Multi-omics subgroups associated with glycaemic deterioration in type 2 diabetes: an IMI-RHAPSODY Study.

Introduction: Type 2 diabetes (T2D) onset, progression and outcomes differ substantially between individuals. Multi-omics analyses may allow a deeper understanding of these differences and ultimately facilitate personalised treatments. Here, in an unsupervised "bottom-up" approach, we attempt to group T2D patients based solely on -omics data generated from plasma. Methods: Circulating plasma lipidomic and proteomic data from two independent clinical cohorts, Hoorn Diabetes Care System (DCS) and Genetics of Diabetes Audit and Research in Tayside Scotland (GoDARTS), were analysed using Similarity Network Fusion. The resulting patient network was analysed with Logistic and Cox regression modelling to explore relationships between plasma -omic profiles and clinical characteristics. Results: From a total of 1,134 subjects in the two cohorts, levels of 180 circulating plasma lipids and 1195 proteins were used to separate patients into two subgroups. These differed in terms of glycaemic deterioration (Hazard Ratio=0.56;0.73), insulin sensitivity and secretion (C-peptide, p=3.7e-11;2.5e-06, DCS and GoDARTS, respectively; Homeostatic model assessment 2 (HOMA2)-B; -IR; -S, p=0.0008;4.2e-11;1.1e-09, only in DCS). The main molecular signatures separating the two groups included triacylglycerols, sphingomyelin, testican-1 and interleukin 18 receptor. Conclusions: Using an unsupervised network-based fusion method on plasma lipidomics and proteomics data from two independent cohorts, we were able to identify two subgroups of T2D patients differing in terms of disease severity. The molecular signatures identified within these subgroups provide insights into disease mechanisms and possibly new prognostic markers for T2D.

A Federated Database for Obesity Research: An IMI-SOPHIA Study.

Obesity is considered by many as a lifestyle choice rather than a chronic progressive disease. The Innovative Medicines Initiative (IMI) SOPHIA (Stratification of Obesity Phenotypes to Optimize Future Obesity Therapy) project is part of a momentum shift aiming to provide better tools for the stratification of people with obesity according to disease risk and treatment response. One of the challenges to achieving these goals is that many clinical cohorts are siloed, limiting the potential of combined data for biomarker discovery. In SOPHIA, we have addressed this challenge by setting up a federated database building on open-source DataSHIELD technology. The database currently federates 16 cohorts that are accessible via a central gateway. The database is multi-modal, including research studies, clinical trials, and routine health data, and is accessed using the R statistical programming environment where statistical and machine learning analyses can be performed at a distance without any disclosure of patient-level data. We demonstrate the use of the database by providing a proof-of-concept analysis, performing a federated linear model of BMI and systolic blood pressure, pooling all data from 16 studies virtually without any analyst seeing individual patient-level data. This analysis provided similar point estimates compared to a meta-analysis of the 16 individual studies. Our approach provides a benchmark for reproducible, safe federated analyses across multiple study types provided by multiple stakeholders.

An in vitro approach to understand contribution of kidney cells to human urinary extracellular vesicles.

Extracellular vesicles (EV) are membranous particles secreted by all cells and found in body fluids. Established EV contents include a variety of RNA species, proteins, lipids and metabolites that are considered to reflect the physiological status of their parental cells. However, to date, little is known about cell-type enriched EV cargo in complex EV mixtures, especially in urine. To test whether EV secretion from distinct human kidney cells in culture differ and can recapitulate findings in normal urine, we comprehensively analysed EV components, (particularly miRNAs, long RNAs and protein) from conditionally immortalised human kidney cell lines (podocyte, glomerular endothelial, mesangial and proximal tubular cells) and compared to EV secreted in human urine. EV from cell culture media derived from immortalised kidney cells were isolated by hydrostatic filtration dialysis (HFD) and characterised by electron microscopy (EM), nanoparticle tracking analysis (NTA) and Western blotting (WB). RNA was isolated from EV and subjected to miRNA and RNA sequencing and proteins were profiled by tandem mass tag proteomics. Representative sets of EV miRNAs, RNAs and proteins were detected in each cell type and compared to human urinary EV isolates (uEV), EV cargo database, kidney biopsy bulk RNA sequencing and proteomics, and single-cell transcriptomics. This revealed that a high proportion of the in vitro EV signatures were also found in in vivo datasets. Thus, highlighting the robustness of our in vitro model and showing that this approach enables the dissection of cell type specific EV cargo in biofluids and the potential identification of cell-type specific EV biomarkers of kidney disease.

Using in vivo transcriptomics and RNA enrichment to identify genes involved in virulence of Candida glabrata.

Candida species are the most commonly isolated opportunistic fungal pathogens in humans. Candida albicans causes most of the diagnosed infections, closely followed by Candida glabrata. C. albicans is well studied, and many genes have been shown to be important for infection and colonization of the host. It is however less clear how C. glabrata infects the host. With the help of fungal RNA enrichment, we here investigated for the first time the transcriptomic profile of C. glabrata during urinary tract infection (UTI) in mice. In the UTI model, bladders and kidneys are major target organs and therefore fungal transcriptomes were addressed in these organs. Our results showed that, next to adhesins and proteases, nitrogen metabolism and regulation play a vital role during C. glabrata UTI. Genes involved in nitrogen metabolism were upregulated and among them we show that DUR1,2 (urea amidolyase) and GAP1 (amino acid permease) were important for virulence. Furthermore, we confirmed the importance of the glyoxylate cycle in the host and identified MLS1 (malate synthase) as an important gene necessary for C. glabrata virulence. In conclusion, our study shows with the support of in vivo transcriptomics how C. glabrata adapts to host conditions.

Candida albicans commensalism in the oral mucosa is favoured by limited virulence and metabolic adaptation.

As part of the human microbiota, the fungus Candida albicans colonizes the oral cavity and other mucosal surfaces of the human body. Commensalism is tightly controlled by complex interactions of the fungus and the host to preclude fungal elimination but also fungal overgrowth and invasion, which can result in disease. As such, defects in antifungal T cell immunity render individuals susceptible to oral thrush due to interrupted immunosurveillance of the oral mucosa. The factors that promote commensalism and ensure persistence of C. albicans in a fully immunocompetent host remain less clear. Using an experimental model of C. albicans oral colonization in mice we explored fungal determinants of commensalism in the oral cavity. Transcript profiling of the oral isolate 101 in the murine tongue tissue revealed a characteristic metabolic profile tailored to the nutrient poor conditions in the stratum corneum of the epithelium where the fungus resides. Metabolic adaptation of isolate 101 was also reflected in enhanced nutrient acquisition when grown on oral mucosa substrates. Persistent colonization of the oral mucosa by C. albicans also correlated inversely with the capacity of the fungus to induce epithelial cell damage and to elicit an inflammatory response. Here we show that these immune evasive properties of isolate 101 are explained by a strong attenuation of a number of virulence genes, including those linked to filamentation. De-repression of the hyphal program by deletion or conditional repression of NRG1 abolished the commensal behaviour of isolate 101, thereby establishing a central role of this factor in the commensal lifestyle of C. albicans in the oral niche of the host.

Function Analysis of MBF1, a Factor Involved in the Response to Amino Acid Starvation and Virulence in Candida albicans.

Candida albicans is a commensal of human mucosae, but also one of the most common fungal pathogens of humans. Systemic infections caused by this fungus, mostly affecting immunocompromised patients, are associated to fatality rates as high as 50% despite the available treatments. In order to improve this situation, it is necessary to fully understand how C. albicans is able to cause disease and how it copes with the host defenses. Our previous studies have revealed the importance of the C. albicans gene MBF1 in virulence and ability to colonize internal organs of mammalian and insect hosts. MBF1 encodes a putative transcriptional regulator, and as such it likely has an impact in the regulation of C. albicans gene expression during host infection. Here, recent advances in RNA-seq technologies were used to obtain a detailed analysis of the impact of MBF1 on C. albicans gene expression both in vitro and during infection. MBF1 was involved in the regulation of several genes with a role in glycolysis and response to stress, particularly to nutritional stress. We also investigated whether an interaction existed between MBF1 and GCN4, a master regulator of response to starvation, and found that both genes were needed for resistance to amino acid starvation, suggesting some level of interaction between the two. Reinforcing this idea, we showed that the proteins encoded by both genes could interact. Consistent with the role of MBF1 in virulence, we also established that GCN4 was necessary for virulence in the mouse model of systemic infection as well as in the Galleria mellonella infection model.

Anti-adipogenic signals at the onset of obesity-related inflammation in white adipose tissue.

Chronic inflammation that affects primarily metabolic organs, such as white adipose tissue (WAT), is considered as a major cause of human obesity-associated co-morbidities. However, the molecular mechanisms initiating this inflammation in WAT are poorly understood. By combining transcriptomics, ChIP-seq and modeling approaches, we studied the global early and late responses to a high-fat diet (HFD) in visceral (vWAT) and subcutaneous (scWAT) AT, the first being more prone to obesity-induced inflammation. HFD rapidly triggers proliferation of adipocyte precursors within vWAT. However, concomitant antiadipogenic signals limit vWAT hyperplastic expansion by interfering with the differentiation of proliferating adipocyte precursors. Conversely, in scWAT, residing beige adipocytes lose their oxidizing properties and allow storage of excessive fatty acids. This phase is followed by tissue hyperplastic growth and increased angiogenic signals, which further enable scWAT expansion without generating inflammation. Our data indicate that scWAT and vWAT differential ability to modulate adipocyte number and differentiation in response to obesogenic stimuli has a crucial impact on the different susceptibility to obesity-related inflammation of these adipose tissue depots.

MetaNetX/MNXref: unified namespace for metabolites and biochemical reactions in the context of metabolic models.

MetaNetX/MNXref is a reconciliation of metabolites and biochemical reactions providing cross-links between major public biochemistry and Genome-Scale Metabolic Network (GSMN) databases. The new release brings several improvements with respect to the quality of the reconciliation, with particular attention dedicated to preserving the intrinsic properties of GSMN models. The MetaNetX website (https://www.metanetx.org/) provides access to the full database and online services. A major improvement is for mapping of user-provided GSMNs to MXNref, which now provides diagnostic messages about model content. In addition to the website and flat files, the resource can now be accessed through a SPARQL endpoint (https://rdf.metanetx.org).

Persistence of Candida albicans in the Oral Mucosa Induces a Curbed Inflammatory Host Response That Is Independent of Immunosuppression.

Controlled immune activation in response to commensal microbes is critical for the maintenance of stable colonization and prevention of microbial overgrowth on epithelial surfaces. Our understanding of the host mechanisms that regulate bacterial commensalism has increased substantially, however, much less data exist regarding host responses to members of the fungal microbiota on colonized surfaces. Using a murine model of oropharyngeal candidiasis, we have recently shown that differences in immune activation in response to diverse natural isolates of Candida albicans are associated with different outcomes of the host-fungal interaction. Here we applied a genome-wide transcriptomic approach to show that rapid induction of a strong inflammatory response characterized by neutrophil-associated genes upon C. albicans colonization inversely correlated with the ability of the fungus to persist in the oral mucosa. Surprisingly, persistent fungal isolates showed no signs of a compensatory regulatory immune response. By combining RNA-seq data, genetic mouse models, and co-infection experiments, we show that attenuation of the inflammatory response at the onset of infection with a persistent isolate is not a consequence of enhanced immunosuppression. Importantly, depletion of regulatory T cells or deletion of the immunoregulatory cytokine IL-10 did not alter host-protective type 17 immunity nor did it impair fungal survival in the oral mucosa, indicating that persistence of C. albicans in the oral mucosa is not a consequence of suppressed antifungal immunity.

Condition-specific series of metabolic sub-networks and its application for gene set enrichment analysis.

MOTIVATION: Genome-scale metabolic networks and transcriptomic data represent complementary sources of knowledge about an organism's metabolism, yet their integration to achieve biological insight remains challenging. RESULTS: We investigate here condition-specific series of metabolic sub-networks constructed by successively removing genes from a comprehensive network. The optimal order of gene removal is deduced from transcriptomic data. The sub-networks are evaluated via a fitness function, which estimates their degree of alteration. We then consider how a gene set, i.e. a group of genes contributing to a common biological function, is depleted in different series of sub-networks to detect the difference between experimental conditions. The method, named metaboGSE, is validated on public data for Yarrowia lipolytica and mouse. It is shown to produce GO terms of higher specificity compared to popular gene set enrichment methods like GSEA or topGO. AVAILABILITY AND IMPLEMENTATION: The metaboGSE R package is available at https://CRAN.R-project.org/package=metaboGSE. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Comparative Genomics of Two Sequential Candida glabrata Clinical Isolates.

Candida glabrata is an important fungal pathogen which develops rapid antifungal resistance in treated patients. It is known that azole treatments lead to antifungal resistance in this fungal species and that multidrug efflux transporters are involved in this process. Specific mutations in the transcriptional regulator PDR1 result in upregulation of the transporters. In addition, we showed that the PDR1 mutations can contribute to enhance virulence in animal models. In this study, we were interested to compare genomes of two specific C. glabrata-related isolates, one of which was azole susceptible (DSY562) while the other was azole resistant (DSY565). DSY565 contained a PDR1 mutation (L280F) and was isolated after a time-lapse of 50 d of azole therapy. We expected that genome comparisons between both isolates could reveal additional mutations reflecting host adaptation or even additional resistance mechanisms. The PacBio technology used here yielded 14 major contigs (sizes 0.18-1.6 Mb) and mitochondrial genomes from both DSY562 and DSY565 isolates that were highly similar to each other. Comparisons of the clinical genomes with the published CBS138 genome indicated important genome rearrangements, but not between the clinical strains. Among the unique features, several retrotransposons were identified in the genomes of the investigated clinical isolates. DSY562 and DSY565 each contained a large set of adhesin-like genes (101 and 107, respectively), which exceed by far the number of reported adhesins (63) in the CBS138 genome. Comparison between DSY562 and DSY565 yielded 17 nonsynonymous SNPs (among which the was the expected PDR1 mutation) as well as small size indels in coding regions (11) but mainly in adhesin-like genes. The genomes contained a DNA mismatch repair allele of MSH2 known to be involved in the so-called hyper-mutator phenotype of this yeast species and the number of accumulated mutations between both clinical isolates is consistent with the presence of a MSH2 defect. In conclusion, this study is the first to compare genomes of C. glabrata sequential clinical isolates using the PacBio technology as an approach. The genomes of these isolates taken in the same patient at two different time points exhibited limited variations, even if submitted to the host pressure.

RNA Sequencing-Based Genome Reannotation of the Dermatophyte Arthroderma benhamiae and Characterization of Its Secretome and Whole Gene Expression Profile during Infection.

Dermatophytes are the most common agents of superficial mycoses in humans and animals. The aim of the present investigation was to systematically identify the extracellular, possibly secreted, proteins that are putative virulence factors and antigenic molecules of dermatophytes. A complete gene expression profile of Arthroderma benhamiae was obtained during infection of its natural host (guinea pig) using RNA sequencing (RNA-seq) technology. This profile was completed with those of the fungus cultivated in vitro in two media containing either keratin or soy meal protein as the sole source of nitrogen and in Sabouraud medium. More than 60% of transcripts deduced from RNA-seq data differ from those previously deposited for A. benhamiae. Using these RNA-seq data along with an automatic gene annotation procedure, followed by manual curation, we produced a new annotation of the A. benhamiae genome. This annotation comprised 7,405 coding sequences (CDSs), among which only 2,662 were identical to the currently available annotation, 383 were newly identified, and 15 secreted proteins were manually corrected. The expression profile of genes encoding proteins with a signal peptide in infected guinea pigs was found to be very different from that during in vitro growth when using keratin as the substrate. Especially, the sets of the 12 most highly expressed genes encoding proteases with a signal sequence had only the putative vacuolar aspartic protease gene PEP2 in common, during infection and in keratin medium. The most upregulated gene encoding a secreted protease during infection was that encoding subtilisin SUB6, which is a known major allergen in the related dermatophyte Trichophyton rubrum. IMPORTANCE Dermatophytoses (ringworm, jock itch, athlete's foot, and nail infections) are the most common fungal infections, but their virulence mechanisms are poorly understood. Combining transcriptomic data obtained from growth under various culture conditions with data obtained during infection led to a significantly improved genome annotation. About 65% of the protein-encoding genes predicted with our protocol did not match the existing annotation for A. benhamiae. Comparing gene expression during infection on guinea pigs with keratin degradation in vitro, which is supposed to mimic the host environment, revealed the critical importance of using real in vivo conditions for investigating virulence mechanisms. The analysis of genes expressed in vivo, encoding cell surface and secreted proteins, particularly proteases, led to the identification of new allergen and virulence factor candidates.

Detection of generic differential RNA processing events from RNA-seq data.

RNA-seq data analysis has revealed abundant alternative splicing in eukaryotic mRNAs. However, splicing is only one of many processing events that transcripts may undergo during their lifetime. We present here RNAprof (RNA profile analysis), a program for the detection of differential processing events from the comparison of RNA-seq experiments. RNAprof implements a specific gene-level normalization procedure and compares RNA-seq coverage profiles at nucleotide resolution to detect regions of significant coverage differences, independently of splice sites or other gene features. We used RNAprof to analyze the effect of alternative-splicing regulators NSRa and NSRb on the Arabidopsis thaliana transcriptome. A number of intron retention events and alternative transcript structures were specifically detected by RNAprof and confirmed by qRT-PCR. Further tests using a public Mus musculus RNA-seq dataset and comparisons with other RNA isoform predictors showed that RNAprof uniquely identified sets of highly significant processing events as well as other relevant library-specific differences in RNA-seq profiles. This highlights an important layer of variation that remains undetected by current protocols for RNA-seq analysis.

RNA Enrichment Method for Quantitative Transcriptional Analysis of Pathogens In Vivo Applied to the Fungus Candida albicans.

UNLABELLED: In vivo transcriptional analyses of microbial pathogens are often hampered by low proportions of pathogen biomass in host organs, hindering the coverage of full pathogen transcriptome. We aimed to address the transcriptome profiles of Candida albicans, the most prevalent fungal pathogen in systemically infected immunocompromised patients, during systemic infection in different hosts. We developed a strategy for high-resolution quantitative analysis of the C. albicans transcriptome directly from early and late stages of systemic infection in two different host models, mouse and the insect Galleria mellonella. Our results show that transcriptome sequencing (RNA-seq) libraries were enriched for fungal transcripts up to 1,600-fold using biotinylated bait probes to capture C. albicans sequences. This enrichment biased the read counts of only ~3% of the genes, which can be identified and removed based on a priori criteria. This allowed an unprecedented resolution of C. albicans transcriptome in vivo, with detection of over 86% of its genes. The transcriptional response of the fungus was surprisingly similar during infection of the two hosts and at the two time points, although some host- and time point-specific genes could be identified. Genes that were highly induced during infection were involved, for instance, in stress response, adhesion, iron acquisition, and biofilm formation. Of the in vivo-regulated genes, 10% are still of unknown function, and their future study will be of great interest. The fungal RNA enrichment procedure used here will help a better characterization of the C. albicans response in infected hosts and may be applied to other microbial pathogens. IMPORTANCE: Understanding the mechanisms utilized by pathogens to infect and cause disease in their hosts is crucial for rational drug development. Transcriptomic studies may help investigations of these mechanisms by determining which genes are expressed specifically during infection. This task has been difficult so far, since the proportion of microbial biomass in infected tissues is often extremely low, thus limiting the depth of sequencing and comprehensive transcriptome analysis. Here, we adapted a technology to capture and enrich C. albicans RNA, which was next used for deep RNA sequencing directly from infected tissues from two different host organisms. The high-resolution transcriptome revealed a large number of genes that were so far unknown to participate in infection, which will likely constitute a focus of study in the future. More importantly, this method may be adapted to perform transcript profiling of any other microbes during host infection or colonization.

miRBoost: boosting support vector machines for microRNA precursor classification.

Identification of microRNAs (miRNAs) is an important step toward understanding post-transcriptional gene regulation and miRNA-related pathology. Difficulties in identifying miRNAs through experimental techniques combined with the huge amount of data from new sequencing technologies have made in silico discrimination of bona fide miRNA precursors from non-miRNA hairpin-like structures an important topic in bioinformatics. Among various techniques developed for this classification problem, machine learning approaches have proved to be the most promising. However these approaches require the use of training data, which is problematic due to an imbalance in the number of miRNAs (positive data) and non-miRNAs (negative data), which leads to a degradation of their performance. In order to address this issue, we present an ensemble method that uses a boosting technique with support vector machine components to deal with imbalanced training data. Classification is performed following a feature selection on 187 novel and existing features. The algorithm, miRBoost, performed better in comparison with state-of-the-art methods on imbalanced human and cross-species data. It also showed the highest ability among the tested methods for discovering novel miRNA precursors. In addition, miRBoost was over 1400 times faster than the second most accurate tool tested and was significantly faster than most of the other tools. miRBoost thus provides a good compromise between prediction efficiency and execution time, making it highly suitable for use in genome-wide miRNA precursor prediction. The software miRBoost is available on our web server http://EvryRNA.ibisc.univ-evry.fr.

Supersecondary structure prediction of transmembrane beta-barrel proteins.

We introduce a graph-theoretic model for predicting the supersecondary structure of transmembrane ß-barrel proteins--a particular class of proteins that performs diverse important functions but it is difficult to determine their structure with experimental methods. This ab initio model resolves the protein folding problem based on pseudo-energy minimization with the aid of a simple probabilistic filter. It also allows for determining structures whose barrel follows a given permutation on the arrangement of ß-strands, and allows for rapidly discriminating the transmembrane ß-barrels from other kinds of proteins. The model is fairly accurate, robust and can be run very efficiently on PC-like computers, thus proving useful for genome screening.

A graph-theoretic approach for classification and structure prediction of transmembrane ß-barrel proteins.

BACKGROUND: Transmembrane ß-barrel proteins are a special class of transmembrane proteins which play several key roles in human body and diseases. Due to experimental difficulties, the number of transmembrane ß-barrel proteins with known structures is very small. Over the years, a number of learning-based methods have been introduced for recognition and structure prediction of transmembrane ß-barrel proteins. Most of these methods emphasize on homology search rather than any biological or chemical basis. RESULTS: We present a novel graph-theoretic model for classification and structure prediction of transmembrane ß-barrel proteins. This model folds proteins based on energy minimization rather than a homology search, avoiding any assumption on availability of training dataset. The ab initio model presented in this paper is the first method to allow for permutations in the structure of transmembrane proteins and provides more structural information than any known algorithm. The model is also able to recognize ß-barrels by assessing the pseudo free energy. We assess the structure prediction on 41 proteins gathered from existing databases on experimentally validated transmembrane ß-barrel proteins. We show that our approach is quite accurate with over 90% F-score on strands and over 74% F-score on residues. The results are comparable to other algorithms suggesting that our pseudo-energy model is close to the actual physical model. We test our classification approach and show that it is able to reject α-helical bundles with 100% accuracy and ß-barrel lipocalins with 97% accuracy. CONCLUSIONS: We show that it is possible to design models for classification and structure prediction for transmembrane ß-barrel proteins which do not depend essentially on training sets but on combinatorial properties of the structures to be proved. These models are fairly accurate, robust and can be run very efficiently on PC-like computers. Such models are useful for the genome screening.