An atlas of human gene expression from massively parallel signature sequencing (MPSS)

Spatial transcriptome analysis of long non-coding RNAs reveals tissue specificity and functional roles in cancer

Long non-coding RNAs (lncRNAs) play a significant role in maintaining tissue morphology and functions, and their precise regulatory effectiveness is closely related to expression patterns. However, the spatial expression patterns of lncRNAs in humans are poorly characterized. Here, we constructed five comprehensive transcriptomic atlases of human lncRNAs covering thousands of major tissue samples in normal and disease states. The lncRNA transcriptomes exhibited high consistency within the same tissues across resources, and even higher complexity in specialized tissues. Tissue-elevated (TE) lncRNAs were identified in each resource and robust TE lncRNAs were refined by integrative analysis. We detected 1 to 4684 robust TE lncRNAs across tissues; the highest number was in testis tissue, followed by brain tissue. Functional analyses of TE lncRNAs indicated important roles in corresponding tissue-related pathways. Moreover, we found that the expression features of robust TE lncRNAs made them be effective biomarkers to distinguish tissues; TE lncRNAs also tended to be associated with cancer, and exhibited differential expression or were correlated with patient survival. In summary, spatial classification of lncRNAs is the starting point for elucidating the function of lncRNAs in both maintenance of tissue morphology and progress of tissue-constricted diseases.

The Transcriptome and Methylome of the Developing and Aging Brain and Their Relations to Gliomas and Psychological Disorders

Mutually linked expression and methylation dynamics in the brain govern genome regulation over the whole lifetime with an impact on cognition, psychological disorders, and cancer. We performed a joint study of gene expression and DNA methylation of brain tissue originating from the human prefrontal cortex of individuals across the lifespan to describe changes in cellular programs and their regulation by epigenetic mechanisms. The analysis considers previous knowledge in terms of functional gene signatures and chromatin states derived from independent studies, aging profiles of a battery of chromatin modifying enzymes, and data of gliomas and neuropsychological disorders for a holistic view on the development and aging of the brain. Expression and methylation changes from babies to elderly adults decompose into different modes associated with the serial activation of (brain) developmental, learning, metabolic and inflammatory functions, where methylation in gene promoters mostly represses transcription. Expression of genes encoding methylome modifying enzymes is very diverse reflecting complex regulations during lifetime which also associates with the marked remodeling of chromatin between permissive and restrictive states. Data of brain cancer and psychotic disorders reveal footprints of pathophysiologies related to brain development and aging. Comparison of aging brains with gliomas supports the view that glioblastoma-like and astrocytoma-like tumors exhibit higher cellular plasticity activated in the developing healthy brain while oligodendrogliomas have a more stable differentiation hierarchy more resembling the aged brain. The balance and specific shifts between volatile and stable and between more irreversible and more plastic epigenomic networks govern the development and aging of healthy and diseased brain.

Is Phylotranscriptomics as Reliable as Phylogenomics?

Phylogenomics, the study of phylogenetic relationships among taxa based on their genome sequences, has emerged as the preferred phylogenetic method because of the wealth of phylogenetic information contained in genome sequences. Genome sequencing, however, can be prohibitively expensive, especially for taxa with huge genomes and when many taxa need sequencing. Consequently, the less costly phylotranscriptomics has seen an increased use in recent years. Phylotranscriptomics reconstructs phylogenies using DNA sequences derived from transcriptomes, which are often orders of magnitude smaller than genomes. However, in the absence of corresponding genome sequences, comparative analyses of transcriptomes can be challenging and it is unclear whether phylotranscriptomics is as reliable as phylogenomics. Here, we respectively compare the phylogenomic and phylotranscriptomic trees of 22 mammals and 15 plants that have both sequenced nuclear genomes and publicly available RNA sequencing data from multiple tissues. We found that phylotranscriptomic analysis can be sensitive to orthologous gene identification. When a rigorous method for identifying orthologs is employed, phylogenomic and phylotranscriptomic trees are virtually identical to each other, regardless of the tissue of origin of the transcriptomes and whether the same tissue is used across species. These findings validate phylotranscriptomics, brighten its prospect, and illustrate the criticality of reliable ortholog detection in such practices.

Advancing from protein interactomes and gene co-expression networks towards multi-omics-based composite networks: approaches for predicting and extracting biological knowledge

Prediction of biological interaction networks from single-omics data has been extensively implemented to understand various aspects of biological systems. However, more recently, there is a growing interest in integrating multi-omics datasets for the prediction of interactomes that provide a global view of biological systems with higher descriptive capability, as compared to single omics. In this review, we have discussed various computational approaches implemented to infer and analyze two of the most important and well studied interactomes: protein-protein interaction networks and gene co-expression networks. We have explicitly focused on recent methods and pipelines implemented to infer and extract biologically important information from these interactomes, starting from utilizing single-omics data and then progressing towards multi-omics data. Accordingly, recent examples and case studies are also briefly discussed. Overall, this review will provide a proper understanding of the latest developments in protein and gene network modelling and will also help in extracting practical knowledge from them.

Analysis of Differentially Expressed Genes in a Chinese Cohort of Esophageal Squamous Cell Carcinoma

Esophageal squamous cell carcinoma (ESCC) is a leading malignancy in China with both high incidence and mortality. Towards improving outcomes, clinically-relevant biomarkers are urgently needed for use as prognostic and treatment targets. Herein we applied RNA-seq for deep sequencing of ten matched pairs of ESCC and adjacent non-cancerous tissues (NT) from Chinese patients. Transcriptomic data mapped to approximately 64% of all annotated genes with 2,047 and 708 unigenes being differentially up-regulated and down-regulated, respectively, between ESCCs and NT samples (p<0.05). Dividing cases by pathological grade revealed significant differentially expressed genes (DEG) between ESCC and NT in both low and high differentiation cases (p<0.05) whereas gene expression differences were not significantly different between high and low differentiation ESCC tissues (p=0.053). Moreover, the majority of ESCC and NT tissues formed clusters in principal component analyses. The veracity of the DEG list was validated in a larger cohort of 45 patient samples, with down-regulated CLIC3, up-regulated CLIC4 and unchanged expression of CLIC2 confirmed in ESCC using quantitative PCR and Western blotting. Our data reveal both previously identified ESCC biomarkers along with novel candidates and represent a ready resource of DEGs in ESCC for further investigation.

Characterization and analysis of the transcriptome in Gymnocypris selincuoensis on the Qinghai-Tibetan Plateau using single-molecule long-read sequencing and RNA-seq

The lakes on the Qinghai-Tibet Plateau (QTP) are the largest and highest lake group in the world. Gymnocypris selincuoensis is the only cyprinid fish living in lake Selincuo, the largest lake on QTP. However, its genetic resource is still blank, limiting studies on molecular and genetic analysis. In this study, the transcriptome of G. selincuoensis was first generated by using PacBio Iso-Seq and Illumina RNA-seq. A full-length (FL) transcriptome with 75,435 transcripts was obtained by Iso-Seq with N50 length of 3,870 bp. Among all transcripts, 75,016 were annotated to public databases, 64,710 contain complete open reading frames and 2,811 were long non-coding RNAs. Based on all- vs.-all BLAST, 2,069 alternative splicing events were detected, and 80% of them were validated by reverse transcription polymerase chain reaction (RT-PCR). Tissue gene expression atlas showed that the number of detected expressed transcripts ranged from 37,397 in brain to 19,914 in muscle, with 10,488 transcripts detected in all seven tissues. Comparative genomic analysis with other cyprinid fishes identified 77 orthologous genes with potential positive selection (Ka/Ks > 0.3). A total of 56,696 perfect simple sequence repeats were identified from FL transcripts. Our results provide valuable genetic resources for further studies on adaptive evolution, gene expression and population genetics in G. selincuoensis and other congeneric fishes.

Reviving the Transcriptome Studies: An Insight Into the Emergence of Single-Molecule Transcriptome Sequencing

Advances in transcriptomics have provided an exceptional opportunity to study functional implications of the genetic variability. Technologies such as RNA-Seq have emerged as state-of-the-art techniques for transcriptome analysis that take advantage of high-throughput next-generation sequencing. However, similar to their predecessors, these approaches continue to impose major challenges on full-length transcript structure identification, primarily due to inherent limitations of read length. With the development of single-molecule sequencing (SMS) from PacBio, a growing number of studies on the transcriptome of different organisms have been reported. SMS has emerged as advantageous for comprehensive genome annotation including identification of novel genes/isoforms, long non-coding RNAs and fusion transcripts. This approach can be used across a broad spectrum of species to better interpret the coding information of the genome, and facilitate the biological function study. We provide an overview of SMS platform and its diverse applications in various biological studies, and our perspective on the challenges associated with the transcriptome studies.

Footprints of Sepsis Framed Within Community Acquired Pneumonia in the Blood Transcriptome

We analyzed the blood transcriptome of sepsis framed within community-acquired pneumonia (CAP) and characterized its molecular and cellular heterogeneity in terms of functional modules of co-regulated genes with impact for the underlying pathophysiological mechanisms. Our results showed that CAP severity is associated with immune suppression owing to T-cell exhaustion and HLA and chemokine receptor deactivation, endotoxin tolerance, macrophage polarization, and metabolic conversion from oxidative phosphorylation to glycolysis. We also found footprints of host's response to viruses and bacteria, altered levels of mRNA from erythrocytes and platelets indicating coagulopathy that parallel severity of sepsis and survival. Finally, our data demonstrated chromatin re-modeling associated with extensive transcriptional deregulation of chromatin modifying enzymes, which suggests the extensive changes of DNA methylation with potential impact for marker selection and functional characterization. Based on the molecular footprints identified, we propose a novel stratification of CAP cases into six groups differing in the transcriptomic scores of CAP severity, interferon response, and erythrocyte mRNA expression with impact for prognosis. Our analysis increases the resolution of transcriptomic footprints of CAP and reveals opportunities for selecting sets of transcriptomic markers with impact for translation of omics research in terms of patient stratification schemes and sets of signature genes.

SV-BR-1-GM, a Clinically Effective GM-CSF-Secreting Breast Cancer Cell Line, Expresses an Immune Signature and Directly Activates CD4+ T Lymphocytes

Targeted cancer immunotherapy with irradiated, granulocyte–macrophage colony-stimulating factor (GM-CSF)-secreting, allogeneic cancer cell lines has been an effective approach to reduce tumor burden in several patients. It is generally assumed that to be effective, these cell lines need to express immunogenic antigens coexpressed in patient tumor cells, and antigen-presenting cells need to take up such antigens then present them to patient T cells. We have previously reported that, in a phase I pilot study (ClinicalTrials.gov NCT00095862), a subject with stage IV breast cancer experienced substantial regression of breast, lung, and brain lesions following inoculation with clinical formulations of SV-BR-1-GM, a GM-CSF-secreting breast tumor cell line. To identify diagnostic features permitting the prospective identification of patients likely to benefit from SV-BR-1-GM, we conducted a molecular analysis of the SV-BR-1-GM cell line and of patient-derived blood, as well as a tumor specimen. Compared to normal human breast cells, SV-BR-1-GM cells overexpress genes encoding tumor-associated antigens (TAAs) such as PRAME, a cancer/testis antigen. Curiously, despite its presumptive breast epithelial origin, the cell line expresses major histocompatibility complex (MHC) class II genes (HLA-DRA, HLA-DRB3, HLA-DMA, HLA-DMB), in addition to several other factors known to play immunostimulatory roles. These factors include MHC class I components (B2M, HLA-A, HLA-B), ADA (encoding adenosine deaminase), ADGRE5 (CD97), CD58 (LFA3), CD74 (encoding invariant chain and CLIP), CD83, CXCL8 (IL8), CXCL16, HLA-F, IL6, IL18, and KITLG. Moreover, both SV-BR-1-GM cells and the responding study subject carried an HLA-DRB3*02:02 allele, raising the question of whether SV-BR-1-GM cells can directly present endogenous antigens to T cells, thereby inducing a tumor-directed immune response. In support of this, SV-BR-1-GM cells (which also carry the HLA-DRB3*01:01 allele) treated with yellow fever virus (YFV) envelope (Env) 43–59 peptides reactivated YFV-DRB3*01:01-specific CD4⁺ T cells. Thus, the partial HLA allele match between SV-BR-1-GM and the clinical responder might have enabled patient T lymphocytes to directly recognize SV-BR-1-GM TAAs as presented on SV-BR-1-GM MHCs. Taken together, our findings are consistent with a potentially unique mechanism of action by which SV-BR-1-GM cells can act as APCs for previously primed CD4⁺ T cells.

Comparative study of gastric cancer and chronic gastritis via network analysis

AIM

In this study the significant differentially expressed genes (DEGs) related to gastric cancer (GC) and chronic gastritis were screened to introduce common and distinctive genes between the two diseases.

BACKGROUND

Diagnosis of gastric cancer as a mortal disease and chronic gastritis the stomach disorder which can be considered as risk factor of GCs required safe and effective molecular biomarkers.

METHODS

Microarray profiles were downloaded from Gene Expression Omnibus (GEO) and analyzed via GEO2R. The candidate DEGs plus relevant genes from STRING database were interacted by Cytoscape software version 3.6.0 the central nodes were determined and analyzed.

RESULTS

JUN, GAPDH, FOS, TP53, PRDM10, VEGFA, and CREB1 as central nodes and TFF1 and ERG1 as the top changed expressed genes were determined as critical nodes related to gastric cancer. GAPDH, PRDM10, TP53, JUN, AKT1, EGFR, MAPK1, EGF, DECR1, and MYC were identified as common remarkable genes between GC and chronic gastritis.

CONCLUSION

Identification of distinctive and common genes between GC and chronic gastritis can be useful in the early stage detection of disease and reducing risk of GCs.

Digging More Missing Proteins Using an Enrichment Approach with ProteoMiner

Human Proteome Project (HPP) aims at mapping entire human proteins with a systematic effort upon all the emerging techniques, which would enhance understanding of human biology and lay a foundation for development of medical applications. Until now, 2563 missing proteins (MPs, PE2-4) are still undetected even using the most sensitive approach of protein detection. Herein, we propose that enrichment of low-abundance proteins benefits MPs finding. ProteoMiner is an equalizing technique by reducing high-abundance proteins and enriching low-abundance proteins in biological liquids. With triton X-100/TBS buffer extraction, ProteoMiner enrichment, and peptide fractionation, 20 MPs (at least two non-nested unique peptides with more than eight a.a. length) with 60 unique peptides were identified from four human tissues including eight membrane/secreted proteins and five nucleus proteins. Then 15 of them were confirmed with two non-nested unique peptides (≥9 a.a.) identified by matching well with their chemically synthetic peptides in PRM assay. Hence, these results demonstrated ProteoMiner as a powerful means in discovery of MPs.

Comprehensive Analysis of Tissue-wide Gene Expression and Phenotype Data Reveals Tissues Affected in Rare Genetic Disorders

Linking putatively pathogenic variants to the tissues they affect is necessary for determining the correct diagnostic workup and therapeutic regime in undiagnosed patients. Here, we explored how gene expression across healthy tissues can be used to infer this link. We integrated 6,665 tissue-wide transcriptomes with genetic disorder knowledge bases covering 3,397 diseases. Receiver-operating characteristics (ROC) analysis using expression levels in each tissue and across tissues indicated significant but modest associations between elevated expression and phenotype for most tissues (maximum area under ROC curve = 0.69). At extreme elevation, associations were marked. Upregulation of disease genes in affected tissues was pronounced for genes associated with autosomal dominant over recessive disorders. Pathways enriched for genes expressed and associated with phenotypes highlighted tissue functionality, including lipid metabolism in spleen and DNA repair in adipose tissue. These results suggest features useful for evaluating the likelihood of particular tissue manifestations in genetic disorders. The web address of an interactive platform integrating these data is provided.

In vivo mapping of tissue- and subcellular-specific proteomes in Caenorhabditis elegans

Multicellular organisms are composed of tissues that have distinct functions requiring specialized proteomes. To define the proteome of a live animal with tissue and subcellular resolution, we adapted a localized proteomics technology for use in the multicellular model organism Caenorhabditis elegans. This approach couples tissue- and location-specific expression of the enzyme ascorbate peroxidase (APX), which enables proximity-based protein labeling in vivo, and quantitative proteomics to identify tissue- and subcellular-restricted proteomes. We identified and localized more than 3000 proteins from strains of C. elegans expressing APX in either the nucleus or cytoplasm of the intestine, epidermis, body wall muscle, or pharyngeal muscle. We also identified several hundred proteins that were specifically localized to one of the four tissues analyzed or specifically localized to the cytoplasm or the nucleus. This approach resulted in the identification both of proteins with previously characterized localizations and of those not known to localize to the nucleus or cytoplasm. Further, we confirmed the tissue- and subcellular-specific localization of a subset of identified proteins using green fluorescent protein tagging and fluorescence microscopy, validating our in vivo proximity-based proteomics technique. Together, these results demonstrate a new approach that enables the tissue- and subcellular-specific identification and quantification of proteins within a live animal.

Analysis of gene expression in the nervous system identifies key genes and novel candidates for health and disease

The incidence of neurodegenerative diseases in the developed world has risen over the last century, concomitant with an increase in average human lifespan. A major challenge is therefore to identify genes that control neuronal health and viability with a view to enhancing neuronal health during ageing and reducing the burden of neurodegeneration. Analysis of gene expression data has recently been used to infer gene functions for a range of tissues from co-expression networks. We have now applied this approach to transcriptomic datasets from the mammalian nervous system available in the public domain. We have defined the genes critical for influencing neuronal health and disease in different neurological cell types and brain regions. The functional contribution of genes in each co-expression cluster was validated using human disease and knockout mouse phenotypes, pathways and gene ontology term annotation. Additionally a number of poorly annotated genes were implicated by this approach in nervous system function. Exploiting gene expression data available in the public domain allowed us to validate key nervous system genes and, importantly, to identify additional genes with minimal functional annotation but with the same expression pattern. These genes are thus novel candidates for a role in neurological health and disease and could now be further investigated to confirm their function and regulation during ageing and neurodegeneration.

Integrative analysis of human omics data using biomolecular networks

High-throughput '-omics' technologies have given rise to an increasing abundance of genome-scale data detailing human biology at the molecular level. Although these datasets have already made substantial contributions to a more comprehensive understanding of human physiology and diseases, their interpretation becomes increasingly cryptic and nontrivial as they continue to expand in size and complexity. Systems biology networks offer a scaffold upon which omics data can be integrated, facilitating the extraction of new and physiologically relevant information from the data. Two of the most prevalent networks that have been used for such integrative analyses of omics data are genome-scale metabolic models (GEMs) and protein-protein interaction (PPI) networks, both of which have demonstrated success among many different omics and sample types. This integrative approach seeks to unite 'top-down' omics data with 'bottom-up' biological networks in a synergistic fashion that draws on the strengths of both strategies. As the volume and resolution of high-throughput omics data continue to grow, integrative network-based analyses are expected to play an increasingly important role in their interpretation.

Microarray Gene Expression Analysis to Evaluate Cell Type Specific Expression of Targets Relevant for Immunotherapy of Hematological Malignancies

Cellular immunotherapy has proven to be effective in the treatment of hematological cancers by donor lymphocyte infusion after allogeneic hematopoietic stem cell transplantation and more recently by targeted therapy with chimeric antigen or T-cell receptor-engineered T cells. However, dependent on the tissue distribution of the antigens that are targeted, anti-tumor responses can be accompanied by undesired side effects. Therefore, detailed tissue distribution analysis is essential to estimate potential efficacy and toxicity of candidate targets for immunotherapy of hematological malignancies. We performed microarray gene expression analysis of hematological malignancies of different origins, healthy hematopoietic cells and various non-hematopoietic cell types from organs that are often targeted in detrimental immune responses after allogeneic stem cell transplantation leading to graft-versus-host disease. Non-hematopoietic cells were also cultured in the presence of IFN-γ to analyze gene expression under inflammatory circumstances. Gene expression was investigated by Illumina HT12.0 microarrays and quality control analysis was performed to confirm the cell-type origin and exclude contamination of non-hematopoietic cell samples with peripheral blood cells. Microarray data were validated by quantitative RT-PCR showing strong correlations between both platforms. Detailed gene expression profiles were generated for various minor histocompatibility antigens and B-cell surface antigens to illustrate the value of the microarray dataset to estimate efficacy and toxicity of candidate targets for immunotherapy. In conclusion, our microarray database provides a relevant platform to analyze and select candidate antigens with hematopoietic (lineage)-restricted expression as potential targets for immunotherapy of hematological cancers.

Alcohol-free fermented blueberry-blackberry beverage phenolic extract attenuates diet-induced obesity and blood glucose in C57BL/6J mice

The aim of this study was to determine the potential of phenolic compounds from a fermented blackberry-blueberry beverage to reduce diet-induced obesity and hyperglycemia in mice fed a 60% high-fat diet (HFD) for 10weeks after 1week of pretreatment. C57BL/6J mice were randomized into six groups and allowed to drink (ad libitum) an alcohol-free blackberry-blueberry beverage [alcohol-free fermented beverage (AFFB), 8.4mg anthocyanin (ANC)/kg body weight (BW)/day]; three doses of a phenolic extract [postamberlite extract (PAE)] from AFFB at 0.1×, 1× and 2× ANC concentrations; sitagliptin (hypoglycemic positive control); or water (negative control). Weight and fat mass gain were attenuated in mice receiving the highest doses of PAE (18.9mg ANC/kg BW/day, P<.05). There were also reductions (P<.05) in percent fat mass, epididymal fat pad weights, mean adipocyte diameters and plasma triglycerides and cholesterol associated with PAE treatments. By the end of the study, fasting blood glucose for mice receiving 9mg (1×) or 18.9mg (2×) ANC/kg BW/day was significantly lower than in the water and the sitagliptin groups (P<.05). Histological and histochemical analyses revealed an unexpected change in liver of mice fed ANC at 1× or 2× doses consisting of liver enlargement and increased lipid deposition. PAE also induced the most differential gene expression changes, including highly significant downstream effects at all doses to reduce d-glucose concentrations. Overall, phenolic compounds from the fermented blueberry-blackberry beverage had an impact to attenuate the development of obesity and fasting blood glucose in C57BL/6J mice.

De novo transcriptome analyses of host-fungal interactions in oil palm (Elaeis guineensis Jacq.)

Background

Basal stem rot (BSR) is a fungal disease in oil palm (Elaeis guineensis Jacq.) which is caused by hemibiotrophic white rot fungi belonging to the Ganoderma genus. Molecular responses of oil palm to these pathogens are not well known although this information is crucial to strategize effective measures to eradicate BSR. In order to elucidate the molecular interactions between oil palm and G. boninense and its biocontrol fungus Trichoderma harzianum, we compared the root transcriptomes of untreated oil palm seedlings with those inoculated with G. boninense and T. harzianum, respectively.

Results

Differential gene expression analyses revealed that jasmonate (JA) and salicylate (SA) may act in an antagonistic manner in affecting the hormone biosynthesis, signaling, and downstream defense responses in G. boninense-treated oil palm roots. In addition, G. boninense may compete with the host to control disease symptom through the transcriptional regulation of ethylene (ET) biosynthesis, reactive oxygen species (ROS) production and scavenging. The strengthening of host cell walls and production of pathogenesis-related proteins as well as antifungal secondary metabolites in host plants, are among the important defense mechanisms deployed by oil palm against G. boninense. Meanwhile, endophytic T. harzianum was shown to improve the of nutrition status and nutrient transportation in host plants.

Conclusion

The findings of this analysis have enhanced our understanding on the molecular interactions of G. boninense and oil palm, and also the biocontrol mechanisms involving T. harzianum, thus contributing to future formulations of better strategies for prevention and treatment of BSR.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-016-2368-0) contains supplementary material, which is available to authorized users.

A common molecular signature in ASD gene expression: following Root 66 to autism

Several gene expression experiments on autism spectrum disorders have been conducted using both blood and brain tissue. Individually, these studies have advanced our understanding of the molecular systems involved in the molecular pathology of autism and have formed the bases of ongoing work to build autism biomarkers. In this study, we conducted an integrated systems biology analysis of 9 independent gene expression experiments covering 657 autism, 9 mental retardation and developmental delay and 566 control samples to determine if a common signature exists and to test whether regulatory patterns in the brain relevant to autism can also be detected in blood. We constructed a matrix of differentially expressed genes from these experiments and used a Jaccard coefficient to create a gene-based phylogeny, validated by bootstrap. As expected, experiments and tissue types clustered together with high statistical confidence. However, we discovered a statistically significant subgrouping of 3 blood and 2 brain data sets from 3 different experiments rooted by a highly correlated regulatory pattern of 66 genes. This Root 66 appeared to be non-random and of potential etiologic relevance to autism, given their enriched roles in neurological processes key for normal brain growth and function, learning and memory, neurodegeneration, social behavior and cognition. Our results suggest that there is a detectable autism signature in the blood that may be a molecular echo of autism-related dysregulation in the brain.

MGFM: a novel tool for detection of tissue and cell specific marker genes from microarray gene expression data

Background

Identification of marker genes associated with a specific tissue/cell type is a fundamental challenge in genetic and cell research. Marker genes are of great importance for determining cell identity, and for understanding tissue specific gene function and the molecular mechanisms underlying complex diseases.

Results

We have developed a new bioinformatics tool called MGFM (Marker Gene Finder in Microarray data) to predict marker genes from microarray gene expression data. Marker genes are identified through the grouping of samples of the same type with similar marker gene expression levels. We verified our approach using two microarray data sets from the NCBI’s Gene Expression Omnibus public repository encompassing samples for similar sets of five human tissues (brain, heart, kidney, liver, and lung). Comparison with another tool for tissue-specific gene identification and validation with literature-derived established tissue markers established functionality, accuracy and simplicity of our tool. Furthermore, top ranked marker genes were experimentally validated by reverse transcriptase-polymerase chain reaction (RT-PCR). The sets of predicted marker genes associated with the five selected tissues comprised well-known genes of particular importance in these tissues. The tool is freely available from the Bioconductor web site, and it is also provided as an online application integrated into the CellFinder platform (http://cellfinder.org/analysis/marker).

Conclusions

MGFM is a useful tool to predict tissue/cell type marker genes using microarray gene expression data. The implementation of the tool as an R-package as well as an application within CellFinder facilitates its use.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1785-9) contains supplementary material, which is available to authorized users.

Human protein interaction networks across tissues and diseases

Protein interaction networks are an important framework for studying protein function, cellular processes, and genotype-to-phenotype relationships. While our view of the human interaction network is constantly expanding, less is known about networks that form in biologically important contexts such as within distinct tissues or in disease conditions. Here we review efforts to characterize these networks and to harness them to gain insights into the molecular mechanisms underlying human disease.

A Genome-Wide Landscape of Retrocopies in Primate Genomes

Gene duplication is a key factor contributing to phenotype diversity across and within species. Although the availability of complete genomes has led to the extensive study of genomic duplications, the dynamics and variability of gene duplications mediated by retrotransposition are not well understood. Here, we predict mRNA retrotransposition and use comparative genomics to investigate their origin and variability across primates. Analyzing seven anthropoid primate genomes, we found a similar number of mRNA retrotranspositions (∼7,500 retrocopies) in Catarrhini (Old Word Monkeys, including humans), but a surprising large number of retrocopies (∼10,000) in Platyrrhini (New World Monkeys), which may be a by-product of higher long interspersed nuclear element 1 activity in these genomes. By inferring retrocopy orthology, we dated most of the primate retrocopy origins, and estimated a decrease in the fixation rate in recent primate history, implying a smaller number of species-specific retrocopies. Moreover, using RNA-Seq data, we identified approximately 3,600 expressed retrocopies. As expected, most of these retrocopies are located near or within known genes, present tissue-specific and even species-specific expression patterns, and no expression correlation to their parental genes. Taken together, our results provide further evidence that mRNA retrotransposition is an active mechanism in primate evolution and suggest that retrocopies may not only introduce great genetic variability between lineages but also create a large reservoir of potentially functional new genomic loci in primate genomes.

How Many Genes Are Expressed in a Transcriptome? Estimation and Results for RNA-Seq

RNA-seq experiments estimate the number of genes expressed in a transcriptome as well as their relative frequencies. However, an undetermined number of genes can remain undetected due to their low expression relative to the sample size (sequence depth). Estimation of the true number of genes expressed in a transcriptome is essential in order to determine which genes are exclusively expressed in specific tissues or under particular conditions. A reliable estimate of the true number of expressed genes is also required to accurately measure transcriptome changes and to predict the sequencing depth needed to increase the proportion of detected genes. This problem is analogous to ecological sampling problems such as estimating the number of species at a given site. Here we present a non-parametric estimator for the number of undetected genes as well as for the extra sample size needed to detect a given proportion of the undetected genes. Our estimators are superior to ones already published by having smaller standard errors and biases. We applied our method to a set of 32 publicly available RNA-seq experiments, including the evaluation of 311 individually sequenced libraries. We found that in the majority of the cases more than one thousand genes are undetected, and that on average approximately 6% of the expressed genes per accession remain undetected. This figure increases to approximately 10% if individual sequencing libraries are analyzed. Our method is also applicable to metagenomic experiments. Using our method, the number of undetected genes as well as the sample size needed to detect them can be calculated, leading to more accurate and complete gene expression studies.

Do housekeeping genes exist?

The searching of human housekeeping (HK) genes has been a long quest since the emergence of transcriptomics, and is instrumental for us to understand the structure of genome and the fundamentals of biological processes. The resolved genes are frequently used in evolution studies and as normalization standards in quantitative gene-expression analysis. Within the past 20 years, more than a dozen HK-gene studies have been conducted, yet none of them sampled human tissues completely. We believe an integration of these results will help remove false positive genes owing to the inadequate sampling. Surprisingly, we only find one common gene across 15 examined HK-gene datasets comprising 187 different tissue and cell types. Our subsequent analyses suggest that it might not be appropriate to rigidly define HK genes as expressed in all tissue types that have diverse developmental, physiological, and pathological states. It might be beneficial to use more robustly identified HK functions for filtering criteria, in which the representing genes can be a subset of genome. These genes are not necessarily the same, and perhaps need not to be the same, everywhere in our body.

Transcriptome assembly, gene annotation and tissue gene expression atlas of the rainbow trout

Efforts to obtain a comprehensive genome sequence for rainbow trout are ongoing and will be complemented by transcriptome information that will enhance genome assembly and annotation. Previously, transcriptome reference sequences were reported using data from different sources. Although the previous work added a great wealth of sequences, a complete and well-annotated transcriptome is still needed. In addition, gene expression in different tissues was not completely addressed in the previous studies. In this study, non-normalized cDNA libraries were sequenced from 13 different tissues of a single doubled haploid rainbow trout from the same source used for the rainbow trout genome sequence. A total of ~1.167 billion paired-end reads were de novo assembled using the Trinity RNA-Seq assembler yielding 474,524 contigs > 500 base-pairs. Of them, 287,593 had homologies to the NCBI non-redundant protein database. The longest contig of each cluster was selected as a reference, yielding 44,990 representative contigs. A total of 4,146 contigs (9.2%), including 710 full-length sequences, did not match any mRNA sequences in the current rainbow trout genome reference. Mapping reads to the reference genome identified an additional 11,843 transcripts not annotated in the genome. A digital gene expression atlas revealed 7,678 housekeeping and 4,021 tissue-specific genes. Expression of about 16,000–32,000 genes (35–71% of the identified genes) accounted for basic and specialized functions of each tissue. White muscle and stomach had the least complex transcriptomes, with high percentages of their total mRNA contributed by a small number of genes. Brain, testis and intestine, in contrast, had complex transcriptomes, with a large numbers of genes involved in their expression patterns. This study provides comprehensive de novo transcriptome information that is suitable for functional and comparative genomics studies in rainbow trout, including annotation of the genome.

The top skin-associated genes: a comparative analysis of human and mouse skin transcriptomes

The mouse represents a key model system for the study of the physiology and biochemistry of skin. Comparison of skin between mouse and human is critical for interpretation and application of data from mouse experiments to human disease. Here, we review the current knowledge on structure and immunology of mouse and human skin. Moreover, we present a systematic comparison of human and mouse skin transcriptomes. To this end, we have recently used a genome-wide database of human gene expression to identify genes highly expressed in skin, with no, or limited expression elsewhere - human skin-associated genes (hSAGs). Analysis of our set of hSAGs allowed us to generate a comprehensive molecular characterization of healthy human skin. Here, we used a similar database to generate a list of mouse skin-associated genes (mSAGs). A comparative analysis between the top human (n=666) and mouse (n=873) skin-associated genes (SAGs) revealed a total of only 30.2% identity between the two lists. The majority of shared genes encode proteins that participate in structural and barrier functions. Analysis of the top functional annotation terms revealed an overlap for morphogenesis, cell adhesion, structure, and signal transduction. The results of this analysis, discussed in the context of published data, illustrate the diversity between the molecular make up of skin of both species and grants a probable explanation, why results generated in murine in vivo models often fail to translate into the human.

The lung-specific proteome defined by integration of transcriptomics and antibody-based profiling

The combined action of multiple cell types is essential for the physiological function of the lung, and increased awareness of the molecular constituents characterizing each cell type is likely to advance the understanding of lung biology and disease. In the current study, we used genome-wide RNA sequencing of normal lung parenchyma and 26 additional tissue types, combined with antibody-based protein profiling, to localize the expression to specific cell types. Altogether, 221 genes were found to be elevated in the lung compared with their expression in other analyzed tissues. Among the gene products were several well-known markers, but also several proteins previously not described in the context of the lung. To link the lung-specific molecular repertoire to human disease, survival associations of pneumocyte-specific genes were assessed by using transcriptomics data from 7 non-small-cell lung cancer (NSCLC) cohorts. Transcript levels of 10 genes (SFTPB, SFTPC, SFTPD, SLC34A2, LAMP3, CACNA2D2, AGER, EMP2, NKX2-1, and NAPSA) were significantly associated with survival in the adenocarcinoma subgroup, thus qualifying as promising biomarker candidates. In summary, based on an integrated omics approach, we identified genes with elevated expression in lung and localized corresponding protein expression to different cell types. As biomarker candidates, these proteins may represent intriguing starting points for further exploration in health and disease.

CD99 is upregulated in placenta and astrocytomas with a differential subcellular distribution according to the malignancy stage

In the present study, we searched for genes highly expressed in placenta and that could contribute to the establishment and maintenance of a malignant phenotype in different types of tumours, and in astrocytomas in particular. We employed a strategy based on the integration of in silico data from previously generated massively parallel signature sequencing and public serial analysis of gene expression databases. Among 12 selected genes, CD99 exhibited the highest relative mRNA expression in GBM compared to non-neoplastic brain tissues. In a larger cohort of astrocytic tumours, we further demonstrated increased CD99 expression in all malignant grades, with GBMs showing the highest values. These findings were confirmed at the protein level by Western blotting and immunohistochemistry. Additionally, we demonstrated the CD99 localisation profile in astrocytic tumours. Interestingly, CD99 expression was confined to the cytoplasm or membrane in more malignant astrocytomas, in contrast to non-neoplastic brain tissue or non-infiltrative pilocytic astrocytoma, which showed no obvious staining in these structures. Comparison of three GBM cell lines revealed higher CD99 expression at the membrane and higher migratory capacity in the A172 and U87MG lines, but lower CD99 expression and no migratory ability in the T98 line. Knocking down CD99 expression by siRNA decreased significantly the migration of both cell lines. These integrated CD99 gene and protein expression results suggest that CD99 expression in astrocytomas of different malignant grades might contribute to the infiltrative ability and support the importance of CD99 as a potential target to reduce infiltrative astrocytoma capacity in migration and invasion.

RNA interference against cancer/testis genes identifies dual specificity phosphatase 21 as a potential therapeutic target in human hepatocellular carcinoma

Cancer/testis (CT) antigens have been considered therapeutic targets for treating cancers. However, a central question is whether their expression contributes to tumorigenesis or if they are functionally irrelevant by-products derived from the process of cellular transformation. In any case, these CT antigens are essential for cancer cell survival and may serve as potential therapeutic targets. Recently, the cell-based RNA interference (RNAi) screen has proven to be a powerful approach for identifying potential therapeutic targets. In this study we sought to identify new CT antigens as potential therapeutic targets for human hepatocellular carcinoma (HCC), and 179 potential CT genes on the X chromosome were screened through a bioinformatics analysis of gene expression profiles. Then an RNAi screen against these potential CT genes identified nine that were required for sustaining the survival of Focus and PLC/PRF/5 cells. Among the nine genes, the physiologically testis-restricted dual specificity phosphatase 21 (DUSP21) encoding a dual specificity phosphatase was up-regulated in 39 (33%) of 118 human HCC specimens. Ectopic DUSP21 had no obvious impact on proliferation and colony formation in HCC cells. However, DUSP21 silencing significantly suppressed cell proliferation, colony formation, and in vivo tumorigenicity in HCC cells. The administration of adenovirus-mediated RNAi and an atelocollagen/siRNA mixture against endogenous DUSP21 significantly suppressed xenograft HCC tumors in mice. Further investigations showed that DUSP21 knockdown led to arrest of the cell cycle in G1 phase, cell senescence, and expression changes of some factors with functions in the cell cycle and/or senescence. Furthermore, the antiproliferative role of DUSP21 knockdown is through activation of p38 mitogen-activated protein kinase in HCC.

CONCLUSION

DUSP21 plays an important role in sustaining HCC cell proliferation and may thus act as a potential therapeutic target in HCC treatment.

Proteomic profiling of exosomes leads to the identification of novel biomarkers for prostate cancer

BACKGROUND

Current markers for prostate cancer, such as PSA lack specificity. Therefore, novel biomarkers are needed. Unfortunately, the complexity of body fluids often hampers biomarker discovery. An attractive alternative approach is the isolation of small vesicles, i.e. exosomes, ∼100 nm, which contain proteins that are specific to the tissue from which they are derived and therefore can be considered as treasure chests for disease-specific biomarker discovery.

MATERIALS AND METHODS

Exosomes were isolated from 2 immortalized primary prostate epithelial cells (PNT2C2 and RWPE-1) and 2 PCa cell lines (PC346C and VCaP) by ultracentrifugation. After tryptic digestion, proteomic analyses utilized a nanoLC coupled with an LTQ-Orbitrap operated in tandem MS (MS/MS) mode. Accurate Mass and Time (AMT) tag approach was employed for peptide identification and quantitation. Candidate biomarkers were validated by Western blotting and Immunohistochemistry.

RESULTS

Proteomic characterization resulted in the identification of 248, 233, 169, and 216 proteins by at least 2 peptides in exosomes from PNT2C2, RWPE-1, PC346C, and VCaP, respectively. Statistical analyses revealed 52 proteins differently abundant between PCa and control cells, 9 of which were more abundant in PCa. Validation by Western blotting confirmed a higher abundance of FASN, XPO1 and PDCD6IP (ALIX) in PCa exosomes.

CONCLUSIONS

Identification of exosomal proteins using high performance LC-FTMS resulted in the discovery of PDCD6IP, FASN, XPO1 and ENO1 as new candidate biomarkers for prostate cancer.

Optimal scaling of digital transcriptomes

Deep sequencing of transcriptomes has become an indispensable tool for biology, enabling expression levels for thousands of genes to be compared across multiple samples. Since transcript counts scale with sequencing depth, counts from different samples must be normalized to a common scale prior to comparison. We analyzed fifteen existing and novel algorithms for normalizing transcript counts, and evaluated the effectiveness of the resulting normalizations. For this purpose we defined two novel and mutually independent metrics: (1) the number of “uniform” genes (genes whose normalized expression levels have a sufficiently low coefficient of variation), and (2) low Spearman correlation between normalized expression profiles of gene pairs. We also define four novel algorithms, one of which explicitly maximizes the number of uniform genes, and compared the performance of all fifteen algorithms. The two most commonly used methods (scaling to a fixed total value, or equalizing the expression of certain ‘housekeeping’ genes) yielded particularly poor results, surpassed even by normalization based on randomly selected gene sets. Conversely, seven of the algorithms approached what appears to be optimal normalization. Three of these algorithms rely on the identification of “ubiquitous” genes: genes expressed in all the samples studied, but never at very high or very low levels. We demonstrate that these include a “core” of genes expressed in many tissues in a mutually consistent pattern, which is suitable for use as an internal normalization guide. The new methods yield robustly normalized expression values, which is a prerequisite for the identification of differentially expressed and tissue-specific genes as potential biomarkers.

Targeted re-sequencing identified rs3106189 at the 5' UTR of TAPBP and rs1052918 at the 3' UTR of TCF3 to be associated with the overall survival of colorectal cancer patients

Recent studies have demonstrated the power of deep re-sequencing of the whole genome or exome in understanding cancer genomes. However, targeted capture of selected genomic whole gene-body regions, rather than the whole exome, have several advantages: 1) the genes can be selected based on biology or a hypothesis; 2) mutations in promoter and intronic regions, which have important regulatory roles, can be investigated; and 3) less expensive than whole genome or whole exome sequencing. Therefore, we designed custom high-density oligonucleotide microarrays (NimbleGen Inc.) to capture approximately 1.7 Mb target regions comprising the genomic regions of 28 genes related to colorectal cancer including genes belonging to the WNT signaling pathway, as well as important transcription factors or colon-specific genes that are over expressed in colorectal cancer (CRC). The 1.7 Mb targeted regions were sequenced with a coverage ranged from 32× to 45× for the 28 genes. We identified a total of 2342 sequence variations in the CRC and corresponding adjacent normal tissues. Among them, 738 were novel sequence variations based on comparisons with the SNP database (dbSNP135). We validated 56 of 66 SNPs in a separate cohort of 30 CRC tissues using Sequenom MassARRAY iPLEX Platform, suggesting a validation rate of at least 85% (56/66). We found 15 missense mutations among the exonic variations, 21 synonymous SNPs that were predicted to change the exonic splicing motifs, 31 UTR SNPs that were predicted to occur at the transcription factor binding sites, 20 intronic SNPs located near the splicing sites, 43 SNPs in conserved transcription factor binding sites and 32 in CpG islands. Finally, we determined that rs3106189, localized to the 5′ UTR of antigen presenting tapasin binding protein (TAPBP), and rs1052918, localized to the 3′ UTR of transcription factor 3 (TCF3), were associated with overall survival of CRC patients.

In vitro-differentiated neural cell cultures progress towards donor-identical brain tissue

Multiple research groups have observed neuropathological phenotypes and molecular symptoms in vitro using induced pluripotent stem cell (iPSC)-derived neural cell cultures (i.e. patient-specific neurons and glia). However, the global differences/similarities that may exist between in vitro neural cells and their tissue-derived counterparts remain largely unknown. In this study, we compared temporal series of iPSC-derived in vitro neural cell cultures to endogenous brain tissue from the same autopsy donor. Specifically, we utilized RNA sequencing (RNA-Seq) to evaluate the transcriptional progression of in vitro-differentiated neural cells (over a timecourse of 0, 35, 70, 105 and 140 days), and compared this with donor-identical temporal lobe tissue. We observed in vitro progression towards the reference brain tissue, and the following three results support this conclusion: (i) there was a significant increasing monotonic correlation between the days of our timecourse and the number of actively transcribed protein-coding genes and long intergenic non-coding RNAs (lincRNAs) (P < 0.05), consistent with the transcriptional complexity of the brain; (ii) there was an increase in CpG methylation after neural differentiation that resembled the epigenomic signature of the endogenous tissue; and (iii) there was a significant decreasing monotonic correlation between the days of our timecourse and the percent of in vitro to brain-tissue differences (P < 0.05) for tissue-specific protein-coding genes and all putative lincRNAs. Taken together, these results are consistent with in vitro neural development and physiological progression occurring predominantly by transcriptional activation of downregulated genes rather than deactivation of upregulated genes.

Alternative polyadenylation in glioblastoma multiforme and changes in predicted RNA binding protein profiles

Alternative polyadenylation (APA) is widely present in the human genome and plays a key role in carcinogenesis. We conducted a comprehensive analysis of the APA products in glioblastoma multiforme (GBM, one of the most lethal brain tumors) and normal brain tissues and further developed a computational pipeline, RNAelements (http://sysbio.zju.edu.cn/RNAelements/), using covariance model from known RNA binding protein (RBP) targets acquired by RNA Immunoprecipitation (RIP) analysis. We identified 4530 APA isoforms for 2733 genes in GBM, and found that 182 APA isoforms from 148 genes showed significant differential expression between normal and GBM brain tissues. We then focused on three genes with long and short APA isoforms that show inconsistent expression changes between normal and GBM brain tissues. These were myocyte enhancer factor 2D, heat shock factor binding protein 1, and polyhomeotic homolog 1 (Drosophila). Using the RNAelements program, we found that RBP binding sites were enriched in the alternative regions between the first and the last polyadenylation sites, which would result in the short APA forms escaping regulation from those RNA binding proteins. To the best of our knowledge, this report is the first comprehensive APA isoform dataset for GBM and normal brain tissues. Additionally, we demonstrated a putative novel APA-mediated mechanism for controlling RNA stability and translation for APA isoforms. These observations collectively lay a foundation for novel diagnostics and molecular mechanisms that can inform future therapeutic interventions for GBM.

Evolutionary conservation and functional roles of ncRNA

Non-coding RNAs (ncRNAs) are a class of transcribed RNA molecules without protein-coding potential. They were regarded as transcriptional noise, or the byproduct of genetic information flow from DNA to protein for a long time. However, in recent years, a number of studies have shown that ncRNAs are pervasively transcribed, and most of them show evidence of evolutionary conservation, although less conserved than protein-coding genes. More importantly, many ncRNAs have been confirmed as playing crucial regulatory roles in diverse biological processes and tumorigenesis. Here we summarize the functional significance of this class of “dark matter” in terms its genomic organization, evolutionary conservation, and broad functional classes.

Impact of experience-dependent and -independent factors on gene expression in songbird brain

Songbirds provide rich natural models for studying the relationships between brain anatomy, behavior, environmental signals, and gene expression. Under the Songbird Neurogenomics Initiative, investigators from 11 laboratories collected brain samples from six species of songbird under a range of experimental conditions, and 488 of these samples were analyzed systematically for gene expression by microarray. ANOVA was used to test 32 planned contrasts in the data, revealing the relative impact of different factors. The brain region from which tissue was taken had the greatest influence on gene expression profile, affecting the majority of signals measured by 18,848 cDNA spots on the microarray. Social and environmental manipulations had a highly variable impact, interpreted here as a manifestation of paradoxical "constitutive plasticity" (fewer inducible genes) during periods of enhanced behavioral responsiveness. Several specific genes were identified that may be important in the evolution of linkages between environmental signals and behavior. The data were also analyzed using weighted gene coexpression network analysis, followed by gene ontology analysis. This revealed modules of coexpressed genes that are also enriched for specific functional annotations, such as "ribosome" (expressed more highly in juvenile brain) and "dopamine metabolic process" (expressed more highly in striatal song control nucleus area X). These results underscore the complexity of influences on neural gene expression and provide a resource for studying how these influences are integrated during natural experience.

Porcine tissue-specific regulatory networks derived from meta-analysis of the transcriptome

The processes that drive tissue identity and differentiation remain unclear for most tissue types. So are the gene networks and transcription factors (TF) responsible for the differential structure and function of each particular tissue, and this is particularly true for non model species with incomplete genomic resources. To better understand the regulation of genes responsible for tissue identity in pigs, we have inferred regulatory networks from a meta-analysis of 20 gene expression studies spanning 480 Porcine Affymetrix chips for 134 experimental conditions on 27 distinct tissues. We developed a mixed-model normalization approach with a covariance structure that accommodated the disparity in the origin of the individual studies, and obtained the normalized expression of 12,320 genes across the 27 tissues. Using this resource, we constructed a network, based on the co-expression patterns of 1,072 TF and 1,232 tissue specific genes. The resulting network is consistent with the known biology of tissue development. Within the network, genes clustered by tissue and tissues clustered by site of embryonic origin. These clusters were significantly enriched for genes annotated in key relevant biological processes and confirm gene functions and interactions from the literature. We implemented a Regulatory Impact Factor (RIF) metric to identify the key regulators in skeletal muscle and tissues from the central nervous systems. The normalization of the meta-analysis, the inference of the gene co-expression network and the RIF metric, operated synergistically towards a successful search for tissue-specific regulators. Novel among these findings are evidence suggesting a novel key role of ERCC3 as a muscle regulator. Together, our results recapitulate the known biology behind tissue specificity and provide new valuable insights in a less studied but valuable model species.

ENU mutagenesis in mice identifies candidate genes for hypogonadism

Genome-wide mutagenesis was performed in mice to identify candidate genes for male infertility, for which the predominant causes remain idiopathic. Mice were mutagenized using N-ethyl-N-nitrosourea (ENU), bred, and screened for phenotypes associated with the male urogenital system. Fifteen heritable lines were isolated and chromosomal loci were assigned using low-density genome-wide SNP arrays. Ten of the 15 lines were pursued further using higher-resolution SNP analysis to narrow the candidate gene regions. Exon sequencing of candidate genes identified mutations in mice with cystic kidneys (Bicc1), cryptorchidism (Rxfp2), restricted germ cell deficiency (Plk4), and severe germ cell deficiency (Prdm9). In two other lines with severe hypogonadism, candidate sequencing failed to identify mutations, suggesting defects in genes with previously undocumented roles in gonadal function. These genomic intervals were sequenced in their entirety and a candidate mutation was identified in SnrpE in one of the two lines. The line harboring the SnrpE variant retains substantial spermatogenesis despite small testis size, an unusual phenotype. In addition to the reproductive defects, heritable phenotypes were observed in mice with ataxia (Myo5a), tremors (Pmp22), growth retardation (unknown gene), and hydrocephalus (unknown gene). These results demonstrate that the ENU screen is an effective tool for identifying potential causes of male infertility.

Inferring high-confidence human protein-protein interactions

Background

As numerous experimental factors drive the acquisition, identification, and interpretation of protein-protein interactions (PPIs), aggregated assemblies of human PPI data invariably contain experiment-dependent noise. Ascertaining the reliability of PPIs collected from these diverse studies and scoring them to infer high-confidence networks is a non-trivial task. Moreover, a large number of PPIs share the same number of reported occurrences, making it impossible to distinguish the reliability of these PPIs and rank-order them. For example, for the data analyzed here, we found that the majority (>83%) of currently available human PPIs have been reported only once.

Results

In this work, we proposed an unsupervised statistical approach to score a set of diverse, experimentally identified PPIs from nine primary databases to create subsets of high-confidence human PPI networks. We evaluated this ranking method by comparing it with other methods and assessing their ability to retrieve protein associations from a number of diverse and independent reference sets. These reference sets contain known biological data that are either directly or indirectly linked to interactions between proteins. We quantified the average effect of using ranked protein interaction data to retrieve this information and showed that, when compared to randomly ranked interaction data sets, the proposed method created a larger enrichment (~134%) than either ranking based on the hypergeometric test (~109%) or occurrence ranking (~46%).

Conclusions

From our evaluations, it was clear that ranked interactions were always of value because higher-ranked PPIs had a higher likelihood of retrieving high-confidence experimental data. Reducing the noise inherent in aggregated experimental PPIs via our ranking scheme further increased the accuracy and enrichment of PPIs derived from a number of biologically relevant data sets. These results suggest that using our high-confidence protein interactions at different levels of confidence will help clarify the topological and biological properties associated with human protein networks.

The Brisbane Systems Genetics Study: genetical genomics meets complex trait genetics

There is growing evidence that genetic risk factors for common disease are caused by hereditary changes of gene regulation acting in complex pathways. Clearly understanding the molecular genetic relationships between genetic control of gene expression and its effect on complex diseases is essential. Here we describe the Brisbane Systems Genetics Study (BSGS), a family-based study that will be used to elucidate the genetic factors affecting gene expression and the role of gene regulation in mediating endophenotypes and complex diseases.

BSGS comprises of a total of 962 individuals from 314 families, for which we have high-density genotype, gene expression and phenotypic data. Families consist of combinations of both monozygotic and dizygotic twin pairs, their siblings, and, for 72 families, both parents. A significant advantage of the inclusion of parents is improved power to disentangle environmental, additive genetic and non-additive genetic effects of gene expression and measured phenotypes. Furthermore, it allows for the estimation of parent-of-origin effects, something that has not previously been systematically investigated in human genetical genomics studies. Measured phenotypes available within the BSGS include blood phenotypes and biochemical traits measured from components of the tissue sample in which transcription levels are determined, providing an ideal test case for systems genetics approaches.

We report results from an expression quantitative trait loci (eQTL) analysis using 862 individuals from BSGS to test for associations between expression levels of 17,926 probes and 528,509 SNP genotypes. At a study wide significance level approximately 15,000 associations were observed between expression levels and SNP genotypes. These associations corresponded to a total of 2,081 expression quantitative trait loci (eQTL) involving 1,503 probes. The majority of identified eQTL (87%) were located within cis-regions.

General trends in the utilization of structural factors contributing to biological complexity

During evolution, proteins containing newly emerged domains and the increasing proportion of multidomain proteins in the full genome-encoded proteome (GEP) have substantially contributed to increasing biological complexity. However, it is not known how these two potential structural factors are preferentially utilized at given physiological states. Here, we classified proteins according to domain number and domain age and explored the general trends across species for the utilization of proteins from GEP to various certain-state proteomes (CSPs, i.e., all the proteins expressed at certain physiological states). We found that multidomain proteins or only older domain-containing proteins are significantly overrepresented in CSPs compared with GEP, which is a trend that is stronger in multicellular organisms than in unicellular organisms. Interestingly, the strengths of overrepresentation decreased during evolution of multicellular eukaryotes. When comparing across CSPs, we found that multidomain proteins are more overrepresented in complex tissues than in simpler ones, whereas no difference among proteins with domains of different ages is evident between complex and simple tissues. Thus, biological complexity under certain conditions is more significantly realized by diverse domain organization than by the emergence of new types of domain. In addition, we found that multidomain or only older domain-containing proteins tend to evolve slowly and generally are under stronger purifying selection, which may partly result from their general overrepresentation trends in CSPs.

A functional assay for microRNA target identification and validation

MicroRNAs (miRNA) are a class of small RNA molecules that regulate numerous critical cellular processes and bind to partially complementary sequences resulting in down-regulation of their target genes. Due to the incomplete homology of the miRNA to its target site identification of miRNA target genes is difficult and currently based on computational algorithms predicting large numbers of potential targets for a given miRNA. To enable the identification of biologically relevant miRNA targets, we describe a novel functional assay based on a 3′-UTR-enriched library and a positive/negative selection strategy. As proof of principle we have used mir-130a and its validated target MAFB to test this strategy. Identification of MAFB and five additional targets and their subsequent confirmation as mir-130a targets by western blot analysis and knockdown experiments validates this strategy for the functional identification of miRNA targets.

A COMT gene haplotype associated with methamphetamine abuse

INTRODUCTION

Methamphetamine (MAMP) use is highly associated with psychiatric disorders with 12-13% of MAMP-dependent patients experiencing psychotic symptoms. Substance abuse and dependence may primarily involve the mesolimbic pathway and dopaminergic brain structures. It follows that dopaminergic genes, particularly COMT (encoding catechol-O-methyltransferase) and its val158met polymorphism (rs4680), are natural candidates for susceptibility loci to addiction. We have previously found an association with rs4680 and MAMP addiction.

METHODS

We present additional genotyping of rs165599 in 423 cases and 502 controls of a Taiwanese MAMP user sample. We carried out an in-silico evaluation of rs165599 for a possible impact on microRNA binding or UTR stability. We also carried out a review of transcript sequences across the COMT 3'UTR.

RESULTS

Genotype counts were (cases/controls): AA 94/110, AG 198/210 and GG 93/109. There were no significant allele or genotype differences between cases and controls for rs165599. However, a haplotype main effect was detected using both rs4680 and rs165599 using the χ²-test in UNPHASED. The global P-value was P=0.0044 with the effect appearing to derive from one haplotype that is underrepresented in cases: A/G for rs4680/rs165599 (haplotype P=0.001). rs165599 is a single nucleotide polymorphism located in the COMT 3' untranslated region (UTR), a noncoding transcript region subject to posttranscriptional down-regulation by mechanisms such as microRNA binding. A review of transcript sequences across the COMT 3'UTR found evidence to suggest antisense interference of COMT from the 3'UTR of the neighbouring 'Armadillo repeat gene deleted in velocardiofacial syndrome' gene.

A general method to derive robust organ-specific gene expression-based differentiation indices: application to thyroid cancer diagnostic

Differentiation is central to development, while dedifferentiation is central to cancer progression. Hence, a quantitative assessment of differentiation would be most useful. We propose an unbiased method to derive organ-specific differentiation indices from gene expression data and demonstrate its usefulness in thyroid cancer diagnosis. We derived a list of thyroid-specific genes by selecting automatically those genes that are expressed at higher level in the thyroid than in any other organ in a normal tissue's genome-wide gene expression compendium. The thyroid index of a tissue was defined as the median expression of these thyroid-specific genes in that tissue. As expected, the thyroid index was inversely correlated with meta-PCNA, a proliferation metagene, across a wide range of thyroid tumors. By contrast, the two indices were positively correlated in a time course of thyroid-stimulating hormone (TSH) activation of primary thyrocytes. Thus, the thyroid index captures biological information not integrated by proliferation rates. The differential diagnostic of follicular thyroid adenomas and follicular thyroid carcinoma is a notorious challenge for pathologists. The thyroid index discriminated them as accurately as did machine-learning classifiers trained on the genome-wide cancer data. Hence, although it was established exclusively from normal tissue data, the thyroid index integrates the relevant diagnostic information contained in tumoral transcriptomes. Similar results were obtained for the classification of the follicular vs classical variants of papillary thyroid cancers, that is, tumors dedifferentiating along a different route. The automated procedures demonstrated in the thyroid are applicable to other organs.

RNA-Seq analyses generate comprehensive transcriptomic landscape and reveal complex transcript patterns in hepatocellular carcinoma

RNA-seq is a powerful tool for comprehensive characterization of whole transcriptome at both gene and exon levels and with a unique ability of identifying novel splicing variants. To date, RNA-seq analysis of HBV-related hepatocellular carcinoma (HCC) has not been reported. In this study, we performed transcriptome analyses for 10 matched pairs of cancer and non-cancerous tissues from HCC patients on Solexa/Illumina GAII platform. On average, about 21.6 million sequencing reads and 10.6 million aligned reads were obtained for samples sequenced on each lane, which was able to identify >50% of all the annotated genes for each sample. Furthermore, we identified 1,378 significantly differently expressed genes (DEGs) and 24, 338 differentially expressed exons (DEEs). Comprehensive function analyses indicated that cell growth-related, metabolism-related and immune-related pathways were most significantly enriched by DEGs, pointing to a complex mechanism for HCC carcinogenesis. Positional gene enrichment analysis showed that DEGs were most significantly enriched at chromosome 8q21.3–24.3. The most interesting findings were from the analysis at exon levels where we characterized three major patterns of expression changes between gene and exon levels, implying a much complex landscape of transcript-specific differential expressions in HCC. Finally, we identified a novel highly up-regulated exon-exon junction in ATAD2 gene in HCC tissues. Overall, to our best knowledge, our study represents the most comprehensive characterization of HBV-related HCC transcriptome including exon level expression changes and novel splicing variants, which illustrated the power of RNA-seq and provided important clues for understanding the molecular mechanisms of HCC pathogenesis at system-wide levels.

In-depth quantitative cardiac proteomics combining electron transfer dissociation and the metalloendopeptidase Lys-N with the SILAC mouse

In quantitative proteomics stable isotope labeling has progressed from cultured cells toward the total incorporation of labeled atoms or amino acids into whole multicellular organisms. For instance, the recently introduced (13)C(6)-lysine labeled SILAC mouse allows accurate comparison of protein expression directly in tissue. In this model, only lysine, but not arginine, residues are isotope labeled, as the latter may cause complications to the quantification by in vivo conversion of arginine to proline. The sole labeling of lysines discourages the use of trypsin, as not all peptides will be quantifiable. Therefore, in the initial work Lys-C was used for digestion. Here, we demonstrate that the lysine-directed protease metalloendopeptidase Lys-N is an excellent alternative. As lysine directed peptides generally yield longer and higher charged peptides, alongside the more traditional collision induced dissociation we also implemented electron transfer dissociation in a quantitative stable isotope labeling with amino acid in cell culture workflow for the first time. The utility of these two complementary approaches is highlighted by investigating the differences in protein expression between the left and right ventricle of a mouse heart. Using Lys-N and electron transfer dissociation yielded coverage to a depth of 3749 proteins, which is similar as earlier investigations into the murine heart proteome. In addition, this strategy yields quantitative information on ∼ 2000 proteins with a median coverage of four peptides per protein in a single strong cation exchange-liquid chromatography-MS experiment, revealing that the left and right ventricle proteomes are very similar qualitatively as well as quantitatively.

Origins and consequences of transcriptional discontinuity

In both prokaryotes and eukaryotes, transcription has been described as being temporally discontinuous, most genes being active mainly during short activity windows interspersed by silent periods. In mammalian cells, recent studies performed at the single cell level have revealed that transcriptional kinetics are highly gene-specific and constrained by the presence of refractory periods of inactivity before a gene can be turned on again. While the underlying mechanisms generating gene-specific kinetic characteristics remain unclear, various biological consequences of transcriptional discontinuity have been unravelled during the past few years. Here we review recent advances on understanding transcriptional kinetics of individual genes at the single cell level and discuss its possible origins and consequences.