Improvement of variables interpretability in kernel PCA.

BACKGROUND: Kernel methods have been proven to be a powerful tool for the integration and analysis of high-throughput technologies generated data. Kernels offer a nonlinear version of any linear algorithm solely based on dot products. The kernelized version of principal component analysis is a valid nonlinear alternative to tackle the nonlinearity of biological sample spaces. This paper proposes a novel methodology to obtain a data-driven feature importance based on the kernel PCA representation of the data. RESULTS: The proposed method, kernel PCA Interpretable Gradient (KPCA-IG), provides a data-driven feature importance that is computationally fast and based solely on linear algebra calculations. It has been compared with existing methods on three benchmark datasets. The accuracy obtained using KPCA-IG selected features is equal to or greater than the other methods' average. Also, the computational complexity required demonstrates the high efficiency of the method. An exhaustive literature search has been conducted on the selected genes from a publicly available Hepatocellular carcinoma dataset to validate the retained features from a biological point of view. The results once again remark on the appropriateness of the computed ranking. CONCLUSIONS: The black-box nature of kernel PCA needs new methods to interpret the original features. Our proposed methodology KPCA-IG proved to be a valid alternative to select influential variables in high-dimensional high-throughput datasets, potentially unravelling new biological and medical biomarkers.

RNA Sequencing Reveals Widespread Transcription of Natural Antisense RNAs in Entamoeba Species.

Entamoeba is a genus of Amoebozoa that includes the intestine-colonizing pathogenic species Entamoeba histolytica. To understand the basis of gene regulation in E. histolytica from an evolutionary perspective, we have profiled the transcriptomes of its closely related species E. dispar, E. moshkovskii and E. invadens. Genome-wide identification of transcription start sites (TSS) and polyadenylation sites (PAS) revealed the similarities and differences of their gene regulatory sequences. In particular, we found the widespread initiation of antisense transcription from within the gene coding sequences is a common feature among all Entamoeba species. Interestingly, we observed the enrichment of antisense transcription in genes involved in several processes that are common to species infecting the human intestine, e.g., the metabolism of phospholipids. These results suggest a potentially conserved and compact gene regulatory system in Entamoeba.

Gene expression analysis in EBV-infected ataxia-telangiectasia cell lines by RNA-sequencing reveals protein synthesis defect and immune abnormalities.

BACKGROUND: Epstein-Barr virus (EBV) targets B-cells where it establishes a latent infection. EBV can transform B-cells in vitro and is recognized as an oncogenic virus, especially in the setting of immune compromise. Indeed, immunodeficient patients may fail to control chronic EBV infection, leading to the development EBV-driven lymphoid malignancies. Ataxia telangiectasia (AT) is a primary immune deficiency caused by mutations in the ATM gene, involved in the repair of double-strand breaks. Patients with AT are at high risk of developing cancers, mostly B-cell lymphoid malignancies, most of which being EBV-related. Aside from immune deficiency secondary to AT, loss of ATM function could also hinder the control of the virus within B-cells, favoring lymphomagenesis in AT patients. RESULTS: We used RNA sequencing on lymphoblastoid cell lines derived from patients with AT and healthy donors to analyze and compare both cellular and viral gene expression. We found numerous deregulated signaling pathways involving transcription, translation, oncogenesis and immune regulation. Specifically, the translational defect was confirmed in vitro, suggesting that the pathogenesis of AT may also involve a ribosomal defect. Concomitant analysis of viral gene expression did not reveal significant differential gene expression, however, analysis of EBV interactome suggests that the viral latency genes EBNA-3A, EBNA-3C and LMP1 may be disrupted in LCL from AT patients. CONCLUSION: Our data support the notion that ATM deficiency deregulates cellular gene expression possibly disrupting interactions with EBV latent genes, promoting the oncogenic potential of the virus. These preliminary findings provide a new step towards the understanding of EBV regulation and of AT pathogenesis.

Systemic viral spreading and defective host responses are associated with fatal Lassa fever in macaques.

Lassa virus (LASV) is endemic in West Africa and induces a viral hemorrhagic fever (VHF) with up to 30% lethality among clinical cases. The mechanisms involved in control of Lassa fever or, in contrast, the ensuing catastrophic illness and death are poorly understood. We used the cynomolgus monkey model to reproduce the human disease with asymptomatic to mild or fatal disease. After initial replication at the inoculation site, LASV reached the secondary lymphoid organs. LASV did not spread further in nonfatal disease and was rapidly controlled by balanced innate and T-cell responses. Systemic viral dissemination occurred during severe disease. Massive replication, a cytokine/chemokine storm, defective T-cell responses, and multiorgan failure were observed. Clinical, biological, immunological, and transcriptomic parameters resembled those observed during septic-shock syndrome, suggesting that similar pathogenesis is induced during Lassa fever. The outcome appears to be determined early, as differentially expressed genes in PBMCs were associated with fatal and non-fatal Lassa fever outcome very early after infection. These results provide a full characterization and important insights into Lassa fever pathogenesis and could help to develop early diagnostic tools.

Genomic determinants for initiation and length of natural antisense transcripts in Entamoeba histolytica.

Natural antisense transcripts (NAT) have been reported in prokaryotes and eukaryotes. While the functions of most reported NATs remain unknown, their potentials in regulating the transcription of their counterparts have been speculated. Entamoeba histolytica, which is a unicellular eukaryotic parasite, has a compact protein-coding genome with very short intronic and intergenic regions. The regulatory mechanisms of gene expression in this compact genome are under-described. In this study, by genome-wide mapping of RNA-Seq data in the genome of E. histolytica, we show that a substantial fraction of its protein-coding genes (28%) has significant transcription on their opposite strand (i.e. NAT). Intriguingly, we found the location of transcription start sites or polyadenylation sites of NAT are determined by the specific motifs encoded on the opposite strand of the gene coding sequences, thereby providing a compact regulatory system for gene transcription. Moreover, we demonstrated that NATs are globally up-regulated under various environmental conditions including temperature stress and pathogenicity. While NATs do not appear to be consequences of spurious transcription, they may play a role in regulating gene expression in E. histolytica, a hypothesis which needs to be tested.

Impact of the gut microbiota on the m6A epitranscriptome of mouse cecum and liver.

The intestinal microbiota modulates host physiology and gene expression via mechanisms that are not fully understood. Here we examine whether host epitranscriptomic marks are affected by the gut microbiota. We use methylated RNA-immunoprecipitation and sequencing (MeRIP-seq) to identify N6-methyladenosine (m6A) modifications in mRNA of mice carrying conventional, modified, or no microbiota. We find that variations in the gut microbiota correlate with m6A modifications in the cecum, and to a lesser extent in the liver, affecting pathways related to metabolism, inflammation and antimicrobial responses. We analyze expression levels of several known writer and eraser enzymes, and find that the methyltransferase Mettl16 is downregulated in absence of a microbiota, and one of its target mRNAs, encoding S-adenosylmethionine synthase Mat2a, is less methylated. We furthermore show that Akkermansia muciniphila and Lactobacillus plantarum affect specific m6A modifications in mono-associated mice. Our results highlight epitranscriptomic modifications as an additional level of interaction between commensal bacteria and their host.

Vaccines inducing immunity to Lassa virus glycoprotein and nucleoprotein protect macaques after a single shot.

Lassa fever is a major threat in Western Africa. The large number of people living at risk for this disease calls for the development of a vaccine against Lassa virus (LASV). We generated live-attenuated LASV vaccines based on measles virus and Mopeia virus platforms and expressing different LASV antigens, with the aim to develop a vaccine able to protect after a single shot. We compared the efficacy of these vaccines against LASV in cynomolgus monkeys. The vaccines were well tolerated and protected the animals from LASV infection and disease after a single immunization but with varying efficacy. Analysis of the immune responses showed that complete protection was associated with robust secondary T cell and antibody responses against LASV. Transcriptomic and proteomic analyses showed an early activation of innate immunity and T cell priming after immunization with the most effective vaccines, with changes detectable as early as 2 days after immunization. The most efficacious vaccine candidate, a measles vector simultaneously expressing LASV glycoprotein and nucleoprotein, has been selected for further clinical evaluation.

Non-Pathogenic Mopeia Virus Induces More Robust Activation of Plasmacytoid Dendritic Cells than Lassa Virus.

Lassa virus (LASV) causes a viral haemorrhagic fever in humans and is a major public health concern in West Africa. An efficient immune response to LASV appears to rely on type I interferon (IFN-I) production and T-cell activation. We evaluated the response of plasmacytoid dendritic cells (pDC) to LASV, as they are an important and early source of IFN-I. We compared the response of primary human pDCs to LASV and Mopeia virus (MOPV), which is very closely related to LASV, but non-pathogenic. We showed that pDCs are not productively infected by either MOPV or LASV, but produce IFN-I. However, the activation of pDCs was more robust in response to MOPV than LASV. In vivo, pDC activation may support the control of viral replication through IFN-I production, but also improve the induction of a global immune response. Therefore, pDC activation could play a role in the control of LASV infection.

Tri-mannose grafting of chitosan nanocarriers remodels the macrophage response to bacterial infection.

BACKGROUND: Infectious diseases are still a leading cause of death and, with the emergence of drug resistance, pose a great threat to human health. New drugs and strategies are thus urgently needed to improve treatment efficacy and limit drug-associated side effects. Nanotechnology-based drug delivery systems are promising approaches, offering hope in the fight against drug resistant bacteria. However, how nanocarriers influence the response of innate immune cells to bacterial infection is mostly unknown. RESULTS: Here, we used Mycobacterium tuberculosis as a model of bacterial infection to examine the impact of mannose functionalization of chitosan nanocarriers (CS-NCs) on the human macrophage response. Both ungrafted and grafted CS-NCs were similarly internalized by macrophages, via an actin cytoskeleton-dependent process. Although tri-mannose ligands did not modify the capacity of CS-NCs to escape lysosomal degradation, they profoundly remodeled the response of M. tuberculosis-infected macrophages. mRNA sequencing showed nearly 900 genes to be differentially expressed due to tri-mannose grafting. Unexpectedly, the set of modulated genes was enriched for pathways involved in cell metabolism, particularly oxidative phosphorylation and sugar metabolism. CONCLUSIONS: The ability to modulate cell metabolism by grafting ligands at the surface of nanoparticles may thus be a promising strategy to reprogram immune cells and improve the efficacy of encapsulated drugs.

Metabolic plasticity of HIV-specific CD8+ T cells is associated with enhanced antiviral potential and natural control of HIV-1 infection.

Spontaneous control of human immunodeficiency virus (HIV) is generally associated with an enhanced capacity of CD8+ T cells to eliminate infected CD4+ T cells, but the molecular characteristics of these highly functional CD8+ T cells are largely unknown. In the present study, using single-cell analysis, it was shown that HIV-specific, central memory CD8+ T cells from spontaneous HIV controllers (HICs) and antiretrovirally treated non-controllers have opposing transcriptomic profiles. Genes linked to effector functions and survival are upregulated in cells from HICs. In contrast, genes associated with activation, exhaustion and glycolysis are upregulated in cells from non-controllers. It was shown that HIV-specific CD8+ T cells from non-controllers are largely glucose dependent, whereas those from HICs have more diverse metabolic resources that enhance both their survival potential and their capacity to develop anti-HIV effector functions. The functional efficiency of the HIV-specific CD8+ T cell response in HICs is thus engraved in their memory population and related to their metabolic programme. Metabolic reprogramming in vitro through interleukin-15 treatment abrogated the glucose dependency and enhanced the antiviral potency of HIV-specific CD8+ T cells from non-controllers.

Lassa virus activates myeloid dendritic cells but suppresses their ability to stimulate T cells.

Lassa virus (LASV) is responsible for a viral hemorrhagic fever in humans and the death of 3,000 to 5,000 people every year. The immune response to LASV is poorly understood, but type I interferon (IFN-I) and T-cell responses appear to be critical for the host. We studied the response of myeloid dendritic cells (mDC) to LASV, as mDCs are involved in both IFN-I production and T-cell activation. We compared the response of primary human mDCs to LASV and Mopeia virus (MOPV), which is similar to LASV, but non-pathogenic. We showed that mDCs produced substantial amounts of IFN-I in response to both LASV and MOPV. However, only MOPV-infected mDCs were able to activate T cells. More surprisingly, coculture with T cells completely inhibited the activation of LASV-infected mDCs. These differences between LASV and MOPV were mostly due to the LASV nucleoprotein, which has major immunosuppressive properties, but the glycoprotein was also involved. Overall, these results suggest that mDCs may be important for the global response to LASV and play a role in the outcome of Lassa fever.

Expansion of the SOS regulon of Vibrio cholerae through extensive transcriptome analysis and experimental validation.

BACKGROUND: The SOS response is an almost ubiquitous response of cells to genotoxic stresses. The full complement of genes in the SOS regulon for Vibrio species has only been addressed through bioinformatic analyses predicting LexA binding box consensus and in vitro validation. Here, we perform whole transcriptome sequencing from Vibrio cholerae treated with mitomycin C as an SOS inducer to characterize the SOS regulon and other pathways affected by this treatment. RESULTS: Comprehensive transcriptional profiling allowed us to define the full landscape of promoters and transcripts active in V. cholerae. We performed extensive transcription start site (TSS) mapping as well as detection/quantification of the coding and non-coding RNA (ncRNA) repertoire in strain N16961. To improve TSS detection, we developed a new technique to treat RNA extracted from cells grown in various conditions. This allowed for identification of 3078 TSSs with an average 5'UTR of 116 nucleotides, and peak distribution between 16 and 64 nucleotides; as well as 629 ncRNAs. Mitomycin C treatment induced transcription of 737 genes and 28 ncRNAs at least 2 fold, while it repressed 231 genes and 17 ncRNAs. Data analysis revealed that in addition to the core genes known to integrate the SOS regulon, several metabolic pathways were induced. This study allowed for expansion of the Vibrio SOS regulon, as twelve genes (ubiEJB, tatABC, smpA, cep, VC0091, VC1190, VC1369-1370) were found to be co-induced with their adjacent canonical SOS regulon gene(s), through transcriptional read-through. Characterization of UV and mitomycin C susceptibility for mutants of these newly identified SOS regulon genes and other highly induced genes and ncRNAs confirmed their role in DNA damage rescue and protection. CONCLUSIONS: We show that genotoxic stress induces a pervasive transcriptional response, affecting almost 20% of the V. cholerae genes. We also demonstrate that the SOS regulon is larger than previously known, and its syntenic organization is conserved among Vibrio species. Furthermore, this specific co-localization is found in other γ-proteobacteria for genes recN-smpA and rmuC-tatABC, suggesting SOS regulon conservation in this phylum. Finally, we comment on the limitations of widespread NGS approaches for identification of all RNA species in bacteria.

Individual co-variation between viral RNA load and gene expression reveals novel host factors during early dengue virus infection of the Aedes aegypti midgut.

Dengue virus (DENV) causes more human infections than any other mosquito-borne virus. The current lack of antiviral strategies has prompted genome-wide screens for host genes that are required for DENV infectivity. Earlier transcriptomic studies that identified DENV host factors in the primary vector Aedes aegypti used inbred laboratory colonies and/or pools of mosquitoes that erase individual variation. Here, we performed transcriptome sequencing on individual midguts in a field-derived Ae. aegypti population to identify new candidate host factors modulating DENV replication. We analyzed the transcriptomic data using an approach that accounts for individual co-variation between viral RNA load and gene expression. This approach generates a prediction about the agonist or antagonist effect of candidate genes on DENV replication based on the sign of the correlation between gene expression and viral RNA load. Using this method, we identified 39 candidate genes that went undetected by conventional pairwise comparison of gene expression levels between DENV-infected midguts and uninfected controls. Only four candidate genes were detected by both methods, emphasizing their complementarity. We demonstrated the value of our approach by functional validation of a candidate agonist gene encoding a sterol regulatory element-binding protein (SREBP), which was identified by correlation analysis but not by pairwise comparison. We confirmed that SREBP promotes DENV infection in the midgut by RNAi-mediated gene knockdown in vivo. We suggest that our approach for transcriptomic analysis can empower genome-wide screens for potential agonist or antagonist factors by leveraging inter-individual variation in gene expression. More generally, this method is applicable to a wide range of phenotypic traits displaying inter-individual variation.

Comparison of biofilm formation and motility processes in arsenic-resistant Thiomonas spp. strains revealed divergent response to arsenite.

Bacteria of the genus Thiomonas are found ubiquitously in arsenic contaminated waters such as acid mine drainage (AMD), where they contribute to the precipitation and the natural bioremediation of arsenic. In these environments, these bacteria have developed a large range of resistance strategies among which the capacity to form particular biofilm structures. The biofilm formation is one of the most ubiquitous adaptive response observed in prokaryotes to various stresses, such as those induced in the presence of toxic compounds. This study focused on the process of biofilm formation in three Thiomonas strains (CB1, CB2 and CB3) isolated from the same AMD. The results obtained here show that these bacteria are all capable of forming biofilms, but the architecture and the kinetics of formation of these biofilms differ depending on whether arsenite is present in the environment and from one strain to another. Indeed, two strains favoured biofilm formation, whereas one favoured motility in the presence of arsenite. To identify the underlying mechanisms, the patterns of expression of some genes possibly involved in the process of biofilm formation were investigated in Thiomonas sp. CB2 in the presence and absence of arsenite, using a transcriptomic approach (RNA-seq). The findings obtained here shed interesting light on how the formation of biofilms, and the motility processes contribute to the adaptation of Thiomonas strains to extreme environments.

Extensive transcriptome analysis correlates the plasticity of Entamoeba histolytica pathogenesis to rapid phenotype changes depending on the environment.

Amoebiasis is a human infectious disease due to the amoeba parasite Entamoeba histolytica. The disease appears in only 20% of the infections. Diversity in phenotypes may occur within the same infectious strain in the gut; for instance, parasites can be commensal (in the intestinal lumen) or pathogenic (inside the tissue). The degree of pathogenesis of clinical isolates varies greatly. These findings raise the hypothesis that genetic derivation may account for amoebic diverse phenotypes. The main goal of this study was to analyse gene expression changes of a single virulent amoebic strain in different environmental contexts where it exhibit different degrees of virulence, namely isolated from humans and maintained through animal liver passages, in contact with the human colon and short or prolonged in vitro culture. The study reveals major transcriptome changes in virulent parasites upon contact with human colon explants, including genes related to sugar metabolism, cytoskeleton rearrangement, stress responses and DNA repair. Furthermore, in long-term cultured parasites, drastic changes in gene expression for proteins with functions for proteasome and tRNA activities were found. Globally we conclude that rapid changes in gene expression rather than genetic derivation can sustain the invasive phenotype of a single virulent isolate of E. histolytica.

MEMHDX: an interactive tool to expedite the statistical validation and visualization of large HDX-MS datasets.

MOTIVATION: With the continued improvement of requisite mass spectrometers and UHPLC systems, Hydrogen/Deuterium eXchange Mass Spectrometry (HDX-MS) workflows are rapidly evolving towards the investigation of more challenging biological systems, including large protein complexes and membrane proteins. The analysis of such extensive systems results in very large HDX-MS datasets for which specific analysis tools are required to speed up data validation and interpretation. RESULTS: We introduce a web application and a new R-package named 'MEMHDX' to help users analyze, validate and visualize large HDX-MS datasets. MEMHDX is composed of two elements. A statistical tool aids in the validation of the results by applying a mixed-effects model for each peptide, in each experimental condition, and at each time point, taking into account the time dependency of the HDX reaction and number of independent replicates. Two adjusted P-values are generated per peptide, one for the 'Change in dynamics' and one for the 'Magnitude of ΔD', and are used to classify the data by means of a 'Logit' representation. A user-friendly interface developed with Shiny by RStudio facilitates the use of the package. This interactive tool allows the user to easily and rapidly validate, visualize and compare the relative deuterium incorporation on the amino acid sequence and 3D structure, providing both spatial and temporal information. AVAILABILITY AND IMPLEMENTATION: MEMHDX is freely available as a web tool at the project home page http://memhdx.c3bi.pasteur.fr CONTACT: marie-agnes.dillies@pasteur.fr or sebastien.brier@pasteur.frSupplementary information: Supplementary data is available at Bioinformatics online.

SARTools: A DESeq2- and EdgeR-Based R Pipeline for Comprehensive Differential Analysis of RNA-Seq Data.

BACKGROUND: Several R packages exist for the detection of differentially expressed genes from RNA-Seq data. The analysis process includes three main steps, namely normalization, dispersion estimation and test for differential expression. Quality control steps along this process are recommended but not mandatory, and failing to check the characteristics of the dataset may lead to spurious results. In addition, normalization methods and statistical models are not exchangeable across the packages without adequate transformations the users are often not aware of. Thus, dedicated analysis pipelines are needed to include systematic quality control steps and prevent errors from misusing the proposed methods. RESULTS: SARTools is an R pipeline for differential analysis of RNA-Seq count data. It can handle designs involving two or more conditions of a single biological factor with or without a blocking factor (such as a batch effect or a sample pairing). It is based on DESeq2 and edgeR and is composed of an R package and two R script templates (for DESeq2 and edgeR respectively). Tuning a small number of parameters and executing one of the R scripts, users have access to the full results of the analysis, including lists of differentially expressed genes and a HTML report that (i) displays diagnostic plots for quality control and model hypotheses checking and (ii) keeps track of the whole analysis process, parameter values and versions of the R packages used. CONCLUSIONS: SARTools provides systematic quality controls of the dataset as well as diagnostic plots that help to tune the model parameters. It gives access to the main parameters of DESeq2 and edgeR and prevents untrained users from misusing some functionalities of both packages. By keeping track of all the parameters of the analysis process it fits the requirements of reproducible research.

GH16 and GH81 family ß-(1,3)-glucanases in Aspergillus fumigatus are essential for conidial cell wall morphogenesis.

The fungal cell wall is a rigid structure because of fibrillar and branched ß-(1,3)-glucan linked to chitin. Softening of the cell wall is an essential phenomenon during fungal morphogenesis, wherein rigid cell wall structures are cleaved by glycosylhydrolases. During the search for glycosylhydrolases acting on ß-(1,3)-glucan, we identified seven genes in the Aspergillus fumigatus genome coding for potential endo-ß-(1,3)-glucanase. ENG1 (previously characterized and named ENGL1, Mouyna et al., ), belongs to the Glycoside-Hydrolase 81 (GH81) family, while ENG2 to ENG7, to GH16 family. ENG1 and four GH16 genes (ENG2-5) were expressed in the resting conidia as well as during germination, suggesting an essential role during A. fumigatus morphogenesis. Here, we report the effect of sequential deletion of AfENG2-5 (GH16) followed by AfENG1 (GH81) deletion in the Δeng2,3,4,5 mutant. The Δeng1,2,3,4,5 mutant showed conidial defects, with linear chains of conidia unable to separate while the germination rate was not affected. These results show, for the first time in a filamentous fungus, that endo ß-(1,3)-glucanases are essential for proper conidial cell wall assembly and thus segregation of conidia during conidiation.

Bacteriocin from epidemic Listeria strains alters the host intestinal microbiota to favor infection.

Listeria monocytogenes is responsible for gastroenteritis in healthy individuals and for a severe invasive disease in immunocompromised patients. Among the three identified L. monocytogenes evolutionary lineages, lineage I strains are overrepresented in epidemic listeriosis outbreaks, but the mechanisms underlying the higher virulence potential of strains of this lineage remain elusive. Here, we demonstrate that Listeriolysin S (LLS), a virulence factor only present in a subset of lineage I strains, is a bacteriocin highly expressed in the intestine of orally infected mice that alters the host intestinal microbiota and promotes intestinal colonization by L. monocytogenes, as well as deeper organ infection. To our knowledge, these results therefore identify LLS as the first bacteriocin described in L. monocytogenes and associate modulation of host microbiota by L. monocytogenes epidemic strains to increased virulence.