Development of method using language processing techniques for extracting information on drug-health food product interactions.

AIMS: Health food products (HFPs) are foods and products related to maintaining and promoting health. HFPs may sometimes cause unforeseen adverse health effects by interacting with drugs. Considering the importance of information on the interactions between HFPs and drugs, this study aimed to establish a workflow to extract information on Drug-HFP Interactions (DHIs) from open resources. METHODS: First, Information on drugs, enzymes, their interactions, and known DHIs was collected from multiple public databases and literature sources. Next, a network consisted of enzymes, HFP, and drugs was constructed, assuming enzymes as candidates for hubs in Drug-HFP interactions (Method 1). Furthermore, we developed methods to analyze the biomedical context of each drug and HFP to predict potential DHIs out of the DHIs obtained in Method 1 by applying BioWordVec, a widely used biomedical terminology quantifier (Method 2-1 and 2-2). RESULTS: 44,965 DHIs (30% known) were identified in Method 1, including 38 metabolic enzymes, 157 HFPs, and 1256 drugs. Method 2-1 selected 7401 DHIs (17% known) from the DHIs of Method 1, while Method 2-2 chose 2819 DHIs (30% known). Based on the different assumptions in these methods where Method 2-1 specifically selects HFPs interacting with specific enzymes and Method 2-2 specifically selects HFPs with similar function with drugs, the propsed methods resulted in extracting a wide variety of DHIs. CONCLUSIONS: By integrating the results of language processing techniques with those of the network analysis, a workflow to efficiently extract unknown and known DHIs was constructed.

EnzymeNet: residual neural networks model for Enzyme Commission number prediction.

Motivation: Enzymes are key targets to biosynthesize functional substances in metabolic engineering. Therefore, various machine learning models have been developed to predict Enzyme Commission (EC) numbers, one of the enzyme annotations. However, the previously reported models might predict the sequences with numerous consecutive identical amino acids, which are found within unannotated sequences, as enzymes. Results: Here, we propose EnzymeNet for prediction of complete EC numbers using residual neural networks. EnzymeNet can exclude the exceptional sequences described above. Several EnzymeNet models were built and optimized to explore the best conditions for removing such sequences. As a result, the models exhibited higher prediction accuracy with macro F1 score up to 0.850 than previously reported models. Moreover, even the enzyme sequences with low similarity to training data, which were difficult to predict using the reported models, could be predicted extensively using EnzymeNet models. The robustness of EnzymeNet models will lead to discover novel enzymes for biosynthesis of functional compounds using microorganisms. Availability and implementation: The source code of EnzymeNet models is freely available at https://github.com/nwatanbe/enzymenet.

Prediction of Metabolic Flux Distribution by Flux Sampling: As a Case Study, Acetate Production from Glucose in Escherichia coli.

Omics data was acquired, and the development and research of metabolic simulation and analysis methods using them were also actively carried out. However, it was a laborious task to acquire such data each time the medium composition, culture conditions, and target organism changed. Therefore, in this study, we aimed to extract and estimate important variables and necessary numbers for predicting metabolic flux distribution as the state of cell metabolism by flux sampling using a genome-scale metabolic model (GSM) and its analysis. Acetic acid production from glucose in Escherichia coli with GSM iJO1366 was used as a case study. Flux sampling obtained by OptGP using 1000 pattern constraints on substrate, product, and growth fluxes produced a wider sample than the default case. The analysis also suggested that the fluxes of iron ions, O2, CO2, and NH4+, were important for predicting the metabolic flux distribution. Additionally, the comparison with the literature value of 13C-MFA using CO2 emission flux as an example of an important flux suggested that the important flux obtained by this method was valid for the prediction of flux distribution. In this way, the method of this research was useful for extracting variables that were important for predicting flux distribution, and as a result, the possibility of contributing to the reduction of measurement variables in experiments was suggested.

Different Recognition of Protein Features Depending on Deep Learning Models: A Case Study of Aromatic Decarboxylase UbiD.

The number of unannotated protein sequences is explosively increasing due to genome sequence technology. A more comprehensive understanding of protein functions for protein annotation requires the discovery of new features that cannot be captured from conventional methods. Deep learning can extract important features from input data and predict protein functions based on the features. Here, protein feature vectors generated by 3 deep learning models are analyzed using Integrated Gradients to explore important features of amino acid sites. As a case study, prediction and feature extraction models for UbiD enzymes were built using these models. The important amino acid residues extracted from the models were different from secondary structures, conserved regions and active sites of known UbiD information. Interestingly, the different amino acid residues within UbiD sequences were regarded as important factors depending on the type of models and sequences. The Transformer models focused on more specific regions than the other models. These results suggest that each deep learning model understands protein features with different aspects from existing knowledge and has the potential to discover new laws of protein functions. This study will help to extract new protein features for the other protein annotations.

Mangrove crab intestine and habitat sediment microbiomes cooperatively work on carbon and nitrogen cycling.

Mangrove ecosystems, where litter and organic components are degraded and converted into detrital materials, support rich coastal fisheries resources. Sesarmid (Grapsidae) crabs, which feed on mangrove litter, play a crucial role in material flow in carbon-rich and nitrogen-limited mangrove ecosystems; however, the process of assimilation and conversion into detritus has not been well studied. In this study, we performed microbiome analyses of intestinal bacteria from three species of mangrove crab and five sediment positions in the mud lobster mounds, including the crab burrow wall, to study the interactive roles of crabs and sediment in metabolism. Metagenome analysis revealed species-dependent intestinal profiles, especially in Neosarmatium smithi, while the sediment microbiome was similar in all positions, albeit with some regional dependency. The microbiome profiles of crab intestines and sediments were significantly different in the MDS analysis based on OTU similarity; however, 579 OTUs (about 70% of reads in the crab intestinal microbiome) were identical between the intestinal and sediment bacteria. In the phenotype prediction, cellulose degradation was observed in the crab intestine. Cellulase activity was detected in both crab intestine and sediment. This could be mainly ascribed to Demequinaceae, which was predominantly found in the crab intestines and burrow walls. Nitrogen fixation was also enriched in both the crab intestines and sediments, and was supported by the nitrogenase assay. Similar to earlier reports, sulfur-related families were highly enriched in the sediment, presumably degrading organic compounds as terminal electron acceptors under anaerobic conditions. These results suggest that mangrove crabs and habitat sediment both contribute to carbon and nitrogen cycling in the mangrove ecosystem via these two key reactions.

Knowledge extraction from literature and enzyme sequences complements FBA analysis in metabolic engineering.

Flux balance analysis (FBA) using genome-scale metabolic model (GSM) is a useful method for improving the bio-production of useful compounds. However, FBA often does not impose important constraints such as nutrients uptakes, by-products excretions and gases (oxygen and carbon dioxide) transfers. Furthermore, important information on metabolic engineering such as enzyme amounts, activities, and characteristics caused by gene expression and enzyme sequences is basically not included in GSM. Therefore, simple FBA is often not sufficient to search for metabolic manipulation strategies that are useful for improving the production of target compounds. In this study, we proposed a method using literature and enzyme search to complement the FBA-based metabolic manipulation strategies. As a case study, this method was applied to shikimic acid production by Corynebacterium glutamicum to verify its usefulness. As unique strategies in literature-mining, overexpression of the transcriptional regulator SugR and gene disruption related to by-products productions were complemented. In the search for alternative enzyme sequences, it was suggested that those candidates are searched for from various species based on features captured by deep learning, which are not simply homologous to amino acid sequences of the base enzymes.

Dynamic Flux Balance Analysis to Evaluate the Strain Production Performance on Shikimic Acid Production in Escherichia coli.

Flux balance analysis (FBA) is used to improve the microbial production of useful compounds. However, a large gap often exists between the FBA solution and the experimental yield, because of growth and byproducts. FBA has been extended to dynamic FBA (dFBA), which is applicable to time-varying processes, such as batch or fed-batch cultures, and has significantly contributed to metabolic and cultural engineering applications. On the other hand, the performance of the experimental strains has not been fully evaluated. In this study, we applied dFBA to the production of shikimic acid from glucose in Escherichia coli, to evaluate the production performance of the strain as a case study. The experimental data of glucose consumption and cell growth were used as FBA constraints. Bi-level FBA optimization with maximized growth and shikimic acid production were the objective functions. Results suggest that the shikimic acid concentration in the high-shikimic-acid-producing strain constructed in the experiment reached up to 84% of the maximum value by simulation. Thus, this method can be used to evaluate the performance of strains and estimate the milestones of strain improvement.

Mechanism-based tuning of insect 3,4-dihydroxyphenylacetaldehyde synthase for synthetic bioproduction of benzylisoquinoline alkaloids.

Previous studies have utilized monoamine oxidase (MAO) and L-3,4-dihydroxyphenylalanine decarboxylase (DDC) for microbe-based production of tetrahydropapaveroline (THP), a benzylisoquinoline alkaloid (BIA) precursor to opioid analgesics. In the current study, a phylogenetically distinct Bombyx mori 3,4-dihydroxyphenylacetaldehyde synthase (DHPAAS) is identified to bypass MAO and DDC for direct production of 3,4-dihydroxyphenylacetaldehyde (DHPAA) from L-3,4-dihydroxyphenylalanine (L-DOPA). Structure-based enzyme engineering of DHPAAS results in bifunctional switching between aldehyde synthase and decarboxylase activities. Output of dopamine and DHPAA products is fine-tuned by engineered DHPAAS variants with Phe79Tyr, Tyr80Phe and Asn192His catalytic substitutions. Balance of dopamine and DHPAA products enables improved THP biosynthesis via a symmetrical pathway in Escherichia coli. Rationally engineered insect DHPAAS produces (R,S)-THP in a single enzyme system directly from L-DOPA both in vitro and in vivo, at higher yields than that of the wild-type enzyme. However, DHPAAS-mediated downstream BIA production requires further improvement.

Author Correction: Mechanism-based tuning of insect 3,4-dihydroxyphenylacetaldehyde synthase for synthetic bioproduction of benzylisoquinoline alkaloids.

In the original version of this Article, the abbreviation of 3,4-dihydroxyphenylacetaldehyde synthase presented in the first paragraph of the Discussion section was given incorrectly as DYPAA. The correct abbreviation for this enzyme is DHPAAS. This error has been corrected in both the PDF and HTML versions of the Article.

Kinetic modeling and sensitivity analysis for higher ethanol production in self-cloning xylose-using Saccharomyces cerevisiae.

We constructed a xylose-utilizing Saccharomyces cerevisiae strain using endogenous xylose-assimilating genes (strain K7-XYL). Such self-cloning yeast is expected to make a great contribution to cost reduction of ethanol production processes. However, it is difficult to modify self-cloning yeast for optimal performance because the available gene source is limited. To improve the ethanol productivity of our self-cloning yeast, a kinetic model of ethanol production was constructed and sensitivity analysis was performed. Alcohol dehydrogenase (ADH1) was identified as a metabolic bottleneck reaction in the ethanol production pathway. An ADH1 overexpression strain (K7-XYL-ADH1) was constructed and evaluated in YP (yeast extract 10 g/L, peptone 20 g/L) medium containing 50 g/L xylose as the sole carbon source. Strain K7-XYL-ADH1 showed higher ethanol productivity (13.8 g/L) than strain K7-XYL (12.5 g/L). Then, K7-XYL-ADH1 was evaluated in YP medium containing 80 g/L glucose and 50 g/L xylose; however, the ethanol productivity did not change relative to that of K7-XYL (K7-XYL 46.3 g/L, K7-XYL-ADH1 45.9 g/L). We presumed that due to the presence of glucose, the internal redox balance of the cells had changed. On culturing in an aerated 5-L jar fermentor to change the internal redox balance of cells, strain K7-XYL-ADH1 showed higher ethanol productivity than K7-XYL (K7-XYL 45.0 g/L, K7-XYL-ADH1 49.4 g/L). Our results confirmed that ADH1 was a metabolic bottleneck in the ethanol production pathway. By eliminating the bottleneck, self-cloning yeast showed almost the same ethanol productivity as genetically modified yeast.

Transporter engineering in biomass utilization by yeast.

Biomass resources are attractive carbon sources for bioproduction because of their sustainability. Many studies have been performed using biomass resources to produce sugars as carbon sources for cell factories. Expression of biomass hydrolyzing enzymes in cell factories is an important approach for constructing biomass-utilizing bioprocesses because external addition of these enzymes is expensive. In particular, yeasts have been extensively engineered to be cell factories that directly utilize biomass because of their manageable responses to many genetic engineering tools, such as gene expression, deletion and editing. Biomass utilizing bioprocesses have also been developed using these genetic engineering tools to construct metabolic pathways. However, sugar input and product output from these cells are critical factors for improving bioproduction along with biomass utilization and metabolic pathways. Transporters are key components for efficient input and output activities. In this review, we focus on transporter engineering in yeast to enhance bioproduction from biomass resources.

Kinetic modeling and sensitivity analysis of acetone-butanol-ethanol production.

A kinetic simulation model of metabolic pathways that describes the dynamic behaviors of metabolites in acetone-butanol-ethanol (ABE) production by Clostridium saccharoperbutylacetonicum N1-4 was proposed using a novel simulator WinBEST-KIT. This model was validated by comparing with experimental time-course data of metabolites in batch cultures over a wide range of initial glucose concentrations (36.1-295 mM). By introducing substrate inhibition, product inhibition of butanol, activation of butyrate and considering the cessation of metabolic reactions in the case of insufficiency of energy after glucose exhaustion, the revised model showed 0.901 of squared correlation coefficient (r(2)) between experimental time-course of metabolites and calculated ones. Thus, the final revised model is assumed to be one of the best candidates for kinetic simulation describing dynamic behavior of metabolites in ABE production. Sensitivity analysis revealed that 5% increase in reaction of reverse pathway of butyrate production (R(17)) and 5% decrease in reaction of CoA transferase for butyrate (R(15)) highly contribute to high production of butanol. These system analyses should be effective in the elucidation which pathway is metabolic bottleneck for high production of butanol.