Accessory subunit NDUFB4 participates in mitochondrial complex I supercomplex formation

Mitochondrial electron transport chain complexes organize into supramolecular structures called respiratory supercomplexes (SCs). The role of respiratory SCs remains largely unconfirmed despite evidence supporting their necessity for mitochondrial respiratory function. The mechanisms underlying the formation of the I1III2IV1 "respirasome" SC are also not fully understood, further limiting insights into these processes in physiology and diseases, including neurodegeneration and metabolic syndromes. NDUFB4 is a complex I accessory subunit that contains residues that interact with the subunit UQCRC1 from complex III, suggesting that NDUFB4 is integral for I1III2IV1 respirasome integrity. Here, we introduced specific point mutations to Asn24 (N24) and Arg30 (R30) residues on NDUFB4 to decipher the role of I1III2-containing respiratory SCs in cellular metabolism while minimizing the functional consequences to complex I assembly. Our results demonstrate that NDUFB4 point mutations N24A and R30A impair I1III2IV1 respirasome assembly and reduce mitochondrial respiratory flux. Steady-state metabolomics also revealed a global decrease in citric acid cycle metabolites, affecting NADH-generating substrates. Taken together, our findings highlight an integral role of NDUFB4 in respirasome assembly and demonstrate the functional significance of SCs in regulating mammalian cell bioenergetics.

The SARS-CoV-2 spike glycoprotein interacts with MAO-B and impairs mitochondrial energetics

SARS-CoV-2 infection is associated with both acute and post-acute neurological symptoms. Emerging evidence suggests that SARS-CoV-2 can alter mitochondrial metabolism, suggesting that changes in brain metabolism may contribute to the development of acute and post-acute neurological complications. Monoamine oxidase B (MAO-B) is a flavoenzyme located on the outer mitochondrial membrane that catalyzes the oxidative deamination of monoamine neurotransmitters. Computational analyses have revealed high similarity between the SARS-CoV-2 spike glycoprotein receptor binding domain on the ACE2 receptor and MAO-B, leading to the hypothesis that SARS-CoV-2 spike glycoprotein may alter neurotransmitter metabolism by interacting with MAO-B. Our results empirically establish that the SARS-CoV-2 spike glycoprotein interacts with MAO-B, leading to increased MAO-B activity in SH-SY5Y neuron-like cells. Common to neurodegenerative disease pathophysiological mechanisms, we also demonstrate that the spike glycoprotein impairs mitochondrial bioenergetics, induces oxidative stress, and perturbs the degradation of depolarized aberrant mitochondria through mitophagy. Our findings also demonstrate that SH-SY5Y neuron-like cells expressing the SARS-CoV-2 spike protein were more susceptible to MPTP-induced necrosis, likely necroptosis. Together, these results reveal novel mechanisms that may contribute to SARS-CoV-2-induced neurodegeneration.

Insights into the Aggressiveness of the Emerging North American Population 3 (NA3) of Fusarium graminearum

In the United States and Canada, Fusarium graminearum (Fg) is the predominant etiological agent of Fusarium head blight (FHB), an economically devastating fungal disease of wheat and other small grains. Besides yield losses, FHB leads to grain contamination with trichothecene mycotoxins that are harmful to plant, human, and livestock health. Three genetic North American populations of Fg, differing in their predominant trichothecene chemotype (i.e., NA1/15ADON, NA2/3ADON, and NA3/NX-2), have been identified. To improve our understanding of the newly discovered population NA3 and how population-level diversity influences FHB outcomes, we inoculated heads of the moderately resistant wheat cultivar Alsen with 15 representative strains from each population and evaluated disease progression, mycotoxin accumulation, and mycotoxin production per unit Fg biomass. Additionally, we evaluated population-specific differences in induced host defense responses. The NA3 population was significantly less aggressive than the NA1 and NA2 populations but posed a similar mycotoxigenic potential. Multiomics analyses revealed patterns in mycotoxin production per unit Fg biomass, expression of Fg aggressiveness-associated genes, and host defense responses that did not always correlate with the NA3-specific severity difference. Our comparative disease assay of NA3/NX-2 and admixed NA1/NX-2 strains indicated that the reduced NA3 aggressiveness is not due solely to the NX-2 chemotype. Notably, the NA1 and NA2 populations did not show a significant advantage over NA3 in perithecia production, a fitness-related trait. Together, our data highlight that the disease outcomes were not due to mycotoxin production or host defense alone, indicating that other virulence factors and/or host defense mechanisms are likely involved.

Integrative metabolomics science in Alzheimer's disease: Relevance and future perspectives

Alzheimer's disease (AD) is determined by various pathophysiological mechanisms starting 10-25 years before the onset of clinical symptoms. As multiple functionally interconnected molecular/cellular pathways appear disrupted in AD, the exploitation of high-throughput unbiased omics sciences is critical to elucidating the precise pathogenesis of AD. Among different omics, metabolomics is a fast-growing discipline allowing for the simultaneous detection and quantification of hundreds/thousands of perturbed metabolites in tissues or biofluids, reproducing the fluctuations of multiple networks affected by a disease. Here, we seek to critically depict the main metabolomics methodologies with the aim of identifying new potential AD biomarkers and further elucidating AD pathophysiological mechanisms. From a systems biology perspective, as metabolic alterations can occur before the development of clinical signs, metabolomics - coupled with existing accessible biomarkers used for AD screening and diagnosis - can support early disease diagnosis and help develop individualized treatment plans. Presently, the majority of metabolomic analyses emphasized that lipid metabolism is the most consistently altered pathway in AD pathogenesis. The possibility that metabolomics may reveal crucial steps in AD pathogenesis is undermined by the difficulty in discriminating between the causal or epiphenomenal or compensatory nature of metabolic findings.

DAPTEV: Deep aptamer evolutionary modelling for COVID-19 drug design

Typical drug discovery and development processes are costly, time consuming and often biased by expert opinion. Aptamers are short, single-stranded oligonucleotides (RNA/DNA) that bind to target proteins and other types of biomolecules. Compared with small-molecule drugs, aptamers can bind to their targets with high affinity (binding strength) and specificity (uniquely interacting with the target only). The conventional development process for aptamers utilizes a manual process known as Systematic Evolution of Ligands by Exponential Enrichment (SELEX), which is costly, slow, dependent on library choice and often produces aptamers that are not optimized. To address these challenges, in this research, we create an intelligent approach, named DAPTEV, for generating and evolving aptamer sequences to support aptamer-based drug discovery and development. Using the COVID-19 spike protein as a target, our computational results suggest that DAPTEV is able to produce structurally complex aptamers with strong binding affinities.

rAAV Manufacturing: The Challenges of Soft Sensing during Upstream Processing

Recombinant adeno-associated virus (rAAV) is the most effective viral vector technology for directly translating the genomic revolution into medicinal therapies. However, the manufacturing of rAAV viral vectors remains challenging in the upstream processing with low rAAV yield in large-scale production and high cost, limiting the generalization of rAAV-based treatments. This situation can be improved by real-time monitoring of critical process parameters (CPP) that affect critical quality attributes (CQA). To achieve this aim, soft sensing combined with predictive modeling is an important strategy that can be used for optimizing the upstream process of rAAV production by monitoring critical process variables in real time. However, the development of soft sensors for rAAV production as a fast and low-cost monitoring approach is not an easy task. This review article describes four challenges and critically discusses the possible solutions that can enable the application of soft sensors for rAAV production monitoring. The challenges from a data scientist's perspective are (i) a predictor variable (soft-sensor inputs) set without AAV viral titer, (ii) multi-step forecasting, (iii) multiple process phases, and (iv) soft-sensor development composed of the mechanistic model.

Diversifying Design of Nucleic Acid Aptamers Using Unsupervised Machine Learning

Inverse design of short single-stranded RNA and DNA sequences (aptamers) is the task of finding sequences that satisfy a set of desired criteria. Relevant criteria may be, for example, the presence of specific folding motifs, binding to molecular ligands, sensing properties, and so on. Most practical approaches to aptamer design identify a small set of promising candidate sequences using high-throughput experiments (e.g., SELEX) and then optimize performance by introducing only minor modifications to the empirically found candidates. Sequences that possess the desired properties but differ drastically in chemical composition will add diversity to the search space and facilitate the discovery of useful nucleic acid aptamers. Systematic diversification protocols are needed. Here we propose to use an unsupervised machine learning model known as the Potts model to discover new, useful sequences with controllable sequence diversity. We start by training a Potts model using the maximum entropy principle on a small set of empirically identified sequences unified by a common feature. To generate new candidate sequences with a controllable degree of diversity, we take advantage of the model's spectral feature: an "energy" bandgap separating sequences that are similar to the training set from those that are distinct. By controlling the Potts energy range that is sampled, we generate sequences that are distinct from the training set yet still likely to have the encoded features. To demonstrate performance, we apply our approach to design diverse pools of sequences with specified secondary structure motifs in 30-mer RNA and DNA aptamers.

Machine Learning and Hybrid Methods for Metabolic Pathway Modeling

Computational cell metabolism models seek to provide metabolic explanations of cell behavior under different conditions or following genetic alterations, help in the optimization of in vitro cell growth environments, or predict cellular behavior in vivo and in vitro. In the extremes, mechanistic models can include highly detailed descriptions of a small number of metabolic reactions or an approximate representation of an entire metabolic network. To date, all mechanistic models have required details of individual metabolic reactions, either kinetic parameters or metabolic flux, as well as information about extracellular and intracellular metabolite concentrations. Despite the extensive efforts and the increasing availability of high-quality data, required in vivo data are not available for the majority of known metabolic reactions; thus, mechanistic models are based primarily on ex vivo kinetic measurements and limited flux information. Machine learning approaches provide an alternative for derivation of functional dependencies from existing data. The increasing availability of metabolomic and lipidomic data, with growing feature coverage as well as sample set size, is expected to provide new data options needed for derivation of machine learning models of cell metabolic processes. Moreover, machine learning analysis of longitudinal data can lead to predictive models of cell behaviors over time. Conversely, machine learning models trained on steady-state data can provide descriptive models for the comparison of metabolic states in different environments or disease conditions. Additionally, inclusion of metabolic network knowledge in these analyses can further help in the development of models with limited data.This chapter will explore the application of machine learning to the modeling of cell metabolism. We first provide a theoretical explanation of several machine learning and hybrid mechanistic machine learning methods currently being explored to model metabolism. Next, we introduce several avenues for improving these models with machine learning. Finally, we provide protocols for specific examples of the utilization of machine learning in the development of predictive cell metabolism models using metabolomic data. We describe data preprocessing, approaches for training of machine learning models for both descriptive and predictive models, and the utilization of these models in synthetic and systems biology. Detailed protocols provide a list of software tools and libraries used for these applications, step-by-step modeling protocols, troubleshooting, as well as an overview of existing limitations to these approaches.

Exercise training enhances muscle mitochondrial metabolism in diet-resistant obesity

Background

Current paradigms for predicting weight loss in response to energy restriction have general validity but a subset of individuals fail to respond adequately despite documented diet adherence. Patients in the bottom 20% for rate of weight loss following a hypocaloric diet (diet-resistant) have been found to have less type I muscle fibres and lower skeletal muscle mitochondrial function, leading to the hypothesis that physical exercise may be an effective treatment when diet alone is inadequate. In this study, we aimed to assess the efficacy of exercise training on mitochondrial function in women with obesity with a documented history of minimal diet-induced weight loss.

Methods

From over 5000 patient records, 228 files were reviewed to identify baseline characteristics of weight loss response from women with obesity who were previously classified in the top or bottom 20% quintiles based on rate of weight loss in the first 6 weeks during which a 900 kcal/day meal replacement was consumed. A subset of 20 women with obesity were identified based on diet-resistance (n=10) and diet sensitivity (n=10) to undergo a 6-week supervised, progressive, combined aerobic and resistance exercise intervention.

Findings

Diet-sensitive women had lower baseline adiposity, higher fasting insulin and triglycerides, and a greater number of ATP-III criteria for metabolic syndrome. Conversely in diet-resistant women, the exercise intervention improved body composition, skeletal muscle mitochondrial content and metabolism, with minimal effects in diet-sensitive women. In-depth analyses of muscle metabolomes revealed distinct group- and intervention- differences, including lower serine-associated sphingolipid synthesis in diet-resistant women following exercise training.

Interpretation

Exercise preferentially enhances skeletal muscle metabolism and improves body composition in women with a history of minimal diet-induced weight loss. These clinical and metabolic mechanism insights move the field towards better personalised approaches for the treatment of distinct obesity phenotypes.

Funding

Canadian Institutes of Health Research (CIHR-INMD and FDN-143278; CAN-163902; CIHR PJT-148634).

Decrease of Intracellular Glutamine by STF-62247 Results in the Accumulation of Lipid Droplets in von Hippel-Lindau Deficient Cells

Kidney cancer is one of the top ten cancer diagnosed worldwide and its incidence has increased the last 20 years. Clear Cell Renal Cell Carcinoma (ccRCC) are characterized by mutations that inactivate the von Hippel-Lindau (VHL) tumor suppressor gene and evidence indicated alterations in metabolic pathways, particularly in glutamine metabolism. We previously identified a small molecule, STF-62247, which target VHL-deficient renal tumors by affecting late-stages of autophagy and lysosomal signaling. In this study, we investigated ccRCC metabolism in VHL-deficient and proficient cells exposed to the small molecule. Metabolomics profiling using 1H NMR demonstrated that STF-62247 increases levels of glucose, pyruvate, glycerol 3-phosphate while glutamate, asparagine, and glutathione significantly decreased. Diminution of glutamate and glutamine was further investigated using mass spectrometry, western blot analyses, enzymatic activities, and viability assays. We found that expression of SLC1A5 increases in VHL-deficient cells treated with STF-62247, possibly to stimulate glutamine uptake intracellularly to counteract the diminution of this amino acid. However, exogenous addition of glutamine was not able to rescue cell viability induced by the small molecule. Instead, our results showed that VHL-deficient cells utilize glutamine to produce fatty acid in response to STF-62247. Surprisingly, this occurs through oxidative phosphorylation in STF-treated cells while control cells use reductive carboxylation to sustain lipogenesis. We also demonstrated that STF-62247 stimulated expression of stearoyl-CoA desaturase (SCD1) and peripilin2 (PLIN2) to generate accumulation of lipid droplets in VHL-deficient cells. Moreover, the carnitine palmitoyltransferase 1A (CPT1A), which control the entry of fatty acid into mitochondria for β-oxidation, also increased in response to STF-62247. CPT1A overexpression in ccRCC is known to limit tumor growth. Together, our results demonstrated that STF-62247 modulates cellular metabolism of glutamine, an amino acid involved in the autophagy-lysosome process, to support lipogenesis, which could be implicated in the signaling driving to cell death.

BATL: Bayesian annotations for targeted lipidomics

MOTIVATION

Bioinformatic tools capable of annotating, rapidly and reproducibly, large, targeted lipidomic datasets are limited. Specifically, few programs enable high-throughput peak assessment of liquid chromatography-electrospray ionization tandem mass spectrometry data acquired in either selected or multiple reaction monitoring modes.

RESULTS

We present here Bayesian Annotations for Targeted Lipidomics, a Gaussian naïve Bayes classifier for targeted lipidomics that annotates peak identities according to eight features related to retention time, intensity, and peak shape. Lipid identification is achieved by modeling distributions of these eight input features across biological conditions and maximizing the joint posterior probabilities of all peak identities at a given transition. When applied to sphingolipid and glycerophosphocholine selected reaction monitoring datasets, we demonstrate over 95% of all peaks are rapidly and correctly identified.

AVAILABILITY AND IMPLEMENTATION

BATL software is freely accessible online at https://complimet.ca/batl/ and is compatible with Safari, Firefox, Chrome and Edge.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Relative Ratios of Human Seasonal Coronavirus Antibodies Predict the Efficiency of Cross-Neutralization of SARS-CoV-2 Spike Binding to ACE2

BACKGROUND

Antibodies raised against human seasonal coronaviruses (sCoVs), which are responsible for the common cold, are known to cross-react with SARS-CoV-2 antigens. This prompts questions about their protective role against SARS-CoV-2 infections and COVID-19 severity. However, the relationship between sCoVs exposure and SARS-CoV-2 correlates of protection are not clearly identified.

METHODS

We performed a cross-sectional analysis of cross-reactivity and cross-neutralization to SARS-CoV-2 antigens (S-RBD, S-trimer, N) using pre-pandemic sera from four different groups: pediatrics and adolescents, individuals 21 to 70 years of age, older than 70 years of age, and individuals living with HCV or HIV. Data was then further analysed using machine learning to identify predictive patterns of neutralization based on sCoVs serology.

FINDINGS

Antibody cross-reactivity to SARS-CoV-2 antigens varied between 1.6% and 15.3% depending on the cohort and the isotype-antigen pair analyzed. We also show a range of neutralizing activity (0-45%) with median inhibition ranging from 17.6 % to 23.3 % in serum that interferes with SARS-CoV-2 spike attachment to ACE2 independently of age group. While the abundance of sCoV antibodies did not directly correlate with neutralization, we show that neutralizing activity is rather dependent on relative ratios of IgGs in sera directed to all four sCoV spike proteins. More specifically, we identified antibodies to NL63 and OC43 as being the most important predictors of neutralization.

INTERPRETATION

Our data support the concept that exposure to sCoVs triggers antibody responses that influence the efficiency of SARS-CoV-2 spike binding to ACE2, which may potentially impact COVID-19 disease severity through other latent variables.

FUNDING

This study was supported by a grant by the CIHR (VR2 -172722) and by a grant supplement by the CITF, and by a NRC Collaborative R&D Initiative Grant (PR031-1).

SMILE: systems metabolomics using interpretable learning and evolution

Background

Direct link between metabolism and cell and organism phenotype in health and disease makes metabolomics, a high throughput study of small molecular metabolites, an essential methodology for understanding and diagnosing disease development and progression. Machine learning methods have seen increasing adoptions in metabolomics thanks to their powerful prediction abilities. However, the “black-box” nature of many machine learning models remains a major challenge for wide acceptance and utility as it makes the interpretation of decision process difficult. This challenge is particularly predominant in biomedical research where understanding of the underlying decision making mechanism is essential for insuring safety and gaining new knowledge.

Results

In this article, we proposed a novel computational framework, Systems Metabolomics using Interpretable Learning and Evolution (SMILE), for supervised metabolomics data analysis. Our methodology uses an evolutionary algorithm to learn interpretable predictive models and to identify the most influential metabolites and their interactions in association with disease. Moreover, we have developed a web application with a graphical user interface that can be used for easy analysis, interpretation and visualization of the results. Performance of the method and utilization of the web interface is shown using metabolomics data for Alzheimer’s disease.

Conclusions

SMILE was able to identify several influential metabolites on AD and to provide interpretable predictive models that can be further used for a better understanding of the metabolic background of AD. SMILE addresses the emerging issue of interpretability and explainability in machine learning, and contributes to more transparent and powerful applications of machine learning in bioinformatics.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-021-04209-1.

Metabolomics and computational analysis of the role of monoamine oxidase activity in delirium and SARS-COV-2 infection

Delirium is an acute change in attention and cognition occurring in ~ 65% of severe SARS-CoV-2 cases. It is also common following surgery and an indicator of brain vulnerability and risk for the development of dementia. In this work we analyzed the underlying role of metabolism in delirium-susceptibility in the postoperative setting using metabolomic profiling of cerebrospinal fluid and blood taken from the same patients prior to planned orthopaedic surgery. Distance correlation analysis and Random Forest (RF) feature selection were used to determine changes in metabolic networks. We found significant concentration differences in several amino acids, acylcarnitines and polyamines linking delirium-prone patients to known factors in Alzheimer's disease such as monoamine oxidase B (MAOB) protein. Subsequent computational structural comparison between MAOB and angiotensin converting enzyme 2 as well as protein-protein docking analysis showed that there potentially is strong binding of SARS-CoV-2 spike protein to MAOB. The possibility that SARS-CoV-2 influences MAOB activity leading to the observed neurological and platelet-based complications of SARS-CoV-2 infection requires further investigation.

NMR analysis of the correlation of metabolic changes in blood and cerebrospinal fluid in Alzheimer model male and female mice

The development of effective therapies as well as early, molecular diagnosis of Alzheimer's disease is impeded by the lack of understanding of the underlying pathological mechanisms. Metabolomics studies of body fluids as well as brain tissues have shown major changes in metabolic profiles of Alzheimer's patients. However, with analysis performed at the late stages of the disease it is not possible to distinguish causes and consequence. The mouse model APP/PS1 expresses a mutant amyloid precursor protein resulting in early Amyloid β (Aβ) accumulation as well as many resulting physiological changes including changes in metabolic profile and metabolism. Analysis of metabolic profile of cerebrospinal fluid (CSF) and blood of APP/PS1 mouse model can provide information about metabolic changes in these body fluids caused by Aβ accumulation. Using our novel method for analysis of correlation and mathematical ranking of significant correlations between metabolites in CSF and blood, we have explored changes in metabolite correlation and connectedness in APP/PS1 and wild type mice. Metabolites concentration and correlation changes in CSF, blood and across the blood brain barrier determined in this work are affected by the production of amyloid plaque. Metabolite changes observed in the APP/PS1 mouse model are the response to the mutation causing plaque formation, not the cause for the plaque suggesting that they are less relevant in the context of early treatment and prevention then the metabolic changes observed only in humans.

A phenylpropanoid diglyceride associates with the leaf rust resistance Lr34res gene in wheat

The gene Lr34res is one of the most long-lasting sources of quantitative fungal resistance in wheat. It is shown to be effective against leaf, stem, and stripe rusts, as well as powdery mildew and spot blotch. Recent biochemical characterizations of the encoded ABC transporter have outlined a number of allocrites, including phospholipids and abscisic acid, consistent with the established general promiscuity of ABC transporters, but ultimately leaving its mechanism of rust resistance unclear. Working with flag leaves of Triticum aestivum L. variety 'Thatcher' (Tc) and a near-isogenic line of 'Thatcher' into which the Lr34res allele was introgressed (Tc+Lr34res; RL6058), a comparative semi-targeted metabolomics analysis of flavonoid-rich extracts revealed virtually identical profiles with the exception of one metabolite accumulating in Tc+Lr34res, which was not present at comparable levels in Tc. Structural characterization of the purified metabolite revealed a phenylpropanoid diglyceride structure, 1-O-p-coumaroyl-3-O-feruloylglycerol (CFG). Additional profiling of CFG across a collection of near-isogenic lines and representative Lr34 haplotypes highlighted a broad association between the presence of Lr34res and elevated accumulations of CFG. Depletion of CFG upon infection, juxtaposed to its relatively lower anti-fungal activity, suggests CFG may serve as a storage form of the more potent anti-microbial hydroxycinnamic acids that are accessed during defense responses. Altogether these findings suggest a role for the encoded LR34res ABC transporter in modifying the accumulation of CFG, leading to increased accumulation of anti-fungal metabolites, essentially priming the wheat plant for defense.

Structurally Related Edaravone Analogues: Synthesis, Antiradical, Antioxidant, and Copper-Chelating Properties

Background::

The current therapeutic options available to patients diagnosed with Amyotrophic Lateral Sclerosis (ALS) are limited and edaravone is a compound that has gained significant interest for its therapeutic potential in this condition.

Objectives: :

The current work was thus undertaken to synthesize and characterize a series of edaravone analogues.

Methods:

A total of 17 analogues were synthesized and characterized for their antioxidant properties, radical scavenging potential and copper-chelating capabilities.

Results:

Radical scavenging and copper-chelating properties were notably observed for edaravone. Analogues bearing hydrogen in position 1 and a phenyl at position 3 and a phenyl in both positions of pyrazol-5 (4H)-one displayed substantial radical scavenging, antioxidants and copper-chelating properties. High accessibility of electronegative groups combined with higher electronegativity and partial charge of the carbonyl moiety in edaravone might explain the observed difference in the activity of edaravone relative to the closely related analogues 6 and 7 bearing hydrogen at position 1 and a phenyl at position 3 (6) and a phenyl in both positions (7).

Conclusion:

Overall, this study reveals a subset of edaravone analogues with interesting properties. Further investigation of these compounds is foreseen in relevant models of oxidative stress-associated diseases in order to assess their therapeutic potential in such conditions.

PROAFTN Classifier for Feature Selection with Application to Alzheimer Metabolomics Data Analysis

Early and accurate Alzheimer’s disease (AD) diagnosis remains a challenge. Recently, increasing efforts have been focused towards utilization of metabolomics data for the discovery of biomarkers for screening and diagnosis of AD. Several machine learning approaches were explored for classifying the blood metabolomics profiles of cognitively healthy and AD patients. Differentiation between AD, mild cognitive impairment (MCI) and cognitively healthy subjects remains difficult. In this paper, we propose a new machine learning approach for the selection of a subset of features that provide an improvement in classification rates between these three levels of cognitive disorders. Our experimental results demonstrate that utilization of these selected metabolic markers improves the performance of several classifiers in comparison to the classification accuracy obtained for the complete metabolomics dataset. The obtained results indicate that our algorithms are effective in discovering a panel of biomarkers of AD and MCI from metabolomics data suggesting the possibility to develop a noninvasive blood diagnostic technique of AD and MCI.

Recent advances from metabolomics and lipidomics application in alzheimer's disease inspiring drug discovery

Introduction: Although age is a major risk factor for Alzheimer's disease (AD), it is not an inevitable consequence of aging nor is it exclusively an old-age disease. Several other major risk factors for AD are strongly associated with metabolism and include lack of exercise, obesity, diabetes, high blood pressure and cholesterol, over-consumption of alcohol and depression in addition to low educational level, social isolation, and cognitive inactivity. Approaches for Alzheimer prevention and treatment through manipulation of metabolism and utilization of active metabolites have great potential either as a primary or secondary treatment avenue or as a preventative strategy in high-risk individuals.Areas covered: This review outlines the current knowledge concerning the relationship between AD and metabolism and the novel treatments attempting to correct changes in AD patients determined through metabolomics or lipidomic analyses.Expert opinion: Metabolites are one of the main driving factors and indicators of AD and can offer many possible avenues for prevention and treatment. However, with the highly interconnected effects of metabolites and metabolism, as well as the many different routes for metabolism dysfunction, successful treatment would have to include the correction of metabolic errors as well as errors in transport and metabolite processing in order to affect and revert AD progression.

Effects of Atmospheric CO2 Level on the Metabolic Response of Resistant and Susceptible Wheat to Fusarium graminearum Infection

Rising atmospheric CO₂ concentrations and associated climate changes are thought to have contributed to the steady increase of Fusarium head blight (FHB) on wheat. However, our understanding of precisely how elevated CO₂ influences the defense response of wheat against Fusarium graminearum remains limited. In this study, we evaluated the metabolic profiles of susceptible (Norm) and moderately resistant (Alsen) spring wheat in response to whole-head inoculation with two deoxynivalenol (DON)-producing F. graminearum isolates (DON⁺), isolates 9F1 and Gz3639, and a DON-deficient (DON^-) isolate (Gzt40) at ambient (400 ppm) and elevated (800 ppm) CO₂ concentrations. The effects of elevated CO₂ were dependent on both the Fusarium strain and the wheat variety, but metabolic differences in the host can explain the observed changes in F. graminearum biomass and DON accumulation. The complexity of abiotic and biotic stress interactions makes it difficult to determine if the observed metabolic changes in wheat are a result of CO₂-induced changes in the host, the pathogen, or a combination of both. However, the effects of elevated CO₂ were not dependent on DON production. Finally, we identified several metabolic biomarkers for wheat that can reliably predict FHB resistance or susceptibility, even as atmospheric CO₂ levels rise.

Machine Learning Methods for Analysis of Metabolic Data and Metabolic Pathway Modeling

Machine learning uses experimental data to optimize clustering or classification of samples or features, or to develop, augment or verify models that can be used to predict behavior or properties of systems. It is expected that machine learning will help provide actionable knowledge from a variety of big data including metabolomics data, as well as results of metabolism models. A variety of machine learning methods has been applied in bioinformatics and metabolism analyses including self-organizing maps, support vector machines, the kernel machine, Bayesian networks or fuzzy logic. To a lesser extent, machine learning has also been utilized to take advantage of the increasing availability of genomics and metabolomics data for the optimization of metabolic network models and their analysis. In this context, machine learning has aided the development of metabolic networks, the calculation of parameters for stoichiometric and kinetic models, as well as the analysis of major features in the model for the optimal application of bioreactors. Examples of this very interesting, albeit highly complex, application of machine learning for metabolism modeling will be the primary focus of this review presenting several different types of applications for model optimization, parameter determination or system analysis using models, as well as the utilization of several different types of machine learning technologies.

Database of resistance related metabolites in Wheat Fusarium head blight Disease (MWFD)

Fungal diseases are an increasing threat to worldwide food security. Fusarium head blight (FHB), primarily caused by Fusarium graminearum, is a devastating disease of Triticum aestivum (bread wheat). Partial resistance to FHB of several wheat and barley cultivars includes specific metabolic responses to inoculation. Investigation of metabolic changes in plants, following pathogen infection, provides valuable data for understanding of the role of metabolites and metabolism in plant-pathogen interaction and resistance. Determination of functions of metabolites in resistance can also inspire the development of antifungals. Metabolic changes induced by FHB in resistant and susceptible plants have been previously investigated. However, the functionality of the majority of these investigated metabolites remains unknown. The ‘Metabolites in the Wheat Fusarium head blight Disease’ (MWFD) database was compiled in order to determine possible targets and roles of these molecules in resistance to FBH and aid in the development of related synthetic antifungals. The MWFD database allows for the quick retrieval of known resistance related metabolites, associated target proteins and their sequence analogues in wheat and Fusarium genomes. The database can be searched for compounds, MeSH terms, as well as protein targets. This comprehensive, manually curated, collection of resistance related metabolites is available at https://bioinfo.nrc.ca/mwfd/index.php.

Database URL:https://bioinfo.nrc.ca/mwfd/index.php

Metabolic Biomarker Panels of Response to Fusarium Head Blight Infection in Different Wheat Varieties

Metabolic changes in spikelets of wheat varieties FL62R1, Stettler, Muchmore and Sumai3 following Fusarium graminearum infection were explored using NMR analysis. Extensive 1D and 2D 1H NMR measurements provided information for detailed metabolite assignment and quantification leading to possible metabolic markers discriminating resistance level in wheat subtypes. In addition, metabolic changes that are observed in all studied varieties as well as wheat variety specific changes have been determined and discussed. A new method for metabolite quantification from NMR data that automatically aligns spectra of standards and samples prior to quantification using multivariate linear regression optimization of spectra of assigned metabolites to samples' 1D spectra is described and utilized. Fusarium infection-induced metabolic changes in different wheat varieties are discussed in the context of metabolic network and resistance.

(1)H NMR metabolomics analysis of renal cell carcinoma cells: Effect of VHL inactivation on metabolism

Von Hippel-Lindau (VHL) is an onco-suppressor involved in oxygen and energy-dependent promotion of protein ubiquitination and proteosomal degradation. Loss of function mutations of VHL (VHL-cells) result in organ specific cancers with the best studied example in renal cell carcinomas. VHL has a well-established role in deactivation of hypoxia-inducible factor (HIF-1) and in regulation of PI3K/AKT/mTOR activity. Cell culture metabolomics analysis was utilized to determined effect of VHL and HIF-1α or HIF-2α on metabolism of renal cell carcinomas (RCC). RCC cells were stably transfected with VHL or shRNA designed to silence HIF-1α or HIF-2α genes. Obtained metabolic data was analysed qualitatively, searching for overall effects on metabolism as well as quantitatively, using methods developed in our group in order to determine specific metabolic changes. Analysis of the effect of VHL and HIF silencing on cellular metabolic footprints and fingerprints provided information about the metabolic pathways affected by VHL through HIF function as well as independently of HIF. Through correlation network analysis as well as statistical analysis of significant metabolic changes we have determined effects of VHL and HIF on energy production, amino acid metabolism, choline metabolism as well as cell regulation and signaling. VHL was shown to influence cellular metabolism through its effect on HIF proteins as well as by affecting activity of other factors.

Leveraging metabolomics to assess the next generation of temozolomide-based therapeutic approaches for glioblastomas

Glioblastoma multiforme (GBM) is the most common adult primary tumor of the central nervous system. The current standard of care for glioblastoma patients involves a combination of surgery, radiotherapy and chemotherapy with the alkylating agent temozolomide. Several mechanisms underlying the inherent and acquired temozolomide resistance have been identified and contribute to treatment failure. Early identification of temozolomide-resistant GBM patients and improvement of the therapeutic strategies available to treat this malignancy are of uttermost importance. This review initially looks at the molecular pathways underlying GBM formation and development with a particular emphasis placed on recent therapeutic advances made in the field. Our focus will next be directed toward the molecular mechanisms modulating temozolomide resistance in GBM patients and the strategies envisioned to circumvent this resistance. Finally, we highlight the diagnostic and prognostic value of metabolomics in cancers and assess its potential usefulness in improving the current standard of care for GBM patients.

Targeting the latest hallmark of cancer: another attempt at 'magic bullet' drugs targeting cancers' metabolic phenotype

The metabolism of tumors is remarkably different from the metabolism of corresponding normal cells and tissues. Metabolic alterations are initiated by oncogenes and are required for malignant transformation, allowing cancer cells to resist some cell death signals while producing energy and fulfilling their biosynthetic needs with limiting resources. The distinct metabolic phenotype of cancers provides an interesting avenue for treatment, potentially with minimal side effects. As many cancers show similar metabolic characteristics, drugs targeting the cancer metabolic phenotype are, perhaps optimistically, expected to be 'magic bullet' treatments. Over the last few years there have been a number of potential drugs developed to specifically target cancer metabolism. Several of these drugs are currently in clinical and preclinical trials. This review outlines examples of drugs developed for different targets of significance to cancer metabolism, with a focus on small molecule leads, chemical biology and clinical results for these drugs.

NMR metabolomics analysis of the effects of 5-lipoxygenase inhibitors on metabolism in glioblastomas

Changes across metabolic networks are emerging as an integral part of cancer development and progression. Increasing comprehension of the importance of metabolic processes as well as metabolites in cancer is stimulating exploration of novel, targeted treatment options. Arachidonic acid (AA) is a major component of phospholipids. Through the cascade catalyzed by cyclooxygenases and lipoxygenases, AA is also a precursor to cellular signaling molecules as well as molecules associated with a variety of diseases including cancer. 5-Lipoxygenase catalyzes the transformation of AA into leukotrienes (LT), important mediators of inflammation. High-throughput analysis of metabolic profiles was used to investigate the response of glioblastoma cell lines to treatment with 5-lipoxygenase inhibitors. Metabolic profiling of cells following drug treatment provides valuable information about the response and metabolic alterations induced by the drug action and give an indication of both on-target and off-target effects of drugs. Four different 5-lipoxygenase inhibitors and antioxidants were tested including zileuton, caffeic acid, and its analogues caffeic acid phenethyl ester and caffeic acid cyclohexethyl ester. A NMR approach identified metabolic signatures resulting from application of these compounds to glioblastoma cell lines, and metabolic data were used to develop a better understanding of the mode of action of these inhibitors.

Integrated analysis of transcriptomics and metabolomics profiles

BACKGROUND

Integrated analysis of transcriptomics and metabolomics data has the potential greatly to increase our understanding of metabolic networks and biological systems leading to various potential clinical applications.

OBJECTIVE

The aim is to present different applications as well as analysis tools utilized for the parallel study of gene and metabolite expressions.

METHODS

Publications dealing with integrated analysis of gene and metabolite expression data as well as publications describing tools that can be used for integrated analysis are reviewed.

RESULTS/CONCLUSION

The full benefit of integrated analysis can be achieved only if data from all utilized methods are treated equally by multidisciplinary teams. This approach can lead to advances in functional genomics with possible clinical developments in diagnostics and improved drug target selection.

1H NMR metabolomics analysis of glioblastoma subtypes: correlation between metabolomics and gene expression characteristics

Glioblastoma multiforme (GBM) is the most common form of malignant glioma, characterized by unpredictable clinical behaviors that suggest distinct molecular subtypes. With the tumor metabolic phenotype being one of the hallmarks of cancer, we have set upon to investigate whether GBMs show differences in their metabolic profiles. (1)H NMR analysis was performed on metabolite extracts from a selection of nine glioblastoma cell lines. Analysis was performed directly on spectral data and on relative concentrations of metabolites obtained from spectra using a multivariate regression method developed in this work. Both qualitative and quantitative sample clustering have shown that cell lines can be divided into four groups for which the most significantly different metabolites have been determined. Analysis shows that some of the major cancer metabolic markers (such as choline, lactate, and glutamine) have significantly dissimilar concentrations in different GBM groups. The obtained lists of metabolic markers for subgroups were correlated with gene expression data for the same cell lines. Metabolic analysis generally agrees with gene expression measurements, and in several cases, we have shown in detail how the metabolic results can be correlated with the analysis of gene expression. Combined gene expression and metabolomics analysis have shown differential expression of transporters of metabolic markers in these cells as well as some of the major metabolic pathways leading to accumulation of metabolites. Obtained lists of marker metabolites can be leveraged for subtype determination in glioblastomas.

Multi-gene biomarker panel for reference free prostate cancer diagnosis: determination and independent validation

Identification of biomarkers that can accurately and reliably diagnose prostate cancer is clinically highly desirable. A novel classification method, K-closest resemblance was applied to several high-quality transcriptomic datasets of prostate cancer leading to the discovery of a panel of eight gene biomarkers that can detect prostate cancer with over 96% specificity and sensitivity in leave-one-out cross-validation. Independent validation on clinical samples confirmed the discriminatory power of this gene panel, yielding over 95% accuracy of diagnosis based on receiver-operating characteristic curve analyses. Different levels of validation of the proposed biomarker panel have shown that it allows extremely accurate diagnosis of prostate cancer. Application of this panel can possibly add a fast and objective tool to the pathologist's arsenal following further clinical testing.

Clustering methods for microarray gene expression data

Within the field of genomics, microarray technologies have become a powerful technique for simultaneously monitoring the expression patterns of thousands of genes under different sets of conditions. A main task now is to propose analytical methods to identify groups of genes that manifest similar expression patterns and are activated by similar conditions. The corresponding analysis problem is to cluster multi-condition gene expression data. The purpose of this paper is to present a general view of clustering techniques used in microarray gene expression data analysis.

Data analysis of alternative splicing microarrays

The importance of alternative splicing in drug and biomarker discovery is best understood through several example genes. For most genes, the identification, detection and particularly quantification of isoforms in different tissues and conditions remain to be carried out. As a result, the focus in drug and biomarker development is increasingly on high-throughput studies of alternative splicing. Initial strategies for the parallel analysis of alternative splicing by microarrays have been recently published. The design specificities and goals of alternative splicing microarrays, in terms of identification and quantification of multiple mRNAs from one gene, are promoting the development of novel methods of analysis.

Microarray Analysis of Alternative Splicing

Alternative splicing, defined as the generation of multiple RNA transcript species from a common mRNA precursor, is one of the mechanisms for the diversification and expansion of cellular proteins from a smaller set of genes. Current estimates indicate that at least 60% of genes in the human genome exhibit alternative splicing. Over the past decade, alternative splicing has increasingly been recognized as a major regulatory process with a critical role in normal development. Furthermore, the importance of alternative splicing in disease development and treatment is starting to be appreciated. Therefore, an increasing number of high-throughput genomics and proteomics studies are being performed in order to delineate (a) the changes in alternative splicing under various conditions; (b) the properties and functions of protein isoforms; and (c) the splicing and alternative splicing regulation process. Strategies for the parallel analysis of alternative splice forms by microarray experiments have been conceived, and examples have been published. In addition to the differences in microarray probe design, the analysis of microarrays with probes for exons, exon/exon junctions as well as specific splice forms is significantly different from the standard experiment. Several methods are being developed in order to address the particular needs of alternative splicing microarrays. Many reviews have already dealt with alternative splicing. However, high-throughput analysis methods that are becoming increasingly popular have not received much attention. Here, we will provide an overview of the tools and analysis methods that were developed specifically for alternative splicing microarrays described in terms of specific experiments.

Carbohydrate Microarrays: Survey of Fabrication Techniques

Carbohydrate microarrays are being developed in order to decipher the information content of the glycome. This postgenomic activity is necessary because of the complexity of protein biosynthesis and post-translational modifications that cannot currently be detected at the genome level. This review looks, in detail, at the experimental approaches that have been taken in the fabrication and preparation of carbohydrate microarrays, glycan arrays and glyco-chips. Tether structures, glycan solution preparation, detection methods and applications have been gathered together in a tabular format.

Determination of tumour marker genes from gene expression data

Cancer classification has traditionally been based on the morphological study of tumours. However, tumours with similar histological appearances can exhibit different responses to therapy, indicating differences in tumour characteristics on the molecular level. Thus, development of a novel, reliable and precise method for classification of tumours is essential for more successful diagnosis and treatment. The high-throughput gene expression data obtained using microarray technology are currently being investigated for diagnostic applications. However, these large datasets introduce a range of challenges, making data analysis a major part of every experiment for any application, including cancer classification and diagnosis. One of the major concerns in the application of microarrays to tumour diagnostics is the fact that the expression levels of many genes are not measurably affected by carcinogenic changes in the cells. Thus, a crucial step in the application of microarrays to cancer diagnostics is the selection of diagnostic marker genes from the gene expression profiles. These molecular markers give valuable additional information for tumour diagnosis, prognosis and therapy development.

Human Pax-5 C-terminal Isoforms Possess Distinct Transactivation Properties and Are Differentially Modulated in Normal and Malignant B Cells

The transcription factor Pax-5 occupies a central role in B cell differentiation and has been implicated in the development of B cell lymphoma. The transcriptional activation function of Pax-5 requires an intact N-terminal DNA-binding domain and is strongly influenced by the C-terminal transactivation domain. We report the identification and characterization of five human Pax-5 isoforms, which occur through the alternative splicing of exons that encode for the C-terminal transactivation domain. These isoforms arise from the inclusion or exclusion of exon 7, exon 8, and/or exon 9. Three of the Pax-5 isoforms generate novel protein sequences rich in proline, serine, and threonine amino acids that are the hallmarks of transactivation domains. The Pax-5 isoforms are expressed in peripheral blood mononuclear cells, cancerous and non-cancerous B cell lines, as well as in primary B cell lymphoma tissue. Electrophoretic mobility shift assays demonstrate that the isoforms possess specific DNA binding activity and recognize the PAX-5 consensus binding sites. In reporter assays using the CD19 promoter, the transactivation properties of the various isoforms were significantly influenced by the changes in the C-terminal protein sequence. Finally, we demonstrate, for the first time, that human Pax-5 isoform expression is modulated by specific signaling pathways in B lymphocytes.

Investigation of interaction between Pax-5 isoforms and thioredoxin using de novo modelling methods

Pax-5 transcription factor plays a crucial role in B-cell development, activation and differentiation. In murine B-cells four different isoforms of Pax-5 have been identified, and their role in the regulation of the activity of the wild-type protein was revealed although still not fully understood. Using theoretical methods, we investigated the properties of one region of the Pax-5e and Pax-5d isoforms (named UDE domain) and we present a possible theoretical model for the interaction of this domain with thioredoxin that have been previously postulated based on the experimental results. Domain UDE (MW 4.8 kDa) is characterised by an extremely high ratio of positively charged residues (8) in comparisons to negatively charged amino acids (3), as well as unusually large concentrations of prolines (11.6%) and cysteines (4.7%). This is indicative of its role in protein-protein interaction. The experimental 3D structure for either UDE domain or for any analogous sequence is not yet available, and therefore we resorted to various bioinformatics methods in order to predict the secondary and 3D structure from the primary sequence of UDE. Physicochemical properties of the predicted UDE structure gave more indication about possibilities for UDE-thioredoxin binding. In addition, UDE domain was shown to have both sequence and structure analogous to a segment of NAD-reducing hydrogenase HOXS a subunit which is believed to interact with thioredoxin. These studies showed that the UDE domain in Pax-5d and Pax-5e represents an ideal binding site for thioredoxin and we developed a model of UDE-TRX complex with two disulphide bridges. The active site of thioredoxin remained exposed after binding to UDE in this model and therefore binding of thioredoxin to Pax-5d could explain the unexpectedly high resistance of this isoform to oxidation. The complex between thioredoxin and Pax-5e can be a method for transportation of thioredoxin into the nucleus and also into the the vicinity of Pax-5a, explaining the observed activator role of Pax-5e.