereby achieving an enhanced accuracy in generating linker structures compared with its base models. Aiming to highlight the significance of linker design strategies in advancing the discovery of water-harvesting MOFs, we conducted a systematic MOF variant expansion upon state-of-the-art MOF-303 utilizing a multidimensional approach that integrates linker extension with multivariate tuning strategies. We synthesized a series of isoreticular aluminum MOFs, termed Long-Arm MOFs (LAMOF-1 to LAMOF-10), featuring linkers that bear various combinations of heteroatoms in their five-membered ring moiety, replacing pyrazole with either thiophene, furan, or thiazole rings or a combination of two. Beyond their consistent and robust architecture, as demonstrated by permanent porosity and thermal stability, the LAMOF series offers a generalizable synthesis strategy. Importantly, these 10 LAMOFs establish new benchmarks for water uptake (up to 0.64 g g-1) and operational humidity ranges (between 13 and 53%), thereby expanding the diversity of water-harvesting MOFs.

Evaluating Scalable Supervised Learning for Synthesize-on-Demand Chemical Libraries.

Traditional small-molecule drug discovery is a time-consuming and costly endeavor. High-throughput chemical screening can only assess a tiny fraction of drug-like chemical space. The strong predictive power of modern machine-learning methods for virtual chemical screening enables training models on known active and inactive compounds and extrapolating to much larger chemical libraries. However, there has been limited experimental validation of these methods in practical applications on large commercially available or synthesize-on-demand chemical libraries. Through a prospective evaluation with the bacterial protein-protein interaction PriA-SSB, we demonstrate that ligand-based virtual screening can identify many active compounds in large commercial libraries. We use cross-validation to compare different types of supervised learning models and select a random forest (RF) classifier as the best model for this target. When predicting the activity of more than 8 million compounds from Aldrich Market Select, the RF substantially outperforms a naïve baseline based on chemical structure similarity. 48% of the RF's 701 selected compounds are active. The RF model easily scales to score one billion compounds from the synthesize-on-demand Enamine REAL database. We tested 68 chemically diverse top predictions from Enamine REAL and observed 31 hits (46%), including one with an IC50 value of 1.3 µM.

Scientific discovery in the age of artificial intelligence.

Artificial intelligence (AI) is being increasingly integrated into scientific discovery to augment and accelerate research, helping scientists to generate hypotheses, design experiments, collect and interpret large datasets, and gain insights that might not have been possible using traditional scientific methods alone. Here we examine breakthroughs over the past decade that include self-supervised learning, which allows models to be trained on vast amounts of unlabelled data, and geometric deep learning, which leverages knowledge about the structure of scientific data to enhance model accuracy and efficiency. Generative AI methods can create designs, such as small-molecule drugs and proteins, by analysing diverse data modalities, including images and sequences. We discuss how these methods can help scientists throughout the scientific process and the central issues that remain despite such advances. Both developers and users of AI toolsneed a better understanding of when such approaches need improvement, and challenges posed by poor data quality and stewardship remain. These issues cut across scientific disciplines and require developing foundational algorithmic approaches that can contribute to scientific understanding or acquire it autonomously, making them critical areas of focus for AI innovation.

The transcription factor CitZAT5 modifies sugar accumulation and hexose proportion in citrus fruit.

Sugars are fundamental to plant developmental processes. For fruits, the accumulation and proportion of sugars play crucial roles in the development of quality and attractiveness. In citrus (Citrus reticulata Blanco.), we found that the difference in sweetness between mature fruits of "Gongchuan" and its bud sport "Youliang" is related to hexose contents. Expression of a SuS (sucrose synthase) gene CitSUS5 and a SWEET (sugars will eventually be exported transporter) gene CitSWEET6, characterized by transcriptome analysis at different developmental stages of these 2 varieties, revealed higher expression levels in "Youliang" fruit. The roles of CitSUS5 and CitSWEET6 were investigated by enzyme activity and transient assays. CitSUS5 promoted the cleavage of sucrose to hexoses, and CitSWEET6 was identified as a fructose transporter. Further investigation identified the transcription factor CitZAT5 (ZINC FINGER OF ARABIDOPSIS THALIANA) that contributes to sucrose metabolism and fructose transportation by positively regulating CitSUS5 and CitSWEET6. The role of CitZAT5 in fruit sugar accumulation and hexose proportion was investigated by homologous transient CitZAT5 overexpression, -VIGS, and -RNAi. CitZAT5 modulates the hexose proportion in citrus by mediating CitSUS5 and CitSWEET6 expression, and the molecular mechanism explained the differences in sugar composition of "Youliang" and "Gongchuan" fruit.

Genome-wide identification and expression analysis of MATE gene family in citrus fruit (Citrus clementina).

Multidrug and toxic compound extrusion (MATE) proteins are a class of secondary active multidrug transporters. In plants, this family has significantly expanded and is involved in numerous plant physiological processes. Although MATE proteins have been identified in an increasing number of species, the understanding about this family in citrus remains unclear. In this study, a total of 69 MATE transporters were identified in the citrus genome (Citrus clementina) and classified into four groups by phylogenetic analysis. Tandem and segmental duplication events were the main causes of the citrus MATE family expansion. RNA-seq and qRT-PCR analyses were performed during citrus fruit development. The results indicated that CitMATE genes showed specific expression profiles in citrus peels and flesh at different developmental stages. Combined with the variations of flavonoids and citrate levels in citrus fruit, we suggested that CitMATE43 and CitMATE66 may be involved in the transport process of flavonoids and citrate in citrus fruit, respectively. In addition, two flavonoids positive regulators, CitERF32 and CitERF33, both directly bind to and activated the CitMATE43 promoter. Our results provide comprehensive information on citrus MATE genes and valuable understanding for the flavonoids and citrate metabolism in citrus fruit.

The Interaction Between CitMYB52 and CitbHLH2 Negatively Regulates Citrate Accumulation by Activating CitALMT in Citrus Fruit.

Citric acid plays significant roles in numerous physiological processes in plants, including carbon metabolism, signal transduction, and tolerance to environmental stress. For fruits, it has a major effect on fruit organoleptic quality by directly influencing consumer taste. Citric acid in citrus is mainly regulated by the balance between synthesis, degradation, and vacuolar storage. The genetic and molecular regulations of citric acid synthesis and degradation have been comprehensively elucidated. However, the transporters for citric acid in fruits are less well understood. Here, an aluminum-activated malate transporter, CitALMT, was characterized. Transient overexpression and stable transformation of CitALMT significantly reduced citrate concentration in citrus fruits and transgenic callus. Correspondingly, transient RNA interference-induced silencing of CitALMT and increased citrate significantly, indicating that CitALMT plays an important role in regulating citrate concentration in citrus fruits. In addition, dual-luciferase assays indicated that CitMYB52 and CitbHLH2 could trans-activate the promoter of CitALMT. EMSA analysis showed that CitbHLH2 could physically interact with the E-box motif in the CitALMT promoter. Bimolecular fluorescence complementation assays, yeast two-hybrid, coimmunoprecipitation and transient overexpression, and RNAi assay indicated that the interaction between CitMYB52 and CitbHLH2 could synergistically trans-activate CitALMT to negatively regulate citrate accumulation.

Citrus heat shock transcription factor CitHsfA7-mediated citric acid degradation in response to heat stress.

Heat stress is a major abiotic stress for plants, which can generate a range of biochemical and genetic responses. In 'Ponkan' mandarin fruit, hot air treatment (HAT) accelerates the degradation of citric acid. However, the transcriptional regulatory mechanisms of citrate degradation in response to HAT remain to be elucidated. Here, 17 heat shock transcription factor sequences were isolated, and dual-luciferase assays were employed to investigate whether the encoded proteins that could trans-activate the promoters of key genes in the GABA shunt, involved in citrate metabolism. We identified four heat shock transcription factors (CitHsfA7, CitHsfA3, CitHsfA4b and CitHsfA8) that showed trans-activation effects on CitAco3, CitIDH3 and CitGAD4, respectively. Transient expression of the CitHsfs in citrus fruits indicated that CitHsfA7 was the only factor that resulted in a significant lowering of the citric acid content, and these results were confirmed by a virus-induced gene silencing system (VIGS). Sub-cellar localization showed that CitHsfA7 is located in the nucleus and is capable of binding directly to a putative HSE in the CitAco3 promoter and enhance its expression. We proposed that the induction of CitHsfA7 transcript level contributes to citric acid degradation in citrus fruit, via modulation of CitAco3 in response to HAT.

The effect of NH4+ on phosphoenolpyruvate carboxykinase gene expression, metabolic flux and citrate content of citrus juice sacs.

Citrate is one of the most important metabolites determining the flavour of citrus fruit. It has been reported that nitrogen supply may have an impact on acid level of fruit. Here, the relationship between nitrogen metabolism and citrate catabolism was studied in pumelo juice sacs. Differences in metabolites, gene expression and flux distributions were analyzed in juice sacs incubated in medium with and without NH4+. Compared with those incubated with NH4+, juice sacs under nitrogen deficiency exhibited enhanced flux through phosphoenolpyruvate carboxykinase (PEPCK) and accelerated consumption of citrate, while the other two TCA cycle efflux points, through malic enzyme (ME) and glutamate dehydrogenase (GDH), were both repressed. Consistent with the estimated fluxes, the expression of PEPCK1 was upregulated under nitrogen deficiency, while that of GDH1, GDH2, NAD-ME1 and NADP-ME2 were all repressed. Thus, we propose that PEPCK1 contributes to citrate degradation under nitrogen limitation.

Citrus CitERF6 Contributes to Citric Acid Degradation via Upregulation of CitAclα1, Encoding ATP-Citrate Lyase Subunit α.

Citric acid is the most abundant organic acid in citrus fruit, and the acetyl-CoA pathway potentially plays an important role in citric acid degradation, which occurs during fruit ripening. Analysis of transcripts during fruit development of key genes in the acetyl-CoA pathway and transient overexpression assay in citrus leaves indicated that CitAclα1 could be a potential target gene involved in citrate degradation. In order to understand more about CitAclα1, 23 transcription factors coexpressed with CitAclα1 in citrus fruit were identified by RNA-seq. Using dual-luciferase assays, CitERF6 was shown to trans-activate the promoter of CitAclα1 and electrophoretic mobility shift assays (EMSAs) showed that CitERF6 directly bound to a 5'-CAACA-3' motif in the CitAclα1 promoter. Furthermore, citric acid content was significantly reduced when CitERF6 was overexpressed in transgenic tobacco leaves. Taken together, these results indicate an important role for CitERF6 in transcriptional regulation of CitAclα1 and control of citrate degradation.

Learning Drug Functions from Chemical Structures with Convolutional Neural Networks and Random Forests.

Empirical testing of chemicals for drug efficacy costs many billions of dollars every year. The ability to predict the action of molecules in silico would greatly increase the speed and decrease the cost of prioritizing drug leads. Here, we asked whether drug function, defined as MeSH "therapeutic use" classes, can be predicted from only a chemical structure. We evaluated two chemical-structure-derived drug classification methods, chemical images with convolutional neural networks and molecular fingerprints with random forests, both of which outperformed previous predictions that used drug-induced transcriptomic changes as chemical representations. This suggests that the structure of a chemical contains at least as much information about its therapeutic use as the transcriptional cellular response to that chemical. Furthermore, because training data based on chemical structure is not limited to a small set of molecules for which transcriptomic measurements are available, our strategy can leverage more training data to significantly improve predictive accuracy to 83-88%. Finally, we explore use of these models for prediction of side effects and drug-repurposing opportunities and demonstrate the effectiveness of this modeling strategy for multilabel classification.

Auto- and mutual-regulation between two CitERFs contribute to ethylene-induced citrus fruit degreening.

Citrus fruit postharvest degreening is a critical stage in marketing, carried out by exposure to ethylene or ethephon. Genome-wide screening of the AP2/ERF superfamily indicated that a novel ERF-II (CitERF6) was shown to trans-activate the CitPPH promoter. Expression of CitERF6 is associated with both developmental and postharvest degreening in citrus fruit. Transient and stable over-expression of CitERF6 in Nicotiana tabacum leaves and 'Ponkan' fruit also results in rapid chlorophyll degradation. Auto- and mutual-regulation was also found between CitERF6 and the previously characterized CitERF13 using the dual-luciferase and yeast one-hybrid assays. Moreover, substitution of the 35S promoter for endogenous promoters showed that both pCitERF6::CitERF6 and pCitERF13::CitERF13 were effective in trans-activating their promoters or triggering chlorophyll degradation. It is proposed that ethylene is one of the triggers activating promoters of CitERF6 and CitERF13, and subsequent auto- and mutual-regulation between CitERF6 and CitERF13 might facilitate the effect of ethylene, leading to fruit degreening.

Practical Model Selection for Prospective Virtual Screening.

Virtual (computational) high-throughput screening provides a strategy for prioritizing compounds for experimental screens, but the choice of virtual screening algorithm depends on the data set and evaluation strategy. We consider a wide range of ligand-based machine learning and docking-based approaches for virtual screening on two protein-protein interactions, PriA-SSB and RMI-FANCM, and present a strategy for choosing which algorithm is best for prospective compound prioritization. Our workflow identifies a random forest as the best algorithm for these targets over more sophisticated neural network-based models. The top 250 predictions from our selected random forest recover 37 of the 54 active compounds from a library of 22,434 new molecules assayed on PriA-SSB. We show that virtual screening methods that perform well on public data sets and synthetic benchmarks, like multi-task neural networks, may not always translate to prospective screening performance on a specific assay of interest.