Biomedical applications of genome-scale metabolic network reconstructions of human pathogens

Reconstruction and Analysis of a Genome-Scale Metabolic Model of Acinetobacter lwoffii

Acinetobacter lwoffii is widely considered to be a harmful bacterium that is resistant to medicines and disinfectants. A. lwoffii NL1 degrades phenols efficiently and shows promise as an aromatic compound degrader in antibiotic-contaminated environments. To gain a comprehensive understanding of A. lwoffii, the first genome-scale metabolic model of A. lwoffii was constructed using semi-automated and manual methods. The iNX811 model, which includes 811 genes, 1071 metabolites, and 1155 reactions, was validated using 39 unique carbon and nitrogen sources. Genes and metabolites critical for cell growth were analyzed, and 12 essential metabolites (mainly in the biosynthesis and metabolism of glycan, lysine, and cofactors) were identified as antibacterial drug targets. Moreover, to explore the metabolic response to phenols, metabolic flux was simulated by integrating transcriptomics, and the significantly changed metabolism mainly included central carbon metabolism, along with some transport reactions. In addition, the addition of substances that effectively improved phenol degradation was predicted and validated using the model. Overall, the reconstruction and analysis of model iNX811 helped to study the antimicrobial systems and biodegradation behavior of A. lwoffii.

Genome-scale metabolic models in translational medicine: the current status and potential of machine learning in improving the effectiveness of the models

The genome-scale metabolic model (GEM) has emerged as one of the leading modeling approaches for systems-level metabolic studies and has been widely explored for a broad range of organisms and applications. Owing to the development of genome sequencing technologies and available biochemical data, it is possible to reconstruct GEMs for model and non-model microorganisms as well as for multicellular organisms such as humans and animal models. GEMs will evolve in parallel with the availability of biological data, new mathematical modeling techniques and the development of automated GEM reconstruction tools. The use of high-quality, context-specific GEMs, a subset of the original GEM in which inactive reactions are removed while maintaining metabolic functions in the extracted model, for model organisms along with machine learning (ML) techniques could increase their applications and effectiveness in translational research in the near future. Here, we briefly review the current state of GEMs, discuss the potential contributions of ML approaches for more efficient and frequent application of these models in translational research, and explore the extension of GEMs to integrative cellular models.

Genome-scale model of Pseudomonas aeruginosa metabolism unveils virulence and drug potentiation

Pseudomonas aeruginosa is one of the leading causes of hospital-acquired infections. To decipher the metabolic mechanisms associated with virulence and antibiotic resistance, we have developed an updated genome-scale model (GEM) of P. aeruginosa. The model (iSD1509) is an extensively curated, three-compartment, and mass-and-charge balanced BiGG model containing 1509 genes, the largest gene content for any P. aeruginosa GEM to date. It is the most accurate with prediction accuracies as high as 92.4% (gene essentiality) and 93.5% (substrate utilization). In iSD1509, we newly added a recently discovered pathway for ubiquinone-9 biosynthesis which is required for anaerobic growth. We used a modified iSD1509 to demonstrate the role of virulence factor (phenazines) in the pathogen survival within biofilm/oxygen-limited condition. Further, the model can mechanistically explain the overproduction of a drug susceptibility biomarker in the P. aeruginosa mutants. Finally, we use iSD1509 to demonstrate the drug potentiation by metabolite supplementation, and elucidate the mechanisms behind the phenotype, which agree with experimental results.

Comprehensive genome-scale metabolic model of the human pathogen Cryptococcus neoformans: A platform for understanding pathogen metabolism and identifying new drug targets

Introduction: The fungal priority pathogen Cryptococcus neoformans causes cryptococcal meningoencephalitis in immunocompromised individuals and leads to hundreds of thousands of deaths per year. The undesirable side effects of existing treatments, the need for long application times to prevent the disease from recurring, the lack of resources for these treatment methods to spread over all continents necessitate the search for new treatment methods. Methods: Genome-scale models have been shown to be valuable in studying the metabolism of many organisms. Here we present the first genome-scale metabolic model for C. neoformans, iCryptococcus. This comprehensive model consists of 1,270 reactions, 1,143 metabolites, 649 genes, and eight compartments. The model was validated, proving accurate when predicting the capability of utilizing different carbon and nitrogen sources and growth rate in comparison to experimental data. Results and Discussion: The compatibility of the in silico Cryptococcus metabolism under infection conditions was assessed. The steroid and amino acid metabolisms found in the essentiality analyses have the potential to be drug targets for the therapeutic strategies to be developed against Cryptococcus species. iCryptococcus model can be applied to explore new targets for antifungal drugs along with essential gene, metabolite and reaction analyses and provides a promising platform for elucidation of pathogen metabolism.

Systems Biology: New Insight into Antibiotic Resistance

Over the past few decades, antimicrobial resistance (AMR) has emerged as an important threat to public health, resulting from the global propagation of multidrug-resistant strains of various bacterial species. Knowledge of the intrinsic factors leading to this resistance is necessary to overcome these new strains. This has contributed to the increased use of omics technologies and their extrapolation to the system level. Understanding the mechanisms involved in antimicrobial resistance acquired by microorganisms at the system level is essential to obtain answers and explore options to combat this resistance. Therefore, the use of robust whole-genome sequencing approaches and other omics techniques such as transcriptomics, proteomics, and metabolomics provide fundamental insights into the physiology of antimicrobial resistance. To improve the efficiency of data obtained through omics approaches, and thus gain a predictive understanding of bacterial responses to antibiotics, the integration of mathematical models with genome-scale metabolic models (GEMs) is essential. In this context, here we outline recent efforts that have demonstrated that the use of omics technology and systems biology, as quantitative and robust hypothesis-generating frameworks, can improve the understanding of antibiotic resistance, and it is hoped that this emerging field can provide support for these new efforts.

Machine Learning Study of Metabolic Networks vs ChEMBL Data of Antibacterial Compounds

Antibacterial drugs (AD) change the metabolic status of bacteria, contributing to bacterial death. However, antibiotic resistance and the emergence of multidrug-resistant bacteria increase interest in understanding metabolic network (MN) mutations and the interaction of AD vs MN. In this study, we employed the IFPTML = Information Fusion (IF) + Perturbation Theory (PT) + Machine Learning (ML) algorithm on a huge dataset from the ChEMBL database, which contains >155,000 AD assays vs >40 MNs of multiple bacteria species. We built a linear discriminant analysis (LDA) and 17 ML models centered on the linear index and based on atoms to predict antibacterial compounds. The IFPTML-LDA model presented the following results for the training subset: specificity (Sp) = 76% out of 70,000 cases, sensitivity (Sn) = 70%, and Accuracy (Acc) = 73%. The same model also presented the following results for the validation subsets: Sp = 76%, Sn = 70%, and Acc = 73.1%. Among the IFPTML nonlinear models, the k nearest neighbors (KNN) showed the best results with Sn = 99.2%, Sp = 95.5%, Acc = 97.4%, and Area Under Receiver Operating Characteristic (AUROC) = 0.998 in training sets. In the validation series, the Random Forest had the best results: Sn = 93.96% and Sp = 87.02% (AUROC = 0.945). The IFPTML linear and nonlinear models regarding the ADs vs MNs have good statistical parameters, and they could contribute toward finding new metabolic mutations in antibiotic resistance and reducing time/costs in antibacterial drug research.

Exploring the metabolic landscape of pancreatic ductal adenocarcinoma cells using genome-scale metabolic modeling

Pancreatic ductal adenocarcinoma (PDAC) is a major research focus because of its poor therapy response and dismal prognosis. PDAC cells adapt their metabolism to the surrounding environment, often relying on diverse nutrient sources. Because traditional experimental techniques appear exhaustive to find a viable therapeutic strategy, a highly curated and omics-informed PDAC genome-scale metabolic model was reconstructed using patient-specific transcriptomics data. From the model-predictions, several new metabolic functions were explored as potential therapeutic targets in addition to the known metabolic hallmarks of PDAC. Significant downregulation in the peroxisomal beta oxidation pathway, flux modulation in the carnitine shuttle system, and upregulation in the reactive oxygen species detoxification pathway reactions were observed. These unique metabolic traits of PDAC were correlated with potential drug combinations targeting genes with poor prognosis in PDAC. Overall, this study provides a better understanding of the metabolic vulnerabilities in PDAC and will lead to novel effective therapeutic strategies.

Interrogation of Essentiality in the Reconstructed Haemophilus influenzae Metabolic Network Identifies Lipid Metabolism Antimicrobial Targets: Preclinical Evaluation of a FabH β-Ketoacyl-ACP Synthase Inhibitor

Expediting drug discovery to fight antibacterial resistance requires holistic approaches at system levels. In this study, we focused on the human-adapted pathogen Haemophilus influenzae, and by constructing a high-quality genome-scale metabolic model, we rationally identified new metabolic drug targets in this organism. Contextualization of available gene essentiality data within in silico predictions identified most genes involved in lipid metabolism as promising targets. We focused on the β-ketoacyl-acyl carrier protein synthase III FabH, responsible for catalyzing the first step in the FASII fatty acid synthesis pathway and feedback inhibition. Docking studies provided a plausible three-dimensional model of FabH in complex with the synthetic inhibitor 1-(5-(2-fluoro-5-(hydroxymethyl)phenyl)pyridin-2-yl)piperidine-4-acetic acid (FabHi). Validating our in silico predictions, FabHi reduced H. influenzae viability in a dose- and strain-dependent manner, and this inhibitory effect was independent of fabH gene expression levels. fabH allelic variation was observed among H. influenzae clinical isolates. Many of these polymorphisms, relevant for stabilization of the dimeric active form of FabH and/or activity, may modulate the inhibitory effect as part of a complex multifactorial process with the overall metabolic context emerging as a key factor tuning FabHi activity. Synergies with antibiotics were not observed and bacteria were not prone to develop resistance. Inhibitor administration during H. influenzae infection on a zebrafish septicemia infection model cleared bacteria without signs of host toxicity. Overall, we highlight the potential of H. influenzae metabolism as a source of drug targets, metabolic models as target-screening tools, and FASII targeting suitability to counteract this bacterial infection. IMPORTANCE Antimicrobial resistance drives the need of synergistically combined powerful computational tools and experimental work to accelerate target identification and drug development. Here, we present a high-quality metabolic model of H. influenzae and show its usefulness both as a computational framework for large experimental data set contextualization and as a tool to discover condition-independent drug targets. We focus on β-ketoacyl-acyl carrier protein synthase III FabH chemical inhibition by using a synthetic molecule with good synthetic and antimicrobial profiles that specifically binds to the active site. The mechanistic complexity of FabH inhibition may go beyond allelic variation, and the strain-dependent effect of the inhibitor tested supports the impact of metabolic context as a key factor driving bacterial cell behavior. Therefore, this study highlights the systematic metabolic evaluation of individual strains through computational frameworks to identify secondary metabolic hubs modulating drug response, which will facilitate establishing synergistic and/or more precise and robust antibacterial treatments.

Computational approaches to understanding Clostridioides difficile metabolism and virulence

The progress of infection by Clostridioides difficile is strongly influenced by metabolic cues it encounters as it colonizes the gastrointestinal tract. Both colonization and regulation of virulence have a multi-factorial interaction between host, microbiome, and gene expression cascades. While these connections with metabolism have been understood for some time, many mechanisms of control have remained difficult to directly assay due to high metabolic variability among C. difficile isolates and difficult genetic systems. Computational systems offer a means to interrogate structure of complex or noisy datasets and generate useful, tractable hypotheses to be tested in the laboratory. Recently, in silico techniques have provided powerful insights into metabolic elements of C. difficile infection ranging from virulence regulation to interactions with the gut microbiota. In this review, we introduce and provide context to the methods of computational modeling that have been applied to C. difficile metabolism and virulence thus far. The techniques discussed here have laid the foundation for future multi-scale efforts aimed at understanding the complex interplay of metabolic activity between pathogen, host, and surrounding microbial community in the regulation of C. difficile pathogenesis.

Bacterial metabolism and pathogenesis intimate intertwining: time for metabolic modelling to come into action

We take a snapshot of the recent understanding of bacterial metabolism and the bacterial-host metabolic interplay during infection, and highlight key outcomes and challenges for the practical implementation of bacterial metabolic modelling computational tools in the pathogenesis field.

Discovering Essential Multiple Gene Effects Through Large Scale Optimization: An Application to Human Cancer Metabolism

Computational modelling of metabolic processes has proven to be a useful approach to formulate our knowledge and improve our understanding of core biochemical systems that are crucial to maintaining cellular functions. Towards understanding the broader role of metabolism on cellular decision-making in health and disease conditions, it is important to integrate the study of metabolism with other core regulatory systems and omics within the cell, including gene expression patterns. After quantitatively integrating gene expression profiles with a genome-scale reconstruction of human metabolism, we propose a set of combinatorial methods to reverse engineer gene expression profiles and to find pairs and higher-order combinations of genetic modifications that simultaneously optimize multi-objective cellular goals. This enables us to suggest classes of transcriptomic profiles that are most suitable to achieve given metabolic phenotypes. We demonstrate how our techniques are able to compute beneficial, neutral or "toxic" combinations of gene expression levels. We test our methods on nine tissue-specific cancer models, comparing our outcomes with the corresponding normal cells, identifying genes as targets for potential therapies. Our methods open the way to a broad class of applications that require an understanding of the interplay among genotype, metabolism, and cellular behaviour, at scale.

Uncovering the Role of Metabolism in Oomycete-Host Interactions Using Genome-Scale Metabolic Models

Metabolism is the set of biochemical reactions of an organism that enables it to assimilate nutrients from its environment and to generate building blocks for growth and proliferation. It forms a complex network that is intertwined with the many molecular and cellular processes that take place within cells. Systems biology aims to capture the complexity of cells, organisms, or communities by reconstructing models based on information gathered by high-throughput analyses (omics data) and prior knowledge. One type of model is a genome-scale metabolic model (GEM) that allows studying the distributions of metabolic fluxes, i.e., the "mass-flow" through the network of biochemical reactions. GEMs are nowadays widely applied and have been reconstructed for various microbial pathogens, either in a free-living state or in interaction with their hosts, with the aim to gain insight into mechanisms of pathogenicity. In this review, we first introduce the principles of systems biology and GEMs. We then describe how metabolic modeling can contribute to unraveling microbial pathogenesis and host-pathogen interactions, with a specific focus on oomycete plant pathogens and in particular Phytophthora infestans. Subsequently, we review achievements obtained so far and identify and discuss potential pitfalls of current models. Finally, we propose a workflow for reconstructing high-quality GEMs and elaborate on the resources needed to advance a system biology approach aimed at untangling the intimate interactions between plants and pathogens.

Next-Generation Genome-Scale Metabolic Modeling through Integration of Regulatory Mechanisms

Genome-scale metabolic models (GEMs) are powerful tools for understanding metabolism from a systems-level perspective. However, GEMs in their most basic form fail to account for cellular regulation. A diverse set of mechanisms regulate cellular metabolism, enabling organisms to respond to a wide range of conditions. This limitation of GEMs has prompted the development of new methods to integrate regulatory mechanisms, thereby enhancing the predictive capabilities and broadening the scope of GEMs. Here, we cover integrative models encompassing six types of regulatory mechanisms: transcriptional regulatory networks (TRNs), post-translational modifications (PTMs), epigenetics, protein-protein interactions and protein stability (PPIs/PS), allostery, and signaling networks. We discuss 22 integrative GEM modeling methods and how these have been used to simulate metabolic regulation during normal and pathological conditions. While these advances have been remarkable, there remains a need for comprehensive and widespread integration of regulatory constraints into GEMs. We conclude by discussing challenges in constructing GEMs with regulation and highlight areas that need to be addressed for the successful modeling of metabolic regulation. Next-generation integrative GEMs that incorporate multiple regulatory mechanisms and their crosstalk will be invaluable for discovering cell-type and disease-specific metabolic control mechanisms.

Advances in flux balance analysis by integrating machine learning and mechanism-based models

The availability of multi-omics data sets and genome-scale metabolic models for various organisms provide a platform for modeling and analyzing genotype-to-phenotype relationships. Flux balance analysis is the main tool for predicting flux distributions in genome-scale metabolic models and various data-integrative approaches enable modeling context-specific network behavior. Due to its linear nature, this optimization framework is readily scalable to multi-tissue or -organ and even multi-organism models. However, both data and model size can hamper a straightforward biological interpretation of the estimated fluxes. Moreover, flux balance analysis simulates metabolism at steady-state and thus, in its most basic form, does not consider kinetics or regulatory events. The integration of flux balance analysis with complementary data analysis and modeling techniques offers the potential to overcome these challenges. In particular machine learning approaches have emerged as the tool of choice for data reduction and selection of most important variables in big data sets. Kinetic models and formal languages can be used to simulate dynamic behavior. This review article provides an overview of integrative studies that combine flux balance analysis with machine learning approaches, kinetic models, such as physiology-based pharmacokinetic models, and formal graphical modeling languages, such as Petri nets. We discuss the mathematical aspects and biological applications of these integrated approaches and outline challenges and future perspectives.

Addressing uncertainty in genome-scale metabolic model reconstruction and analysis

The reconstruction and analysis of genome-scale metabolic models constitutes a powerful systems biology approach, with applications ranging from basic understanding of genotype-phenotype mapping to solving biomedical and environmental problems. However, the biological insight obtained from these models is limited by multiple heterogeneous sources of uncertainty, which are often difficult to quantify. Here we review the major sources of uncertainty and survey existing approaches developed for representing and addressing them. A unified formal characterization of these uncertainties through probabilistic approaches and ensemble modeling will facilitate convergence towards consistent reconstruction pipelines, improved data integration algorithms, and more accurate assessment of predictive capacity.

Reconstruction of a Context-Specific Model Based on Genome-Scale Metabolic Simulation for Identification of Prochloraz Resistance Mechanisms in Penicillium digitatum

Penicillium digitatum is the most destructive postharvest pathogen of citrus fruits, causing substantial economic losses. Prochloraz-resistant strains have emerged due to overuse of imidazole fungicides in agriculture. To study the prochloraz resistance mechanisms at the system level, a genome-scale metabolic model (GEM, iPD1512) of P. digitatum was reconstructed and constrained based on context-specific transcriptome data of the prochloraz-resistant strain, PdF6, from our previous work, a newly sequenced, context-specific transcriptome result of the major facilitator superfamily transporter-encoding gene mfs2 knockout mutant PdF6Δmfs2, and experimentally derived growth rate data. Through the model, iPD1512, the processes of prochloraz resistance in P. digitatum were well simulated. In detail, the growth rates of both wild-type and mutant P. digitatum under different prochloraz concentrations were simulated using constraint-based reconstruction and analysis. The growth rates of the mutant strains (sterol regulatory element-binding protein-encoding gene sreA knockout mutant PdF6ΔsreA and PdF6Δmfs2) were calculated and confirmed to be consistent with the simulation results. Furthermore, correlations between genes and prochloraz resistance were predicted and showed a great difference when compared with correlation results based on p-values from the hypothesis testing used by comparative transcriptomics. To sum up, in contrast to traditional transcriptome analysis, the GEM provides a systemic and dynamic drug resistance mechanism, which might help to detect some key upstream regulatory genes, but with small expression changes, and might provide more efficient targets to control prochloraz-resistant P. digitatum.

Genome-Scale Metabolic Modeling for Unraveling Molecular Mechanisms of High Threat Pathogens

Pathogens give rise to a wide range of diseases threatening global health and hence drawing public health agencies' attention to establish preventative and curative solutions. Genome-scale metabolic modeling is ever increasingly used tool for biomedical applications including the elucidation of antibiotic resistance, virulence, single pathogen mechanisms and pathogen-host interaction systems. With this approach, the sophisticated cellular system of metabolic reactions inside the pathogens as well as between pathogen and host cells are represented in conjunction with their corresponding genes and enzymes. Along with essential metabolic reactions, alternate pathways and fluxes are predicted by performing computational flux analyses for the growth of pathogens in a very short time. The genes or enzymes responsible for the essential metabolic reactions in pathogen growth are regarded as potential drug targets, as a priori guide to researchers in the pharmaceutical field. Pathogens alter the key metabolic processes in infected host, ultimately the objective of these integrative constraint-based context-specific metabolic models is to provide novel insights toward understanding the metabolic basis of the acute and chronic processes of infection, revealing cellular mechanisms of pathogenesis, identifying strain-specific biomarkers and developing new therapeutic approaches including the combination drugs. The reaction rates predicted during different time points of pathogen development enable us to predict active pathways and those that only occur during certain stages of infection, and thus point out the putative drug targets. Among others, fatty acid and lipid syntheses reactions are recent targets of new antimicrobial drugs. Genome-scale metabolic models provide an improved understanding of how intracellular pathogens utilize the existing microenvironment of the host. Here, we reviewed the current knowledge of genome-scale metabolic modeling in pathogen cells as well as pathogen host interaction systems and the promising applications in the extension of curative strategies against pathogens for global preventative healthcare.

Novel antimicrobial development using genome-scale metabolic model of Gram-negative pathogens: a review

Antimicrobial resistance (AMR) threatens the effective prevention and treatment of a wide range of infections. Governments around the world are beginning to devote effort for innovative treatment development to treat these resistant bacteria. Systems biology methods have been applied extensively to provide valuable insights into metabolic processes at system level. Genome-scale metabolic models serve as platforms for constraint-based computational techniques which aid in novel drug discovery. Tools for automated reconstruction of metabolic models have been developed to support system level metabolic analysis. We discuss features of such software platforms for potential users to best fit their purpose of research. In this work, we focus to review the development of genome-scale metabolic models of Gram-negative pathogens and also metabolic network approach for identification of antimicrobial drugs targets.

Novel Approaches for Systems Biology of Metabolism-Oriented Pathogen-Human Interactions: A Mini-Review

Pathogenic microorganisms exploit host metabolism for sustained survival by rewiring its metabolic interactions. Therefore, several metabolic changes are induced in both pathogen and host cells in the course of infection. A systems-based approach to elucidate those changes includes the integrative use of genome-scale metabolic networks and molecular omics data, with the overall goal of better characterizing infection mechanisms for novel treatment strategies. This review focuses on novel aspects of metabolism-oriented systems-based investigation of pathogen-human interactions. The reviewed approaches are the generation of dual-omics data for the characterization of metabolic signatures of pathogen-host interactions, the reconstruction of pathogen-host integrated genome-scale metabolic networks, which has a high potential to be applied to pathogen-gut microbiota interactions, and the structure-based analysis of enzymes playing role in those interactions. The integrative use of those approaches will pave the way for the identification of novel biomarkers and drug targets for the prediction and prevention of infectious diseases.

An integrated computational and experimental study to investigate Staphylococcus aureus metabolism

Staphylococcus aureus is a metabolically versatile pathogen that colonizes nearly all organs of the human body. A detailed and comprehensive knowledge of staphylococcal metabolism is essential to understand its pathogenesis. To this end, we have reconstructed and experimentally validated an updated and enhanced genome-scale metabolic model of S. aureus USA300_FPR3757. The model combined genome annotation data, reaction stoichiometry, and regulation information from biochemical databases and previous strain-specific models. Reactions in the model were checked and fixed to ensure chemical balance and thermodynamic consistency. To further refine the model, growth assessment of 1920 nonessential mutants from the Nebraska Transposon Mutant Library was performed, and metabolite excretion profiles of important mutants in carbon and nitrogen metabolism were determined. The growth and no-growth inconsistencies between the model predictions and in vivo essentiality data were resolved using extensive manual curation based on optimization-based reconciliation algorithms. Upon intensive curation and refinements, the model contains 863 metabolic genes, 1379 metabolites (including 1159 unique metabolites), and 1545 reactions including transport and exchange reactions. To improve the accuracy and predictability of the model to environmental changes, condition-specific regulation information curated from the existing knowledgebase was incorporated. These critical additions improved the model performance significantly in capturing gene essentiality, substrate utilization, and metabolite production capabilities and increased the ability to generate model-based discoveries of therapeutic significance. Use of this highly curated model will enhance the functional utility of omics data, and therefore, serve as a resource to support future investigations of S. aureus and to augment staphylococcal research worldwide.

Integration of in vivo experiment with a newly developed model of Staphylococcus aureus metabolism helps explore its metabolic versatility. A multidisciplinary team led by Rajib Saha at the University of Nebraska developed a new genome-scale metabolic model of the multi-drug resistant pathogen S. aureus by combining genome annotation data, reaction stoichiometry, and condition- and mutant-specific regulations from biochemical databases and previous strain-specific models. Extensive manual curation and incorporation of newly generated experimental data on growth and metabolite production improved the accuracy and predictability of the model and increased its ability to generate model-based discoveries of therapeutic significance. Use of this highly curated model will enhance the functional utility of omics data and, therefore, serve as a resource to support future investigations of S. aureus and to augment staphylococcal research worldwide.

A White-Box Machine Learning Approach for Revealing Antibiotic Mechanisms of Action

Current machine learning techniques enable robust association of biological signals with measured phenotypes, but these approaches are incapable of identifying causal relationships. Here, we develop an integrated "white-box" biochemical screening, network modeling, and machine learning approach for revealing causal mechanisms and apply this approach to understanding antibiotic efficacy. We counter-screen diverse metabolites against bactericidal antibiotics in Escherichia coli and simulate their corresponding metabolic states using a genome-scale metabolic network model. Regression of the measured screening data on model simulations reveals that purine biosynthesis participates in antibiotic lethality, which we validate experimentally. We show that antibiotic-induced adenine limitation increases ATP demand, which elevates central carbon metabolism activity and oxygen consumption, enhancing the killing effects of antibiotics. This work demonstrates how prospective network modeling can couple with machine learning to identify complex causal mechanisms underlying drug efficacy.

In Vivo Gene Essentiality and Metabolism in Bordetella pertussis

Bordetella pertussis is the causative agent of whooping cough, a serious respiratory illness affecting children and adults, associated with prolonged cough and potential mortality. Whooping cough has reemerged in recent years, emphasizing a need for increased knowledge of basic mechanisms of B. pertussis growth and pathogenicity. While previous studies have provided insight into in vitro gene essentiality of this organism, very little is known about in vivo gene essentiality, a critical gap in knowledge, since B. pertussis has no previously identified environmental reservoir and is isolated from human respiratory tract samples. We hypothesize that the metabolic capabilities of B. pertussis are especially tailored to the respiratory tract and that many of the genes involved in B. pertussis metabolism would be required to establish infection in vivo In this study, we generated a diverse library of transposon mutants and then used it to probe gene essentiality in vivo in a murine model of infection. Using the CON-ARTIST pipeline, 117 genes were identified as conditionally essential at 1 day postinfection, and 169 genes were identified as conditionally essential at 3 days postinfection. Most of the identified genes were associated with metabolism, and we utilized two existing genome-scale metabolic network reconstructions to probe the effects of individual essential genes on biomass synthesis. This analysis suggested a critical role for glucose metabolism and lipooligosaccharide biosynthesis in vivo This is the first genome-wide evaluation of in vivo gene essentiality in B. pertussis and provides tools for future exploration.IMPORTANCE Our study describes the first in vivo transposon sequencing (Tn-seq) analysis of B. pertussis and identifies genes predicted to be essential for in vivo growth in a murine model of intranasal infection, generating key resources for future investigations into B. pertussis pathogenesis and vaccine design.

Modeling Antibacterial Activity with Machine Learning and Fusion of Chemical Structure Information with Microorganism Metabolic Networks

Predicting the activity of new chemical compounds over pathogenic microorganisms with different metabolic reaction networks (MRN _s) is an important goal due to the different susceptibility to antibiotics. The ChEMBL database contains >160 000 outcomes of preclinical assays of antimicrobial activity for 55 931 compounds with >365 parameters of activity (MIC, IC₅₀, etc.) and >90 bacteria strains of >25 bacterial species. In addition, the Leong and Barabàsi data set includes >40 MRNs of microorganisms. However, there are no models able to predict antibacterial activity for multiple assays considering both drug and MRN structures at the same time. In this work, we combined perturbation theory, machine learning, and information fusion techniques to develop the first PTMLIF model. The best linear model found presented values of specificity = 90.31/90.40 and sensitivity = 88.14/88.07 in training/validation series. We carried out a comparison to nonlinear artificial neural network (ANN) techniques and previous models from the literature. Next, we illustrated the practical use of the model with an experimental case of study. We reported for the first time the isolation and characterization of terpenes from the plant Cissus incisa. The antibacterial activity of the terpenes was experimentally determined. The more active compounds were phytol and α-amyrin, with MIC = 100 μg/mL for Vancomycin-resistant Enterococcus faecium and Acinetobacter baumannii resistant to carbapenems. These compounds are already known from other sources. However, they have been isolated and evaluated for the first time here against several strains of multidrug-resistant bacteria including World Health Organization (WHO) priority pathogens. Last, we used the model to predict the activity of these compounds versus other microorganisms with different MRNs in order to find other potential targets.

Constraints-based analysis identifies NAD+ recycling through metabolic reprogramming in antibiotic resistant Chromobacterium violaceum

In the post genomic era, high throughput data augment stoichiometric flux balance models to compute accurate metabolic flux states, growth and energy phenotypes. Investigating altered metabolism in the context of evolved resistant genotypes potentially provide simple strategies to overcome drug resistance and induce susceptibility to existing antibiotics. A genome-scale metabolic model (GSMM) for Chromobacterium violaceum, an opportunistic human pathogen, was reconstructed using legacy data. Experimental constraints were used to represent antibiotic susceptible and resistant populations. Model predictions were validated using growth and respiration data successfully. Differential flux distribution and metabolic reprogramming were identified as a response to antibiotics, chloramphenicol and streptomycin. Streptomycin resistant populations (StrpR) redirected tricarboxylic acid (TCA) cycle flux through the glyoxylate shunt. Chloramphenicol resistant populations (ChlR) resorted to overflow metabolism producing acetate and formate. This switch to fermentative metabolism is potentially through excess reducing equivalents and increased NADH/NAD ratios. Reduced proton gradients and changed Proton Motive Force (PMF) induced by antibiotics were also predicted and verified experimentally using flow cytometry based membrane potential measurements. Pareto analysis of NADH and ATP maintenance showed the decoupling of electron transfer and ATP synthesis in StrpR. Redox homeostasis and NAD+ cycling through rewiring metabolic flux was implicated in re-sensitizing antibiotic resistant C. violaceum. These approaches can be used to probe metabolic vulnerabilities of resistant pathogens. On the verge of a post-antibiotic era, we foresee a critical need for systems level understanding of pathogens and host interaction to extend shelf life of antibiotics and strategize novel therapies.

Integrated Experimental and Computational Analyses Reveal Differential Metabolic Functionality in Antibiotic-Resistant Pseudomonas aeruginosa

Metabolic adaptations accompanying the development of antibiotic resistance in bacteria remain poorly understood. To study this relationship, we profiled the growth of lab-evolved antibiotic-resistant lineages of the opportunistic pathogen Pseudomonas aeruginosa across 190 unique carbon sources. Our data revealed that the evolution of antibiotic resistance resulted in systems-level changes to growth dynamics and metabolic phenotype. A genome-scale metabolic network reconstruction of P. aeruginosa was paired with whole-genome sequencing data to predict genes contributing to observed changes in metabolism. We experimentally validated computational predictions to identify mutations in resistant P. aeruginosa affecting loss of catabolic function. Finally, we found a shared metabolic phenotype between lab-evolved P. aeruginosa and clinical isolates with similar mutational landscapes. Our results build upon previous knowledge of antibiotic-induced metabolic adaptation and provide a framework for the identification of metabolic limitations in antibiotic-resistant pathogens.