Genome-Scale Metabolic Modelling of Lifestyle Changes in Rhizobium leguminosarum

Rhizobia are soil bacteria that induce nodule formation on plant roots and differentiate into nitrogen-fixing bacteroids. A detailed understanding of this complex symbiosis is essential for advancing ongoing efforts to engineer novel symbioses with cereal crops for sustainable agriculture.

Staphylococcus aureus Transcriptome Data and Metabolic Modelling Investigate the Interplay of Ser/Thr Kinase PknB, Its Phosphatase Stp, the glmR/yvcK Regulon and the cdaA Operon for Metabolic Adaptation

Serine/threonine kinase PknB and its corresponding phosphatase Stp are important regulators of many cell functions in the pathogen S. aureus. Genome-scale gene expression data of S. aureus strain NewHG (sigB+) elucidated their effect on physiological functions. Moreover, metabolic modelling from these data inferred metabolic adaptations. We compared wild-type to deletion strains lacking pknB, stp or both. Ser/Thr phosphorylation of target proteins by PknB switched amino acid catabolism off and gluconeogenesis on to provide the cell with sufficient components. We revealed a significant impact of PknB and Stp on peptidoglycan, nucleotide and aromatic amino acid synthesis, as well as catabolism involving aspartate transaminase. Moreover, pyrimidine synthesis was dramatically impaired by stp deletion but only slightly by functional loss of PknB. In double knockouts, higher activity concerned genes involved in peptidoglycan, purine and aromatic amino acid synthesis from glucose but lower activity of pyrimidine synthesis from glucose compared to the wild type. A second transcriptome dataset from S. aureus NCTC 8325 (sigB-) validated the predictions. For this metabolic adaptation, PknB was found to interact with CdaA and the yvcK/glmR regulon. The involved GlmR structure and the GlmS riboswitch were modelled. Furthermore, PknB phosphorylation lowered the expression of many virulence factors, and the study shed light on S. aureus infection processes.

Integrating genome-scale metabolic modelling and transfer learning for human gene regulatory network reconstruction

Motivation Gene regulation is responsible for controlling numerous physiological functions and dynamically responding to environmental fluctuations. Reconstructing the human network of gene regulatory interactions is thus paramount to understanding the cell functional organisation across cell types, as well as to elucidating pathogenic processes and identifying molecular drug targets. Although significant effort has been devoted towards this direction, existing computational methods mainly rely on gene expression levels, possibly ignoring the information conveyed by mechanistic biochemical knowledge. Moreover, except for a few recent attempts, most of the existing approaches only consider the information of the organism under analysis, without exploiting the information of related model organisms. Results We propose a novel method for the reconstruction of the human gene regulatory network, based on a transfer learning strategy that synergically exploits information from human and mouse, conveyed by gene-related metabolic features generated in-silico from gene expression data. Specifically, we learn a predictive model from metabolic activity inferred via tissue-specific metabolic modelling of artificial gene knockouts. Our experiments show that the combination of our transfer learning approach with the constructed metabolic features provides a significant advantage in terms of reconstruction accuracy, as well as additional clues on the contribution of each constructed metabolic feature. Availability The system, the datasets and all the results obtained in this study are available at: https://doi.org/10.6084/m9.figshare.c.5237687 Supplementary information Supplementary data are available at Bioinformatics online.

Integrative Gene Expression and Metabolic Analysis tool IgemRNA

Genome scale metabolic modelling is widely used technique to research metabolism impacts on organism properties. Additional omics data integration enables a more precise genotype-phenotype analysis for biotechnology, medicine and life sciences. Transcriptome data amounts rapidly increase each year. Many transcriptome analysis tools with integrated genome scale metabolic modelling are proposed. But these tools have own restrictions, compatibility issues and the necessity of previous experience and advanced user skills. We have analysed and classified published tools, summarized possible transcriptome pre-processing, and analysis methods and implemented them in the new transcriptome analysis tool IgemRNA. Tool novelty is the possibility of transcriptomics data pre-processing approach, analysis of transcriptome with or without genome scale metabolic models and different thresholding and gene mapping approach availability. In comparison with usual Gene set enrichment analysis methods, IgemRNA options provide additional transcriptome data validation, where minimal metabolic network connectivity and flux requirements are met. IgemRNA allows to process transcriptome datasets, compare data between different phenotypes, execute multiple analysis and data filtering functions. All this is done via graphical user interface. IgemRNA is compatible with Cobra Toolbox 3.0 and uses some of its functions for genome scale metabolic model optimization tasks. IgemRNA is open access software available at https://github.com/BigDataInSilicoBiologyGroup/IgemRNA.

Modelling hCDKL5 Heterologous Expression in Bacteria

hCDKL5 refers to the human cyclin-dependent kinase like 5 that is primarily expressed in the brain. Mutations in its coding sequence are often causative of hCDKL5 deficiency disorder, a devastating neurodevelopmental disorder currently lacking a cure. The large-scale recombinant production of hCDKL5 is desirable to boost the translation of preclinical therapeutic approaches into the clinic. However, this is hampered by the intrinsically disordered nature of almost two-thirds of the hCDKL5 sequence, making this region more susceptible to proteolytic attack, and the observed toxicity when the enzyme is accumulated in the cytoplasm of eukaryotic host cells. The bacterium Pseudoalteromonas haloplanktis TAC125 (PhTAC125) is the only prokaryotic host in which the full-length production of hCDKL5 has been demonstrated. To date, a system-level understanding of the metabolic burden imposed by hCDKL5 production is missing, although it would be crucial for upscaling of the production process. Here, we combined experimental data on protein production and nutrients assimilation with metabolic modelling to infer the global consequences of hCDKL5 production in PhTAC125 and to identify potential overproduction targets. Our analyses showed a remarkable accuracy of the model in simulating the recombinant strain phenotype and also identified priority targets for optimised protein production.

Modelling hCDKL5 heterologous expression in bacteria

hCDKL5 refers to the human cyclin-dependent kinase that is primarily expressed in the brain where it exerts its function in several neuron districts. Mutations in its coding sequence are often causative of hCDKL5 deficiency disorder. The large-scale recombinant production of hCDKL5 is desirable to boost the translation of current therapeutic approaches into the clinic. However, this is hampered by the following features: i) almost two-thirds of hCDKL5 sequence are predicted to be intrinsically disordered, making this region more susceptible to proteolytic attack; ii) the cytoplasmic accumulation of the enzyme in eukaryotic host cells is associated to toxicity. The bacterium Pseudoalteromonas haloplanktis TAC125 (PhTAC125) is the only prokaryotic host in which the full-length production of hCDKL5 has been demonstrated. To date, a system-level understanding of the metabolic burden imposed by hCDKL5 production is missing, although it would be crucial for the upscaling of the production process. Here, we have combined experimental data on protein production and nutrients assimilation with metabolic modelling to infer the global consequences of hCDKL5 production in PhTAC125 and to identify potential overproduction targets. Our analyses showed a remarkable accuracy of the model in simulating the recombinant strain phenotype and also identified priority targets for optimized protein production.

An evolutionary algorithm for designing microbial communities via environmental modification

Despite a growing understanding of how environmental composition affects microbial communities, it remains difficult to apply this knowledge to the rational design of synthetic multispecies consortia. This is because natural microbial communities can harbour thousands of different organisms and environmental substrates, making up a vast combinatorial space that precludes exhaustive experimental testing and computational prediction. Here, we present a method based on the combination of machine learning and metabolic modelling that selects optimal environmental compositions to produce target community phenotypes. In this framework, dynamic flux balance analysis is used to model the growth of a community in candidate environments. A genetic algorithm is then used to evaluate the behaviour of the community relative to a target phenotype, and subsequently adjust the environment to allow the organisms to approach this target. We apply this iterative process to thousands of in silico communities of varying sizes, showing how it can rapidly identify environments that yield desired taxonomic compositions and patterns of metabolic exchange. Moreover, this combination of approaches produces testable predictions for the assembly of experimental microbial communities with specific properties and can facilitate rational environmental design processes for complex microbiomes.