A DNA microarray survey of gene expression in normal human tissues

A tissue-specific gene expression template portrays heart development and pathology

Congenital heart defects (CHD) are the most common cause of death in children under the age of 1. Tetralogy of Fallot (TOF) is a severe CHD that results from developmental defects in the conotruncal outflow tract. Recently, a tissue-specific gene expression template (GET) was derived from microarray data that accurately characterized multiple normal human tissues. We used the GET to examine spatial, temporal, and a pathological condition (TOF) within a single organ, the heart. The GET, as previously defined, generally identified temporal and spatial differences in the cardiac tissue. Differences in the stoichiometry of the GET reflected the severe developmental disturbance associated with TOF. Our analysis suggests that the homoeostatic equilibrium assessed by the GET at the inter-organ level is generally maintained at the intra-organ level as well.

An expression signature at diagnosis to estimate prostate cancer patients' overall survival

Background:

This study aimed to identify biomarkers for estimating the overall and prostate cancer (PCa)-specific survival in PCa patients at diagnosis.

Methods:

To explore the importance of embryonic stem cell (ESC) gene signatures, we identified 641 ESC gene predictors (ESCGPs) using published microarray data sets. ESCGPs were selected in a stepwise manner, and were combined with reported genes. Selected genes were analyzed by multiplex quantitative polymerase chain reaction using prostate fine-needle aspiration samples taken at diagnosis from a Swedish cohort of 189 PCa patients diagnosed between 1986 and 2001. Of these patients, there was overall and PCa-specific survival data available for 97.9%, and 77.9% were primarily treated by hormone therapy only. Univariate and multivariate Cox proportional hazard ratios and Kaplan–Meier plots were used for the survival analysis, and a k-nearest neighbor (kNN) algorithm for estimating overall survival.

Results:

An expression signature of VGLL3, IGFBP3 and F3 was shown sufficient to categorize the patients into high-, intermediate- and low-risk subtypes. The median overall survival times of the subtypes were 3.23, 4.00 and 9.85 years, respectively. The difference corresponded to hazard ratios of 5.86 (95% confidence interval (CI): 2.91–11.78, P<0.001) for the high-risk subtype and 3.45 (95% CI: 1.79–6.66, P<0.001) for the intermediate-risk compared with the low-risk subtype. The kNN models that included the gene expression signature outperformed the one designed on clinical parameters alone.

Conclusions:

The expression signature can potentially be used to estimate overall survival time. When validated in future studies, it could be integrated in the routine clinical diagnostic and prognostic procedure of PCa for an optimal treatment decision based on the estimated survival benefit.

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.

Body maps on the human genome

BACKGROUND

Chromosomes have territories, or preferred locales, in the cell nucleus. When these sites are taken into account, some large-scale structure of the human genome emerges.

RESULTS

The synoptic picture is that genes highly expressed in particular topologically compact tissues are not randomly distributed on the genome. Rather, such tissue-specific genes tend to map somatotopically onto the complete chromosome set. They seem to form a "genome homunculus": a multi-dimensional, genome-wide body representation extending across chromosome territories of the entire spermcell nucleus. The antero-posterior axis of the body significantly corresponds to the head-tail axis of the nucleus, and the dorso-ventral body axis to the central-peripheral nucleus axis.

CONCLUSIONS

This large-scale genomic structure includes thousands of genes. One rationale for a homuncular genome structure would be to minimize connection costs in genetic networks. Somatotopic maps in cerebral cortex have been reported for over a century.

Transcriptome profiling identifies differentially expressed genes in postnatal developing pituitary gland of miniature pig

In recent years, Tibetan pig and Bama pig are popularly used as animal models for medical researches. However, little genomic information is available for the two breeds, particularly regarding gene expression pattern at the whole-transcriptome level. In this study, we characterized the pituitary transcriptome profile along their postnatal developmental stages within and between the two breeds in order to illustrate the differential dynamics and functions of differentially expressed genes. We obtained a total of ∼300 million 80-bp paired-end reads, detected 15 715 previously annotated genes. Most of the genes (90.33%) were shared between the two breeds with the main functions in metabolic process. Four hormone genes (GH, PRL, LHB, and FSHB) were detected in all samples with extremely high levels of expression. Functional differences between the three developmental stages (infancy, puberty and adulthood) in each breed were dominantly presented by the gene expressions at the first stage. That is, Bama pig was over-represented in the genes involved in the cellular process, while Tibetan pig was over-represented in the genes represented by the reproductive process. The identified SNPs indicated that the divergence between the miniature pig breeds and the large pig (Duroc) were greater than that between the two miniature pig breeds. This study substantially expands our knowledge concerning the genes transcribed in the pig pituitary gland and provides an overview of pituitary transcriptome dynamics throughout the period of postnatal development.

Characterization of regulatory features of housekeeping and tissue-specific regulators within tissue regulatory networks

Background

Transcription factors (TFs) and miRNAs are essential for the regulation of gene expression; however, the global view of human gene regulatory networks remains poorly understood. For example, how is the expression of so many genes regulated by limited cohorts of regulators and how are genes differentially expressed in different tissues despite the genetic code being the same in all tissues?

Results

We analyzed the network properties of housekeeping and tissue-specific genes in gene regulatory networks from seven human tissues. Our results show that different classes of genes behave quite differently in these networks. Tissue-specific miRNAs show a higher average target number compared with non-tissue specific miRNAs, which indicates that tissue-specific miRNAs tend to regulate different sets of targets. Tissue-specific TFs exhibit higher in-degree, out-degree, cluster coefficient and betweenness values, indicating that they occupy central positions in the regulatory network and that they transfer genetic information from upstream genes to downstream genes more quickly than other TFs. Housekeeping TFs tend to have higher cluster coefficients compared with other genes that are neither housekeeping nor tissue specific, indicating that housekeeping TFs tend to regulate their targets synergistically. Several topological properties of disease-associated miRNAs and genes were found to be significantly different from those of non-disease-associated miRNAs and genes.

Conclusions

Tissue-specific miRNAs, TFs and disease genes have particular topological properties within the transcriptional regulatory networks of the seven human tissues examined. The tendency of tissue-specific miRNAs to regulate different sets of genes shows that a particular tissue-specific miRNA and its target gene set may form a regulatory module to execute particular functions in the process of tissue differentiation. The regulatory patterns of tissue-specific TFs reflect their vital role in regulatory networks and their importance to biological functions in their respective tissues. The topological differences between disease and non-disease genes may aid the discovery of new disease genes or drug targets. Determining the network properties of these regulatory factors will help define the basic principles of human gene regulation and the molecular mechanisms of disease.

Ontology-aware classification of tissue and cell-type signals in gene expression profiles across platforms and technologies

Motivation: Leveraging gene expression data through large-scale integrative analyses for multicellular organisms is challenging because most samples are not fully annotated to their tissue/cell-type of origin. A computational method to classify samples using their entire gene expression profiles is needed. Such a method must be applicable across thousands of independent studies, hundreds of gene expression technologies and hundreds of diverse human tissues and cell-types.

Results: We present Unveiling RNA Sample Annotation (URSA) that leverages the complex tissue/cell-type relationships and simultaneously estimates the probabilities associated with hundreds of tissues/cell-types for any given gene expression profile. URSA provides accurate and intuitive probability values for expression profiles across independent studies and outperforms other methods, irrespective of data preprocessing techniques. Moreover, without re-training, URSA can be used to classify samples from diverse microarray platforms and even from next-generation sequencing technology. Finally, we provide a molecular interpretation for the tissue and cell-type models as the biological basis for URSA’s classifications.

Availability and implementation: An interactive web interface for using URSA for gene expression analysis is available at: ursa.princeton.edu. The source code is available at https://bitbucket.org/youngl/ursa_backend.

Contact:ogt@cs.princeton.edu

Supplementary information:Supplementary data are available at Bioinformatics online.

Yin Yang gene expression ratio signature for lung cancer prognosis

Many studies have established gene expression-based prognostic signatures for lung cancer. All of these signatures were built from training data sets by learning the correlation of gene expression with the patients' survival time. They require all new sample data to be normalized to the training data, ultimately resulting in common problems of low reproducibility and impracticality. To overcome these problems, we propose a new signature model which does not involve data training. We hypothesize that the imbalance of two opposing effects in lung cancer cells, represented by Yin and Yang genes, determines a patient's prognosis. We selected the Yin and Yang genes by comparing expression data from normal lung and lung cancer tissue samples using both unsupervised clustering and pathways analyses. We calculated the Yin and Yang gene expression mean ratio (YMR) as patient risk scores. Thirty-one Yin and thirty-two Yang genes were identified and selected for the signature development. In normal lung tissues, the YMR is less than 1.0; in lung cancer cases, the YMR is greater than 1.0. The YMR was tested for lung cancer prognosis prediction in four independent data sets and it significantly stratified patients into high- and low-risk survival groups (p = 0.02, HR = 2.72; p = 0.01, HR = 2.70; p = 0.007, HR = 2.73; p = 0.005, HR = 2.63). It also showed prediction of the chemotherapy outcomes for stage II & III. In multivariate analysis, the YMR risk factor was more successful at predicting clinical outcomes than other commonly used clinical factors, with the exception of tumor stage. The YMR can be measured in an individual patient in the clinic independent of gene expression platform. This study provided a novel insight into the biology of lung cancer and shed light on the clinical applicability.

The architecture of the gene regulatory networks of different tissues

Summary: The great variety of human cell types in morphology and function is due to the diverse gene expression profiles that are governed by the distinctive regulatory networks in different cell types. It is still a challenging task to explain how the regulatory networks achieve the diversity of different cell types. Here, we report on our studies of the design principles of the tissue regulatory system by constructing the regulatory networks of eight human tissues, which subsume the regulatory interactions between transcription factors (TFs), microRNAs (miRNAs) and non-TF target genes. The results show that there are in-/out-hubs of high in-/out-degrees in tissue networks. Some hubs (strong hubs) maintain the hub status in all the tissues where they are expressed, whereas others (weak hubs), in spite of their ubiquitous expression, are hubs only in some tissues. The network motifs are mostly feed-forward loops. Some of them having no miRNAs are the common motifs shared by all tissues, whereas the others containing miRNAs are the tissue-specific ones owned by one or several tissues, indicating that the transcriptional regulation is more conserved across tissues than the post-transcriptional regulation. In particular, a common bow-tie framework was found that underlies the motif instances and shows diverse patterns in different tissues. Such bow-tie framework reflects the utilization efficiency of the regulatory system as well as its high variability in different tissues, and could serve as the model to further understand the structural adaptation of the regulatory system to the specific requirements of different cell functions.

Contact:edgar.wingender@bioinf.med.uni-goettingen.de; jwang@nju.edu.cn

Supplementary information:Supplementary data are available at Bioinformatics online.

Analysis of in vivo oocyte maturation, in vitro embryo development and gene expression in cumulus cells of dairy cows and heifers selected for one fertility quantitative trait loci (QTL) located on BTA3

We have previously shown that Holstein cows selected for their homozygous favorable ("fertil+") or unfavorable ("fertil-") haplotype at one quantitative trait loci (QTL) of female fertility located on chromosome 3 (QTL-F-Fert-BTA3) had a different success rate 35 and 90 days after the first artificial insemination. To determine whether the lower fertility in "fertil-" animals could be related to oocyte quality, we analyzed the embryo development rate in vitro and the oocyte meiotic maturation in vivo in "fertil+" and "fertil-" heifers. In vitro maturation and fertilization of immature oocytes recovered by ovum pick-up from "fertil+" and "fertil-" heifers resulted in similar cleavage and blastocyst rates in the two haplotypes. However the percentage of expanded blastocysts and the number of cells per blastocyst were significantly higher in "fertil+". Oocytes from presumptive preovulatory follicles were analyzed after ovarian stimulation. A similar rate of immature (from prophase to metaphase-I) and mature oocytes (metaphase-II) was obtained in the two haplotypes, whereas a significantly higher percentage of oocytes from metaphase-I to metaphase-II was observed in "fertil+" compared to "fertil-" heifers. Since cumulus cells (CCs) could reflect the developmental competence of oocytes, we analyzed the expression of seven genes included in the QTL-F-Fert-BTA3 using real-time PCR in bovine CCs after in vivo or in vitro maturation, as a model of higher and lower competence, respectively. Transcript levels of TAGLN2, EEF1A1 and PIGM were higher in CCs after in vitro maturation (IVM) compared to in vivo maturation, whereas no difference was observed for IFI16, KIRREL, SPTA1 and PEX19 expression. The expression levels of all these genes in in preovulatory CCs were not significantly different between "fertil+" and "fertil-" heifers. In conclusion, the lower fertility of "fertil-" females could be partially due to a lowest quality of the oocytes and consequently of preimplantation embryo development.

Toward a systems level view of the ECM and related proteins: a framework for the systematic definition and analysis of biological systems

Advances in high throughput 'omic technologies are starting to provide unprecedented insights into how components of biological systems are organized and interact. Key to exploiting these datasets is the definition of the components that comprise the system of interest. Although a variety of knowledge bases exist that capture such information, a major challenge is determining how these resources may be best utilized. Here we present a systematic curation strategy to define a systems-level view of the human extracellular matrix (ECM)--a three-dimensional meshwork of proteins and polysaccharides that impart structure and mechanical stability to tissues. Employing our curation strategy we define a set of 357 proteins that represent core components of the ECM, together with an additional 524 genes that mediate related functional roles, and construct a map of their physical interactions. Topological properties help identify modules of functionally related proteins, including those involved in cell adhesion, bone formation and blood clotting. Because of its major role in cell adhesion, proliferation and morphogenesis, defects in the ECM have been implicated in cancer, atherosclerosis, asthma, fibrosis, and arthritis. We use MeSH annotations to identify modules enriched for specific disease terms that aid to strengthen existing as well as predict novel gene-disease associations. Mapping expression and conservation data onto the network reveal modules evolved in parallel to convey tissue-specific functionality on otherwise broadly expressed units. In addition to demonstrating an effective workflow for defining biological systems, this study crystallizes our current knowledge surrounding the organization of the ECM.

Development of a Pacific oyster (Crassostrea gigas) 31,918-feature microarray: identification of reference genes and tissue-enriched expression patterns

BACKGROUND

Research using the Pacific oyster Crassostrea gigas as a model organism has experienced rapid growth in recent years due to the development of high-throughput molecular technologies. As many as 56,268 EST sequences have been sequenced to date, representing a genome-wide resource that can be used for transcriptomic investigations.

RESULTS

In this paper, we developed a Pacific oyster microarray containing oligonucleotides representing 31,918 transcribed sequences selected from the publicly accessible GigasDatabase. This newly designed microarray was used to study the transcriptome of male and female gonads, mantle, gills, posterior adductor muscle, visceral ganglia, hemocytes, labial palps and digestive gland. Statistical analyses identified genes differentially expressed among tissues and clusters of tissue-enriched genes. These genes reflect major tissue-specific functions at the molecular level, such as tissue formation in the mantle, filtering in the gills and labial palps, and reproduction in the gonads. Hierarchical clustering predicted the involvement of unannotated genes in specific functional pathways such as the insulin/NPY pathway, an important pathway under study in our model species. Microarray data also accurately identified reference genes whose mRNA level appeared stable across all the analyzed tissues. Adp-ribosylation factor 1 (arf1) appeared to be the most robust reference for normalizing gene expression data across different tissues and is therefore proposed as a relevant reference gene for further gene expression analysis in the Pacific oyster.

CONCLUSIONS

This study provides a new transcriptomic tool for studies of oyster biology, which will help in the annotation of its genome and which identifies candidate reference genes for gene expression analysis.

NRF2 deficiency reduces life span of mice administered thoracic irradiation

Subsets of cancer survivors who have been subjected to thoracic irradiation face the prospect of developing pulmonary injury. Radiation-induced pulmonary fibrosis is an insidious injury that presents 6 to 24 months after irradiation and continues to progress over a period of years. TGF-β and reactive oxygen species contribute significantly to the pathogenesis of this injury. The transcription factor NRF2 controls antioxidant gene expression and therefore regulates the cellular oxidant burden. This work demonstrates an additional paradigm for NRF2: suppression of TGF-β-mediated signaling, assessed by measuring expression of a surrogate TGF-β1 target gene (PAI-1) in lung fibroblasts. Thoracic irradiation of Nfe2l2(-/-) mice resulted in rapid expression of PAI-1 and FSP-1 compared to irradiated wild-type mice. Examination of lung tissue 16 weeks after thoracic irradiation of Nfe2l2(-/-) mice revealed the presence of distended alveoli and decreased numbers of alveoli compared to wild-type mice. Suppression of NRF2 expression shortened life span in mice administered 16 Gy to the thorax. Nfe2l2(+/-) and Nfe2l2(-/-) mice exhibited a mean life span of 176 days compared to wild-type mice, which lived an average of 212 days. These novel results identify NRF2 as a susceptibility factor for the development of late tissue injury.

Cancer/Testis Antigens as potential predictors of biochemical recurrence of prostate cancer following radical prostatectomy

Background

The Cancer/Testis Antigens (CTAs) are an important group of proteins that are typically restricted to the testis in the normal adult but are aberrantly expressed in several types of cancers. As a result of their restricted expression patterns, the CTAs could serve as unique biomarkers for cancer diagnosis/prognosis. The aim of this study was to identify promising CTAs that are associated with prostate cancer (PCa) recurrence following radical prostatectomy (RP).

Methods

The expression of 5 CTAs was measured by quantitative multiplex real-time PCR using prostate tissue samples obtained from 72 patients with apparently clinically localized PCa with a median of two years follow-up (range, 1 to 14 years).

Results

The expression of CTAs namely, CEP55, NUF2, PBK and TTK were significantly higher while PAGE4 was significantly lower in patients with recurrent disease. All CTAs with the exception of TTK were significantly correlated with the prostatectomy Gleason score, but none were correlated with age, stage, or preoperative PSA levels. In univariate proportional hazards models, CEP55 (HR = 3.59, 95% CI: 1.50-8.60), p = 0.004; NUF2 (HR = 2.28, 95% CI: 1.11-4.67), p = 0.024; and PAGE4 (HR = 0.44, 95% CI: 0.21-0.93), p = 0.031 were significantly associated with the risk of PCa recurrence. However, the results were no longer significant after adjustment for prostatectomy Gleason score.

Conclusions

To our knowledge, this is the first study to identify CTAs as biomarkers that can differentiate patients with recurrent and non-recurrent disease following RP and underscores its potential impact on PCa prognosis and treatment.

Tissue-specific gene expression templates for accurate molecular characterization of the normal physiological states of multiple human tissues with implication in development and cancer studies

BACKGROUND

To elucidate the molecular complications in many complex diseases, we argue for the priority to construct a model representing the normal physiological state of a cell/tissue.

RESULTS

By analyzing three independent microarray datasets on normal human tissues, we established a quantitative molecular model GET, which consists of 24 tissue-specific Gene Expression Templates constructed from a set of 56 genes, for predicting 24 distinct tissue types under disease-free condition. 99.2% correctness was reached when a large-scale validation was performed on 61 new datasets to test the tissue-prediction power of GET. Network analysis based on molecular interactions suggests a potential role of these 56 genes in tissue differentiation and carcinogenesis.Applying GET to transcriptomic datasets produced from tissue development studies the results correlated well with developmental stages. Cancerous tissues and cell lines yielded significantly lower correlation with GET than the normal tissues. GET distinguished melanoma from normal skin tissue or benign skin tumor with 96% sensitivity and 89% specificity.

CONCLUSIONS

These results strongly suggest that a normal tissue or cell may uphold its normal functioning and morphology by maintaining specific chemical stoichiometry among genes. The state of stoichiometry can be depicted by a compact set of representative genes such as the 56 genes obtained here. A significant deviation from normal stoichiometry may result in malfunction or abnormal growth of the cells.

Systematic spatial bias in DNA microarray hybridization is caused by probe spot position-dependent variability in lateral diffusion

BACKGROUND

The hybridization of nucleic acid targets with surface-immobilized probes is a widely used assay for the parallel detection of multiple targets in medical and biological research. Despite its widespread application, DNA microarray technology still suffers from several biases and lack of reproducibility, stemming in part from an incomplete understanding of the processes governing surface hybridization. In particular, non-random spatial variations within individual microarray hybridizations are often observed, but the mechanisms underpinning this positional bias remain incompletely explained.

METHODOLOGY/PRINCIPAL FINDINGS

This study identifies and rationalizes a systematic spatial bias in the intensity of surface hybridization, characterized by markedly increased signal intensity of spots located at the boundaries of the spotted areas of the microarray slide. Combining observations from a simplified single-probe block array format with predictions from a mathematical model, the mechanism responsible for this bias is found to be a position-dependent variation in lateral diffusion of target molecules. Numerical simulations reveal a strong influence of microarray well geometry on the spatial bias.

CONCLUSIONS

Reciprocal adjustment of the size of the microarray hybridization chamber to the area of surface-bound probes is a simple and effective measure to minimize or eliminate the diffusion-based bias, resulting in increased uniformity and accuracy of quantitative DNA microarray hybridization.

Expression cartography of human tissues using self organizing maps

Background

Parallel high-throughput microarray and sequencing experiments produce vast quantities of multidimensional data which must be arranged and analyzed in a concerted way. One approach to addressing this challenge is the machine learning technique known as self organizing maps (SOMs). SOMs enable a parallel sample- and gene-centered view of genomic data combined with strong visualization and second-level analysis capabilities. The paper aims at bridging the gap between the potency of SOM-machine learning to reduce dimension of high-dimensional data on one hand and practical applications with special emphasis on gene expression analysis on the other hand.

Results

The method was applied to generate a SOM characterizing the whole genome expression profiles of 67 healthy human tissues selected from ten tissue categories (adipose, endocrine, homeostasis, digestion, exocrine, epithelium, sexual reproduction, muscle, immune system and nervous tissues). SOM mapping reduces the dimension of expression data from ten of thousands of genes to a few thousand metagenes, each representing a minicluster of co-regulated single genes. Tissue-specific and common properties shared between groups of tissues emerge as a handful of localized spots in the tissue maps collecting groups of co-regulated and co-expressed metagenes. The functional context of the spots was discovered using overrepresentation analysis with respect to pre-defined gene sets of known functional impact. We found that tissue related spots typically contain enriched populations of genes related to specific molecular processes in the respective tissue. Analysis techniques normally used at the gene-level such as two-way hierarchical clustering are better represented and provide better signal-to-noise ratios if applied to the metagenes. Metagene-based clustering analyses aggregate the tissues broadly into three clusters containing nervous, immune system and the remaining tissues.

Conclusions

The SOM technique provides a more intuitive and informative global view of the behavior of a few well-defined modules of correlated and differentially expressed genes than the separate discovery of the expression levels of hundreds or thousands of individual genes. The program is available as R-package 'oposSOM'.

A compendium of canine normal tissue gene expression

BACKGROUND

Our understanding of disease is increasingly informed by changes in gene expression between normal and abnormal tissues. The release of the canine genome sequence in 2005 provided an opportunity to better understand human health and disease using the dog as clinically relevant model. Accordingly, we now present the first genome-wide, canine normal tissue gene expression compendium with corresponding human cross-species analysis.

METHODOLOGY/PRINCIPAL FINDINGS

The Affymetrix platform was utilized to catalogue gene expression signatures of 10 normal canine tissues including: liver, kidney, heart, lung, cerebrum, lymph node, spleen, jejunum, pancreas and skeletal muscle. The quality of the database was assessed in several ways. Organ defining gene sets were identified for each tissue and functional enrichment analysis revealed themes consistent with known physio-anatomic functions for each organ. In addition, a comparison of orthologous gene expression between matched canine and human normal tissues uncovered remarkable similarity. To demonstrate the utility of this dataset, novel canine gene annotations were established based on comparative analysis of dog and human tissue selective gene expression and manual curation of canine probeset mapping. Public access, using infrastructure identical to that currently in use for human normal tissues, has been established and allows for additional comparisons across species.

CONCLUSIONS/SIGNIFICANCE

These data advance our understanding of the canine genome through a comprehensive analysis of gene expression in a diverse set of tissues, contributing to improved functional annotation that has been lacking. Importantly, it will be used to inform future studies of disease in the dog as a model for human translational research and provides a novel resource to the community at large.

Novel tumour-specific promoters for transcriptional targeting of hepatocellular carcinoma by herpes simplex virus vectors

BACKGROUND

Hepatocellular carcinoma (HCC) is a cancer of poor prognosis, with limited success in patient treatment, which it makes an excellent target for gene therapy and viral oncolysis. Accordingly, herpes virus simplex type-1 (HSV-1) is one of the most promising viral platforms for transferring therapeutic genes and the development of oncolytic vectors that can target, multiply in, and eradicate hepatoma cells via their lytic cycle. Enhanced efficacy and specificity of HSV-1-based vectors towards HCC may be achieved by using HCC-specific gene promoters to drive selective viral gene expression and accomplish conditional replication and/or to control the expression of therapeutic genes. However, careful verification of promoter function in the context of the replication-competent HSV-1 vectors is required. The present study aimed to identify novel HCC-specific promoters that could efficiently direct transgene expression to HCC cells and maintain their activity during active viral replication.

METHODS

Publicly available microarray data from human HCC biopsies were analysed in order to detect novel candidate genes induced primarily in HCC compared to normal liver. HCC specificity and promoter activity were evaluated by RT-PCR and chromatin immunoprecipitation. Additionally, transcriptional activity of promoters was further evaluated in the context of HSV-1 genome, using luciferase assays in cultured cells and animal models.

RESULTS

Eight HCC-specific genes were characterised in this study: Angiopoietin-like-3, Cytochrome P450, family 2, subfamily C, polypeptide 8, Vitronectin, Alcohol dehydrogenase 6-class V, Apolipoprotein B, Fibrinogen beta chain, Inter-alpha-globulin-inhibitor H3 and Inter-alpha-globulin-inhibitor H1. Specific HCC expression and active gene transcription were confirmed in human liver and non-liver cell lines and further evaluated in primary neoplastic cells from hepatitis C and B virus (HCV- and HBV)-associated HCC patients. High promoter activity and specificity in the presence of HSV-1 infection and from within the viral genome, was validated, both in vitro and in vivo.

CONCLUSIONS

We identified and experimentally characterized novel hepatoma-specific promoters, which were valuable for cancer-specific gene therapy, using HSV-1 vectors.

Regulator of G protein signaling 5 is highly expressed in parathyroid tumors and inhibits signaling by the calcium-sensing receptor

The molecular mechanisms responsible for aberrant calcium signaling in parathyroid disease are poorly understood. The loss of appropriate calcium-responsive modulation of PTH secretion observed in parathyroid disease is commonly attributed to decreased expression of the calcium-sensing receptor (CaSR), a G protein-coupled receptor. However, CaSR expression is highly variable in parathyroid adenomas, and the lack of correlation between CaSR abundance and calcium-responsive PTH kinetics indicates that mechanisms independent of CaSR expression may contribute to aberrant calcium sensing in parathyroid disease. To gain a better understanding of parathyroid tumors and the molecular determinants that drive parathyroid adenoma development, we performed gene expression profiling on a panel of 64 normal and neoplastic parathyroid tissues. The microarray data revealed high-level expression of genes known to be involved in parathyroid biology (PTH, VDR, CGA, CaSR, and GCM2). Moreover, our screen identified regulator of G protein signaling 5 (RGS5) as a candidate inhibitor of CaSR signaling. We confirmed RGS5 to be highly expressed in parathyroid adenomas relative to matched-pair normal glands. Transient expression of RGS5 in cells stably expressing CaSR resulted in dose-dependent abrogation of calcium-stimulated inositol trisphosphate production and ERK1/2 phosphorylation. Furthermore, we found that RGS5-nullizygous mice display reduced plasma PTH levels, an outcome consistent with attenuated opposition to CaSR activity. Collectively, these data suggest that RGS5 can act as a physiological regulator of calcium sensing by CaSR in the parathyroid gland. The abnormally elevated expression of RGS5 observed in parathyroid adenomas could thus represent a novel mechanism of CaSR desensitization in patients with primary hyperparathyroidism.

Survey of the effect of genetic variations on gene expression in human prefrontal cortex and its application to genetics of psychiatric disorders

Identifying the genetic basis of gene expression variation in the human brain is important for understanding brain physiology and pathophysiology. We investigated the genetic basis of gene expression variation in human prefrontal cortex using single nucleotide polymorphisms (SNPs) and taking into consideration brain sample pH. From approximately 12,000 brain-expressed transcripts, we identified 187 cis-regulated transcripts. Some of the transcripts were identified as cis-regulated in the lymphoblastoid cells or lymphocytes, which suggests common cis-regulation across different tissues. Knowledge of genetic variations contributing to differences in gene expression in the brain would be particularly useful in the study of neuropsychiatric disorders in combination with a large-scale genome-wide association study. Using Wellcome Trust Case Control Consortium association study data, we identified SNPs associated with bipolar disorder and gene expression variation in the human brain. We found that SNPs in the AKAP10 and PRKCI genes are significantly associated with bipolar disorder and gene expression variation.

Principal manifolds and graphs in practice: from molecular biology to dynamical systems

We present several applications of non-linear data modeling, using principal manifolds and principal graphs constructed using the metaphor of elasticity (elastic principal graph approach). These approaches are generalizations of the Kohonen's self-organizing maps, a class of artificial neural networks. On several examples we show advantages of using non-linear objects for data approximation in comparison to the linear ones. We propose four numerical criteria for comparing linear and non-linear mappings of datasets into the spaces of lower dimension. The examples are taken from comparative political science, from analysis of high-throughput data in molecular biology, from analysis of dynamical systems.

Module discovery by exhaustive search for densely connected, co-expressed regions in biomolecular interaction networks

Background

Computational prediction of functionally related groups of genes (functional modules) from large-scale data is an important issue in computational biology. Gene expression experiments and interaction networks are well studied large-scale data sources, available for many not yet exhaustively annotated organisms. It has been well established, when analyzing these two data sources jointly, modules are often reflected by highly interconnected (dense) regions in the interaction networks whose participating genes are co-expressed. However, the tractability of the problem had remained unclear and methods by which to exhaustively search for such constellations had not been presented.

Methodology/Principal Findings

We provide an algorithmic framework, referred to as Densely Connected Biclustering (DECOB), by which the aforementioned search problem becomes tractable. To benchmark the predictive power inherent to the approach, we computed all co-expressed, dense regions in physical protein and genetic interaction networks from human and yeast. An automatized filtering procedure reduces our output which results in smaller collections of modules, comparable to state-of-the-art approaches. Our results performed favorably in a fair benchmarking competition which adheres to standard criteria. We demonstrate the usefulness of an exhaustive module search, by using the unreduced output to more quickly perform GO term related function prediction tasks. We point out the advantages of our exhaustive output by predicting functional relationships using two examples.

Conclusion/Significance

We demonstrate that the computation of all densely connected and co-expressed regions in interaction networks is an approach to module discovery of considerable value. Beyond confirming the well settled hypothesis that such co-expressed, densely connected interaction network regions reflect functional modules, we open up novel computational ways to comprehensively analyze the modular organization of an organism based on prevalent and largely available large-scale datasets.

Availability

Software and data sets are available at http://www.sfu.ca/~ester/software/DECOB.zip.

A panel of cancer-testis genes exhibiting broad-spectrum expression in haematological malignancies

Cancer-testis (CT) antigens/genes show restricted expression in normal tissues but widespread expression in many tumour types. This, coupled with their ability to induce cytotoxic T-lymphocyte responses, makes them attractive vaccine candidates. Following our identification of PASD1, we have used RT-PCR to analyse the mRNA expression profile of a large panel of CT genes in cell lines derived from haematological malignancies, and have studied Sp17 protein expression in the same cell lines and diffuse large B-cell lymphoma (DLBCL) biopsies. Our extensive analysis revealed multiple CT transcripts exhibiting widespread expression across cell lines derived from 21 B- and 4 T-cell malignancies. The broadest mRNA expression profiles were observed for the following eight CT genes: Sp17 (25/25), PRAME (25/25), CSAGE (24/25), PASD1 (22/25), CAGE/DDX53 (19/25), CTAGE1 (19/25), HAGE/DDX43 (16/25) and PLU-1/JARID1B (15/25). Cell lines derived from more aggressive lymphoma subtypes generally expressed CT transcripts at higher frequency. Sp17 protein was detected in a number of cell lines and in six of eleven (54.5%) DLBCL biopsies. Analysis of Sp17 protein expression, by immunohistochemistry and Western blotting, broadens the scope of this CT antigen as a potentially relevant clinical target in haematological malignancies. Further studies of protein expression are now needed to validate these antigens as vaccine candidates.

Could protein tertiary structure influence mammary transgene expression more than tissue specific codon usage?

Animal mammary glands have been successfully employed to produce therapeutic recombinant human proteins. However, considerable variation in animal mammary transgene expression efficiency has been reported. We now consider whether aspects of codon usage and/or protein tertiary structure underlie this variation in mammary transgene expression.

Comparison and consolidation of microarray data sets of human tissue expression

Background

Human tissue displays a remarkable diversity in structure and function. To understand how such diversity emerges from the same DNA, systematic measurements of gene expression across different tissues in the human body are essential. Several recent studies addressed this formidable task using microarray technologies. These large tissue expression data sets have provided us an important basis for biomedical research. However, it is well known that microarray data can be compromised by high noise level and various experimental artefacts. Critical comparison of different data sets can help to reveal such errors and to avoid pitfalls in their application.

Results

We present here the first comparison and integration of four freely available tissue expression data sets generated using three different microarray platforms and containing a total of 377 microarray hybridizations. When assessing the tissue expression of genes, we found that the results considerably depend on the chosen data set. Nevertheless, the comparison also revealed statistically significant similarity of gene expression profiles across different platforms. This enabled us to construct consolidated lists of platform-independent tissue-specific genes using a set of complementary measures. Follow-up analyses showed that results based on consolidated data tend to be more reliable.

Conclusions

Our study strongly indicates that the consolidation of the four different tissue expression data sets can increase data quality and can lead to biologically more meaningful results. The provided compendium of platform-independent gene lists should facilitate the identification of novel tissue-specific marker genes.

Gene expression profiles identify inflammatory signatures in dendritic cells

Dendritic cells (DCs) constitute a heterogeneous group of antigen-presenting leukocytes important in activation of both innate and adaptive immunity. We studied the gene expression patterns of DCs incubated with reagents inducing their activation or inhibition. Total RNA was isolated from DCs and gene expression profiling was performed with oligonucleotide microarrays. Using a supervised learning algorithm based on Random Forest, we generated a molecular signature of inflammation from a training set of 77 samples. We then validated this molecular signature in a testing set of 38 samples. Supervised analysis identified a set of 44 genes that distinguished very accurately between inflammatory and non inflammatory samples. The diagnostic performance of the signature genes was assessed against an independent set of samples, by qRT-PCR. Our findings suggest that the gene expression signature of DCs can provide a molecular classification for use in the selection of anti-inflammatory or adjuvant molecules with specific effects on DC activity.

Positive selection neighboring functionally essential sites and disease-implicated regions of mammalian reproductive proteins

Background

Reproductive proteins are central to the continuation of all mammalian species. The evolution of these proteins has been greatly influenced by environmental pressures induced by pathogens, rival sperm, sexual selection and sexual conflict. Positive selection has been demonstrated in many of these proteins with particular focus on primate lineages. However, the mammalia are a diverse group in terms of mating habits, population sizes and germ line generation times. We have examined the selective pressures at work on a number of novel reproductive proteins across a wide variety of mammalia.

Results

We show that selective pressures on reproductive proteins are highly varied. Of the 10 genes analyzed in detail, all contain signatures of positive selection either across specific sites or in specific lineages or a combination of both. Our analysis of SP56 and Col1a1 are entirely novel and the results show positively selected sites present in each gene. Our findings for the Col1a1 gene are suggestive of a link between positive selection and severe disease type. We find evidence in our dataset to suggest that interacting proteins are evolving in symphony: most likely to maintain interacting functionality.

Conclusion

Our in silico analyses show positively selected sites are occurring near catalytically important regions suggesting selective pressure to maximize efficient fertilization. In those cases where a mechanism of protein function is not fully understood, the sites presented here represent ideal candidates for mutational study. This work has highlighted the widespread rate heterogeneity in mutational rates across the mammalia and specifically has shown that the evolution of reproductive proteins is highly varied depending on the species and interacting partners. We have shown that positive selection and disease are closely linked in the Col1a1 gene.

Comparison of breast cancer to healthy control tissue discovers novel markers with potential for prognosis and early detection

This study was initiated to identify biomarkers with potential value for the early detection of poor-outcome breast cancer. Two sets of well-characterized tissues were utilized: one from breast cancer patients with favorable vs. poor outcome and the other from healthy women undergoing reduction mammaplasty. Over 46 differentially expressed genes were identified from a large list of potential targets by a) mining publicly available expression data (identifying 134 genes for quantitative PCR) and b) utilizing a commercial PCR array. Three genes show elevated expression in cancers with poor outcome and low expression in all other tissues, warranting further investigation as potential blood markers for early detection of cancers with poor outcome. Twelve genes showed lower expression in cancers with poor outcome than in cancers with favorable outcome but no differential expression between aggressive cancers and most healthy controls. These genes are more likely to be useful as prognostic tissue markers than as serum markers for early detection of aggressive disease. As a secondary finding was that, when histologically normal breast tissue was removed from a distant site in a breast with cancer, 7 of 38 specimens displayed a cancer-like expression profile, while the remaining 31 were genetically similar to the reduction mammaplasty control group. This finding suggests that some regions of ipsilateral histologically ‘normal’ breast tissue are predisposed to becoming malignant and that normal-appearing tissue with malignant signature might warrant treatment to prevent new primary tumors.

Splicing factor and exon profiling across human tissues

It has been shown that alternative splicing is especially prevalent in brain and testis when compared to other tissues. To test whether there is a specific propensity of these tissues to generate splicing variants, we used a single source of high-density microarray data to perform both splicing factor and exon expression profiling across 11 normal human tissues. Paired comparisons between tissues and an original exon-based statistical group analysis demonstrated after extensive RT-PCR validation that the cerebellum, testis, and spleen had the largest proportion of differentially expressed alternative exons. Variations at the exon level correlated with a larger number of splicing factors being expressed at a high level in the cerebellum, testis and spleen than in other tissues. However, this splicing factor expression profile was similar to a more global gene expression pattern as a larger number of genes had a high expression level in the cerebellum, testis and spleen. In addition to providing a unique resource on expression profiling of alternative splicing variants and splicing factors across human tissues, this study demonstrates that the higher prevalence of alternative splicing in a subset of tissues originates from the larger number of genes, including splicing factors, being expressed than in other tissues.

Disease gene characterization through large-scale co-expression analysis

BACKGROUND

In the post genome era, a major goal of biology is the identification of specific roles for individual genes. We report a new genomic tool for gene characterization, the UCLA Gene Expression Tool (UGET).

RESULTS

Celsius, the largest co-normalized microarray dataset of Affymetrix based gene expression, was used to calculate the correlation between all possible gene pairs on all platforms, and generate stored indexes in a web searchable format. The size of Celsius makes UGET a powerful gene characterization tool. Using a small seed list of known cartilage-selective genes, UGET extended the list of known genes by identifying 32 new highly cartilage-selective genes. Of these, 7 of 10 tested were validated by qPCR including the novel cartilage-specific genes SDK2 and FLJ41170. In addition, we retrospectively tested UGET and other gene expression based prioritization tools to identify disease-causing genes within known linkage intervals. We first demonstrated this utility with UGET using genetically heterogeneous disorders such as Joubert syndrome, microcephaly, neuropsychiatric disorders and type 2 limb girdle muscular dystrophy (LGMD2) and then compared UGET to other gene expression based prioritization programs which use small but discrete and well annotated datasets. Finally, we observed a significantly higher gene correlation shared between genes in disease networks associated with similar complex or Mendelian disorders.

DISCUSSION

UGET is an invaluable resource for a geneticist that permits the rapid inclusion of expression criteria from one to hundreds of genes in genomic intervals linked to disease. By using thousands of arrays UGET annotates and prioritizes genes better than other tools especially with rare tissue disorders or complex multi-tissue biological processes. This information can be critical in prioritization of candidate genes for sequence analysis.

Global expression profile of highly enriched cardiomyocytes derived from human embryonic stem cells

Human embryonic stem cells (hESC), with their ability to differentiate into cardiomyocytes in culture, hold great potential for cell replacement therapies and provide an in vitro model of human heart development. A genomewide characterization of the molecular phenotype of hESC-derived cardiomyocytes is important for their envisioned applications. We have employed a lineage selection strategy to generate a pure population of cardiomyocytes (>99%) from transgenic hESC lines. Global gene expression profiling showed that these cardiomyocytes are distinct from pluripotent and differentiated hESC cultures. Pure cardiomyocytes displayed similarities with heart tissue, but in many aspects presented an individual transcriptome pattern. A subset of 1,311 cardiac-enriched transcripts was identified, which were significantly overpresented (p < .01) in the Gene Ontology (GO) categories of heart function and heart development. Focused analysis of the GO categories ion transport, sarcomere, and heart development uncovered a unique molecular signature of hESC cardiomyocytes. Pathway analysis revealed an extensive cardiac transcription factor network and novel peroxisome proliferator-activated receptor signaling components within the cardiac-enriched genes. Notably, approximately 80% of these genes were previously uncharacterized. We have evaluated the biological relevance of four candidates-Rbm24, Tcea3, Fhod3, and C15orf52-by in situ hybridization during early mouse development and report that all were prominently expressed in cardiac structures. Our results provide the fundamental basis for a comprehensive understanding of gene expression patterns of hESC cardiomyocytes and will greatly help define biological processes and signaling pathways involved in hESC cardiomyogenic differentiation and in human heart development.

Novel rank-based statistical methods reveal microRNAs with differential expression in multiple cancer types

Background

microRNAs (miRNAs) regulate target genes at the post-transcriptional level and play important roles in cancer pathogenesis and development. Variation amongst individuals is a significant confounding factor in miRNA (or other) expression studies. The true character of biologically or clinically meaningful differential expression can be obscured by inter-patient variation. In this study we aim to identify miRNAs with consistent differential expression in multiple tumor types using a novel data analysis approach.

Methods

Using microarrays we profiled the expression of more than 700 miRNAs in 28 matched tumor/normal samples from 8 different tumor types (breast, colon, liver, lung, lymphoma, ovary, prostate and testis). This set is unique in putting emphasis on minimizing tissue type and patient related variability using normal and tumor samples from the same patient. We develop scores for comparing miRNA expression in the above matched sample data based on a rigorous characterization of the distribution of order statistics over a discrete state set, including exact p-values. Specifically, we compute a Rank Consistency Score (RCoS) for every miRNA measured in our data. Our methods are also applicable in various other contexts. We compare our methods, as applied to matched samples, to paired t-test and to the Wilcoxon Signed Rank test.

Results

We identify consistent (across the cancer types measured) differentially expressed miRNAs. 41 miRNAs are under-expressed in cancer compared to normal, at FDR (False Discovery Rate) of 0.05 and 17 are over-expressed at the same FDR level. Differentially expressed miRNAs include known oncomiRs (e.g miR-96) as well as miRNAs that were not previously universally associated with cancer. Specific examples include miR-133b and miR-486-5p, which are consistently down regulated and mir-629* which is consistently up regulated in cancer, in the context of our cohort. Data is available in GEO. Software is available at: http://bioinfo.cs.technion.ac.il/people/zohar/RCoS/

Collagen and calcium-binding EGF domains 1 is frequently inactivated in ovarian cancer by aberrant promoter hypermethylation and modulates cell migration and survival

Background:

Collagen and calcium-binding EGF domains 1 (CCBE1) is an uncharacterised gene that has down-regulated expression in breast cancer. As CCBE1 maps to 18q21.32, a region frequently exhibiting loss of heterozygosity in ovarian cancer, the aim of this study was to determine the expression and function of CCBE1 in ovarian cancer.

Methods:

Expression and methylation patterns of CCBE1 were determined in ovarian cancer cell lines and primary tumours. CCBE1 contains collagen repeats and an aspartic acid/asparagine hydroxylation/EGF-like domain, suggesting a function in extracellular matrix remodelling and migration, which was determined using small-interfering RNA (siRNA)-mediated knockdown and over-expression of CCBE1 in cell lines.

Results:

CCBE1 is expressed in normal ovary, but is reduced in ovarian cancer cell lines and primary carcinomas. Pharmacological demethylation/deacetylation in ovarian cancer cell lines re-induced CCBE1 expression, indicating that epigenetic mechanisms contribute to its silencing in cancer. CCBE1 promoter hypermethylation was detected in 6/11 (55%) ovarian cancer cell lines and 38/81 (41%) ovarian carcinomas. siRNA-mediated knockdown of CCBE1 in ovarian cancer cell lines enhanced their migration; conversely, re-expression of CCBE1 reduced migration and survival. Hence, loss of CCBE1 expression may promote ovarian carcinogenesis by enhancing migration and cell survival.

Conclusions:

These data suggest that CCBE1 is a new candidate tumour suppressor in ovarian cancer.

CRX is a diagnostic marker of retinal and pineal lineage tumors

Background

CRX is a homeobox transcription factor whose expression and function is critical to maintain retinal and pineal lineage cells and their progenitors. To determine the biologic and diagnostic potential of CRX in human tumors of the retina and pineal, we examined its expression in multiple settings.

Methodology/Principal Findings

Using situ hybridization and immunohistochemistry we show that Crx RNA and protein expression are exquisitely lineage restricted to retinal and pineal cells during normal mouse and human development. Gene expression profiling analysis of a wide range of human cancers and cancer cell lines also supports that CRX RNA is highly lineage restricted in cancer. Immunohistochemical analysis of 22 retinoblastomas and 13 pineal parenchymal tumors demonstrated strong expression of CRX in over 95% of these tumors. Importantly, CRX was not detected in the majority of tumors considered in the differential diagnosis of pineal region tumors (n = 78). The notable exception was medulloblastoma, 40% of which exhibited CRX expression in a heterogeneous pattern readily distinguished from that seen in retino-pineal tumors.

Conclusions/Significance

These findings describe new potential roles for CRX in human cancers and highlight the general utility of lineage restricted transcription factors in cancer biology. They also identify CRX as a sensitive and specific clinical marker and a potential lineage dependent therapeutic target in retinoblastoma and pineoblastoma.

Expression profiling of repetitive elements by melting temperature analysis: variation in HERV-W gag expression across human individuals and tissues

Background

Human endogenous retroviruses (HERV) constitute approximately 8% of the human genome and have long been considered "junk". The sheer number and repetitive nature of these elements make studies of their expression methodologically challenging. Hence, little is known of transcription of genomic regions harboring such elements.

Results

Applying a recently developed technique for obtaining high resolution melting temperature data, we examined the frequency distributions of HERV-W gag element into 13 Tm categories in human tissues. Transcripts containing HERV-W gag sequences were expressed in non-random patterns with extensive variations in the expression between both tissues, including different brain regions, and individuals. Furthermore, the patterns of such transcripts varied more between individuals in brain regions than other tissues.

Conclusion

Thus, regulated expression of non-coding regions of the human genome appears to include the HERV-W family of repetitive elements. Although it remains to be established whether such expression patterns represent leakage from transcription of functional regions or specific transcription, the current approach proves itself useful for studying detailed expression patterns of repetitive regions.

Evolution of biomolecular networks: lessons from metabolic and protein interactions

Despite only becoming popular at the beginning of this decade, biomolecular networks are now frameworks that facilitate many discoveries in molecular biology. The nodes of these networks are usually proteins (specifically enzymes in metabolic networks), whereas the links (or edges) are their interactions with other molecules. These networks are made up of protein-protein interactions or enzyme-enzyme interactions through shared metabolites in the case of metabolic networks. Evolutionary analysis has revealed that changes in the nodes and links in protein-protein interaction and metabolic networks are subject to different selection pressures owing to distinct topological features. However, many evolutionary constraints can be uncovered only if temporal and spatial aspects are included in the network analysis.

Complex architecture and regulated expression of the Sox2ot locus during vertebrate development

The Sox2 gene is a key regulator of pluripotency embedded within an intron of a long noncoding RNA (ncRNA), termed Sox2 overlapping transcript (Sox2ot), which is transcribed in the same orientation. However, this ncRNA remains uncharacterized. Here we show that Sox2ot has multiple transcription start sites associated with genomic features that indicate regulated expression, including highly conserved elements (HCEs) and chromatin marks characteristic of gene promoters. To identify biological processes in which Sox2ot may be involved, we analyzed its expression in several developmental systems, compared to expression of Sox2. We show that Sox2ot is a stable transcript expressed in mouse embryonic stem cells, which, like Sox2, is down-regulated upon induction of embryoid body (EB) differentiation. However, in contrast to Sox2, Sox2ot is up-regulated during EB mesoderm-lineage differentiation. In adult mouse, Sox2ot isoforms were detected in tissues where Sox2 is expressed, as well as in different tissues, supporting independent regulation of expression of the ncRNA. Sox2dot, an isoform of Sox2ot transcribed from a distal HCE located >500 kb upstream of Sox2, was detected exclusively in the mouse brain, with enrichment in regions of adult neurogenesis. In addition, Sox2ot isoforms are transcribed from HCEs upstream of Sox2 in other vertebrates, including in several regions of the human brain. We also show that Sox2ot is dynamically regulated during chicken and zebrafish embryogenesis, consistently associated with central nervous system structures. These observations provide insight into the structure and regulation of the Sox2ot gene, and suggest conserved roles for Sox2ot orthologs during vertebrate development.