A computational/functional genomics approach for the enrichment of the retinal transcriptome and the identification of positional candidate retinopathy genes

Transcriptome of the human retina, retinal pigmented epithelium and choroid

The retina and its adjacent supporting tissues - retinal pigmented epithelium (RPE) and choroid - are critical structures in human eyes required for normal visual perception. Abnormal changes in these layers have been implicated in diseases such as age-related macular degeneration and glaucoma. With the advent of high-throughput methods, such as serial analysis of gene expression, cDNA microarray, and RNA sequencing, there is unprecedented opportunity to facilitate our understanding of the normal retina, RPE, and choroid. This information can be used to identify dysfunction in age-related macular degeneration and glaucoma. In this review, we describe the current status in our understanding of these transcriptomes through the use of high-throughput techniques.

Genome-wide analysis of immune system genes by expressed sequence Tag profiling

Profiling studies of mRNA and microRNA, particularly microarray-based studies, have been extensively used to create compendia of genes that are preferentially expressed in the immune system. In some instances, functional studies have been subsequently pursued. Recent efforts such as the Encyclopedia of DNA Elements have demonstrated the benefit of coupling RNA sequencing analysis with information from expressed sequence tags (ESTs) for transcriptomic analysis. However, the full characterization and identification of transcripts that function as modulators of human immune responses remains incomplete. In this study, we demonstrate that an integrated analysis of human ESTs provides a robust platform to identify the immune transcriptome. Beyond recovering a reference set of immune-enriched genes and providing large-scale cross-validation of previous microarray studies, we discovered hundreds of novel genes preferentially expressed in the immune system, including noncoding RNAs. As a result, we have established the Immunogene database, representing an integrated EST road map of gene expression in human immune cells, which can be used to further investigate the function of coding and noncoding genes in the immune system. Using this approach, we have uncovered a unique metabolic gene signature of human macrophages and identified PRDM15 as a novel overexpressed gene in human lymphomas. Thus, we demonstrate the utility of EST profiling as a basis for further deconstruction of physiologic and pathologic immune processes.

Genetic networks in the mouse retina: growth associated protein 43 and phosphatase tensin homolog network

PURPOSE

The present study examines the structure and covariance of endogenous variation in gene expression across the recently expanded family of C57BL/6J (B) X DBA/2J (D) Recombinant Inbred (BXD RI) strains of mice. This work is accompanied by a highly interactive database that can be used to generate and test specific hypotheses. For example, we define the genetic network regulating growth associated protein 43 (Gap43) and phosphatase tensin homolog (Pten).

METHODS

The Hamilton Eye Institute (HEI) Retina Database within GeneNetwork features the data analysis of 346 Illumina Sentrix BeadChip Arrays (mouse whole genome-6 version 2). Eighty strains of mice are presented, including 75 BXD RI strains, the parental strains (C57BL/6J and DBA/2J), the reciprocal crosses, and the BALB/cByJ mice. Independent biologic samples for at least two animals from each gender were obtained with a narrow age range (48 to 118 days). Total RNA was prepared followed by the production of biotinylated cRNAs, which were pipetted into the Mouse WG-6V2 arrays. The data was globally normalized with rank invariant and stabilization (2z+8).

RESULTS

The HEI Retina Database is located on the GeneNetwork website. The database was used to extract unique transcriptome signatures for specific cell types in the retina (retinal pigment epithelial, amacrine, and retinal ganglion cells). Two genes associated with axonal outgrowth (Gap43 and Pten) were used to display the power of this new retina database. Bioinformatic tools located within GeneNetwork in conjunction with the HEI Retina Database were used to identify the unique signature Quantitative Trait Loci (QTLs) for Gap43 and Pten on chromosomes 1, 2, 12, 15, 16, and 19. Gap43 and Pten possess networks that are similar to ganglion cell networks that may be associated with axonal growth in the mouse retina. This network involves high correlations of transcription factors (SRY sex determining region Y-box 2 [Sox2], paired box gene 6 [Pax6], and neurogenic differentiation 1 [Neurod1]), and genes involved in DNA binding (proliferating cell nuclear antigen [Pcna] and zinc finger, BED-type containing 4 [Zbed4]), as well as an inhibitor of DNA binding (inhibitor of DNA binding 2, dominant negative helix-loop-helix protein [Id2]). Furthermore, we identified the potential upstream modifiers on chromosome 2 (teashirt zinc finger homeobox 2 [Tshz2], RNA export 1 homolog [Rae1] and basic helix-loop-helix domain contatining, class B4 [Bhlhb4]) on chromosome 15 (RAB, member of RAS oncogene family-like 2a [Rabl2a], phosphomannomutase 1 [Pmm1], copine VIII [Cpne8], and fibulin 1 [Fbln1]).

CONCLUSIONS

The endogenous variation in mRNA levels among BXD RI strains can be used to explore and test expression networks underlying variation in retina structure, function, and disease susceptibility. The Gap43 and Pten network highlights the covariance of gene expression and forms a molecular network associated with axonal outgrowth in the adult retina.

Alcohol impairment of saccadic and smooth pursuit eye movements: impact of risk factors for alcohol dependence

RATIONALE

While persons at risk for alcohol dependence by virtue of heavy drinking patterns or family history (FH) of alcohol use disorders have exhibited differential alcohol responses on a variety of measures, few studies have examined alcohol's effects on eye movements in these subgroups.

OBJECTIVES

The purpose of this study was to (1) conduct a placebo-controlled, dose-ranging study of alcohol's effects on eye movements and (2) examine the impact of these risk factors on oculomotor response to alcohol.

METHODS

A within-subject, double-blind laboratory study was conducted in N = 138 heavy (HD; n = 78) and light social drinkers (LD; n = 60) with self-reported positive (FH+) or negative (FH-) family history. Subjects participated in three laboratory sessions in which they consumed a beverage containing a high (0.8 g/kg) or low (0.4 g/kg) dose of alcohol or placebo. Smooth pursuit, pro-saccadic, and anti-saccadic eye movements were recorded before and at two intervals after alcohol consumption.

RESULTS

Alcohol significantly impaired smooth pursuit gain and pro- and anti-saccade latency, velocity, and accuracy in a dose and time specific matter. HD and LD showed similar impairment on smooth pursuit gain and anti-saccade measures, but HD were less impaired in pro-saccade latency, velocity, and accuracy. FH+ and FH- subjects were equally impaired in nearly all pro- and anti-saccade measures, but FH+ were less impaired in smooth pursuit gain.

CONCLUSIONS

In sum, alcohol produced systematic impairment on oculomotor functioning, even at a non-intoxicating dose. Furthermore, high- and low-risk drinkers may be vulnerable to select performance deficits relative to eye movement task.

Exploiting ESTs in human health

Expressed Sequence Tags (ESTs) are fragments of cDNA clones. They correspond to the most abundant type of cDNA information available in the public databases. ESTs have been used for expression profiling, gene identification, characterization of differentially expressed genes, and identification of transcript variants among other utilities. In this review I will discuss the major features of the collection of ESTs available in the public domain giving a special emphasis on how this dataset has been used in studies about human diseases.

In silico strategy for detection of target candidates for antibody therapy of solid tumors

In contrast to earlier attempts for the identification of target candidates suitable for monoclonal antibody (mAb) based cancer therapies we concentrated on highly selective lineage-specific genes additionally preserved or even overexpressed in orthotopic cancers. In a script aided workflow we reduced all human entries of the RefSeq mRNA database to those encoding transmembrane domain bearing gene products and subjected them to BLAST analysis against the human EST database. All BLAST results were validated in a gene centric way allowing two types of data curation prior to expression profiling of matching ESTs in selected healthy tissues: (i) exclusion of questionable ESTs arising e.g. from genomic contamination and (ii) elimination of erroneously predicted mRNAs as well as transcripts with only weak EST coverage. The impact of such stringent input control on accuracy of prediction is underlined by RT-PCR confirmation of predicted tissue distribution patterns for a number of selected candidates.

Genomic expression during human myelopoiesis

Background

Human myelopoiesis is an exciting biological model for cellular differentiation since it represents a plastic process where multipotent stem cells gradually limit their differentiation potential, generating different precursor cells which finally evolve into distinct terminally differentiated cells. This study aimed at investigating the genomic expression during myeloid differentiation through a computational approach that integrates gene expression profiles with functional information and genome organization.

Results

Gene expression data from 24 experiments for 8 different cell types of the human myelopoietic lineage were used to generate an integrated myelopoiesis dataset of 9,425 genes, each reliably associated to a unique genomic position and chromosomal coordinate. Lists of genes constitutively expressed or silent during myelopoiesis and of genes differentially expressed in commitment phase of myelopoiesis were first identified using a classical data analysis procedure. Then, the genomic distribution of myelopoiesis genes was investigated integrating transcriptional and functional characteristics of genes. This approach allowed identifying specific chromosomal regions significantly highly or weakly expressed, and clusters of differentially expressed genes and of transcripts related to specific functional modules.

Conclusion

The analysis of genomic expression during human myelopoiesis using an integrative computational approach allowed discovering important relationships between genomic position, biological function and expression patterns and highlighting chromatin domains, including genes with coordinated expression and lineage-specific functions.

Head and neck squamous cell carcinoma transcriptome analysis by comprehensive validated differential display

Head and neck squamous cell carcinoma (HNSCC) is common worldwide and is associated with a poor rate of survival. Identification of new markers and therapeutic targets, and understanding the complex transformation process, will require a comprehensive description of genome expression, that can only be achieved by combining different methodologies. We report here the HNSCC transcriptome that was determined by exhaustive differential display (DD) analysis coupled with validation by different methods on the same patient samples. The resulting 820 nonredundant sequences were analysed by high throughput bioinformatics analysis. Human proteins were identified for 73% (596) of the DD sequences. A large proportion (>50%) of the remaining unassigned sequences match ESTs (expressed sequence tags) from human tumours. For the functionally annotated proteins, there is significant enrichment for relevant biological processes, including cell motility, protein biosynthesis, stress and immune responses, cell death, cell cycle, cell proliferation and/or maintenance and transport. Three of the novel proteins (TMEM16A, PHLDB2 and ARHGAP21) were analysed further to show that they have the potential to be developed as therapeutic targets.

In silico identification and comparative analysis of differentially expressed genes in human and mouse tissues

Background

Screening for differentially expressed genes on the genomic scale and comparative analysis of the expression profiles of orthologous genes between species to study gene function and regulation are becoming increasingly feasible. Expressed sequence tags (ESTs) are an excellent source of data for such studies using bioinformatic approaches because of the rich libraries and tremendous amount of data now available in the public domain. However, any large-scale EST-based bioinformatics analysis must deal with the heterogeneous, and often ambiguous, tissue and organ terms used to describe EST libraries.

Results

To deal with the issue of tissue source, in this work, we carefully screened and organized more than 8 million human and mouse ESTs into 157 human and 108 mouse tissue/organ categories, to which we applied an established statistic test using different thresholds of the p value to identify genes differentially expressed in different tissues. Further analysis of the tissue distribution and level of expression of human and mouse orthologous genes showed that tissue-specific orthologs tended to have more similar expression patterns than those lacking significant tissue specificity. On the other hand, a number of orthologs were found to have significant disparity in their expression profiles, hinting at novel functions, divergent regulation, or new ortholog relationships.

Conclusion

Comprehensive statistics on the tissue-specific expression of human and mouse genes were obtained in this very large-scale, EST-based analysis. These statistical results have been organized into a database, freely accessible at our website , for easy searching of human and mouse tissue-specific genes and for investigating gene expression profiles in the context of comparative genomics. Comparative analysis showed that, although highly tissue-specific genes tend to exhibit similar expression profiles in human and mouse, there are significant exceptions, indicating that orthologous genes, while sharing basic genomic properties, could result in distinct phenotypes.

Machine learning approaches to supporting the identification of photoreceptor-enriched genes based on expression data

Background

Retinal photoreceptors are highly specialised cells, which detect light and are central to mammalian vision. Many retinal diseases occur as a result of inherited dysfunction of the rod and cone photoreceptor cells. Development and maintenance of photoreceptors requires appropriate regulation of the many genes specifically or highly expressed in these cells. Over the last decades, different experimental approaches have been developed to identify photoreceptor enriched genes. Recent progress in RNA analysis technology has generated large amounts of gene expression data relevant to retinal development. This paper assesses a machine learning methodology for supporting the identification of photoreceptor enriched genes based on expression data.

Results

Based on the analysis of publicly-available gene expression data from the developing mouse retina generated by serial analysis of gene expression (SAGE), this paper presents a predictive methodology comprising several in silico models for detecting key complex features and relationships encoded in the data, which may be useful to distinguish genes in terms of their functional roles. In order to understand temporal patterns of photoreceptor gene expression during retinal development, a two-way cluster analysis was firstly performed. By clustering SAGE libraries, a hierarchical tree reflecting relationships between developmental stages was obtained. By clustering SAGE tags, a more comprehensive expression profile for photoreceptor cells was revealed. To demonstrate the usefulness of machine learning-based models in predicting functional associations from the SAGE data, three supervised classification models were compared. The results indicated that a relatively simple instance-based model (KStar model) performed significantly better than relatively more complex algorithms, e.g. neural networks. To deal with the problem of functional class imbalance occurring in the dataset, two data re-sampling techniques were studied. A random over-sampling method supported the implementation of the most powerful prediction models. The KStar model was also able to achieve higher predictive sensitivities and specificities using random over-sampling techniques.

Conclusion

The approaches assessed in this paper represent an efficient and relatively inexpensive in silico methodology for supporting large-scale analysis of photoreceptor gene expression by SAGE. They may be applied as complementary methodologies to support functional predictions before implementing more comprehensive, experimental prediction and validation methods. They may also be combined with other large-scale, data-driven methods to facilitate the inference of transcriptional regulatory networks in the developing retina. Furthermore, the methodology assessed may be applied to other data domains.

A text-mining analysis of the human phenome

A number of large-scale efforts are underway to define the relationships between genes and proteins in various species. But, few attempts have been made to systematically classify all such relationships at the phenotype level. Also, it is unknown whether such a phenotype map would carry biologically meaningful information. We have used text mining to classify over 5000 human phenotypes contained in the Online Mendelian Inheritance in Man database. We find that similarity between phenotypes reflects biological modules of interacting functionally related genes. These similarities are positively correlated with a number of measures of gene function, including relatedness at the level of protein sequence, protein motifs, functional annotation, and direct protein-protein interaction. Phenotype grouping reflects the modular nature of human disease genetics. Thus, phenotype mapping may be used to predict candidate genes for diseases as well as functional relations between genes and proteins. Such predictions will further improve if a unified system of phenotype descriptors is developed. The phenotype similarity data are accessible through a web interface at http://www.cmbi.ru.nl/MimMiner/.

Identification of regulatory targets of tissue-specific transcription factors: application to retina-specific gene regulation

Identification of tissue-specific gene regulatory networks can yield insights into the molecular basis of a tissue's development, function and pathology. Here, we present a computational approach designed to identify potential regulatory target genes of photoreceptor cell-specific transcription factors (TFs). The approach is based on the hypothesis that genes related to the retina in terms of expression, disease and/or function are more likely to be the targets of retina-specific TFs than other genes. A list of genes that are preferentially expressed in retina was obtained by integrating expressed sequence tag, SAGE and microarray datasets. The regulatory targets of retina-specific TFs are enriched in this set of retina-related genes. A Bayesian approach was employed to integrate information about binding site location relative to a gene's transcription start site. Our method was applied to three retina-specific TFs, CRX, NRL and NR2E3, and a number of potential targets were predicted. To experimentally assess the validity of the bioinformatic predictions, mobility shift, transient transfection and chromatin immunoprecipitation assays were performed with five predicted CRX targets, and the results were suggestive of CRX regulation in 5/5, 3/5 and 4/5 cases, respectively. Together, these experiments strongly suggest that RP1, GUCY2D, ABCA4 are novel targets of CRX.

Comprehensive in silico functional specification of mouse retina transcripts

BACKGROUND

The retina is a well-defined portion of the central nervous system (CNS) that has been used as a model for CNS development and function studies. The full specification of transcripts in an individual tissue or cell type, like retina, can greatly aid the understanding of the control of cell differentiation and cell function. In this study, we have integrated computational bioinformatics and microarray experimental approaches to classify the tissue specificity and developmental distribution of mouse retina transcripts.

RESULTS

We have classified a set of retina-specific genes using sequence-based screening integrated with computational and retina tissue-specific microarray approaches. 33,737 non-redundant sequences were identified as retina transcript clusters (RTCs) from more than 81,000 mouse retina ESTs. We estimate that about 19,000 to 20,000 genes might express in mouse retina from embryonic to adult stages. 39.1% of the RTCs are not covered by 60,770 RIKEN full-length cDNAs. Through comparison with 2 million mouse ESTs, spectra of neural, retinal, late-generated retinal, and photoreceptor -enriched RTCs have been generated. More than 70% of these RTCs have data from biological experiments confirming their tissue-specific expression pattern. The highest-grade retina-enriched pool covered almost all the known genes encoding proteins involved in photo-transduction.

CONCLUSION

This study provides a comprehensive mouse retina transcript profile for further gene discovery in retina and suggests that tissue-specific transcripts contribute substantially to the whole transcriptome.

Integration of text- and data-mining using ontologies successfully selects disease gene candidates

Genome-wide techniques such as microarray analysis, Serial Analysis of Gene Expression (SAGE), Massively Parallel Signature Sequencing (MPSS), linkage analysis and association studies are used extensively in the search for genes that cause diseases, and often identify many hundreds of candidate disease genes. Selection of the most probable of these candidate disease genes for further empirical analysis is a significant challenge. Additionally, identifying the genes that cause complex diseases is problematic due to low penetrance of multiple contributing genes. Here, we describe a novel bioinformatic approach that selects candidate disease genes according to their expression profiles. We use the eVOC anatomical ontology to integrate text-mining of biomedical literature and data-mining of available human gene expression data. To demonstrate that our method is successful and widely applicable, we apply it to a database of 417 candidate genes containing 17 known disease genes. We successfully select the known disease gene for 15 out of 17 diseases and reduce the candidate gene set to 63.3% (±18.8%) of its original size. This approach facilitates direct association between genomic data describing gene expression and information from biomedical texts describing disease phenotype, and successfully prioritizes candidate genes according to their expression in disease-affected tissues.

The Retinome - defining a reference transcriptome of the adult mammalian retina/retinal pigment epithelium

BACKGROUND

The mammalian retina is a valuable model system to study neuronal biology in health and disease. To obtain insight into intrinsic processes of the retina, great efforts are directed towards the identification and characterization of transcripts with functional relevance to this tissue.

RESULTS

With the goal to assemble a first genome-wide reference transcriptome of the adult mammalian retina, referred to as the retinome, we have extracted 13,037 non-redundant annotated genes from nearly 500,000 published datasets on redundant retina/retinal pigment epithelium (RPE) transcripts. The data were generated from 27 independent studies employing a wide range of molecular and biocomputational approaches. Comparison to known retina-/RPE-specific pathways and established retinal gene networks suggest that the reference retinome may represent up to 90% of the retinal transcripts. We show that the distribution of retinal genes along the chromosomes is not random but exhibits a higher order organization closely following the previously observed clustering of genes with increased expression.

CONCLUSION

The genome wide retinome map offers a rational basis for selecting suggestive candidate genes for hereditary as well as complex retinal diseases facilitating elaborate studies into normal and pathological pathways. To make this unique resource freely available we have built a database providing a query interface to the reference retinome 1.

A transcript finishing initiative for closing gaps in the human transcriptome

We report the results of a transcript finishing initiative, undertaken for the purpose of identifying and characterizing novel human transcripts, in which RT-PCR was used to bridge gaps between paired EST clusters, mapped against the genomic sequence. Each pair of EST clusters selected for experimental validation was designated a transcript finishing unit (TFU). A total of 489 TFUs were selected for validation, and an overall efficiency of 43.1% was achieved. We generated a total of 59,975 bp of transcribed sequences organized into 432 exons, contributing to the definition of the structure of 211 human transcripts. The structure of several transcripts reported here was confirmed during the course of this project, through the generation of their corresponding full-length cDNA sequences. Nevertheless, for 21% of the validated TFUs, a full-length cDNA sequence is not yet available in public databases, and the structure of 69.2% of these TFUs was not correctly predicted by computer programs. The TF strategy provides a significant contribution to the definition of the complete catalog of human genes and transcripts, because it appears to be particularly useful for identification of low abundance transcripts expressed in a restricted set of tissues as well as for the delineation of gene boundaries and alternatively spliced isoforms.

Annotation and analysis of 10,000 expressed sequence tags from developing mouse eye and adult retina

BACKGROUND

As a biomarker of cellular activities, the transcriptome of a specific tissue or cell type during development and disease is of great biomedical interest. We have generated and analyzed 10,000 expressed sequence tags (ESTs) from three mouse eye tissue cDNA libraries: embryonic day 15.5 (M15E) eye, postnatal day 2 (M2PN) eye and adult retina (MRA).

RESULTS

Annotation of 8,633 non-mitochondrial and non-ribosomal high-quality ESTs revealed that 57% of the sequences represent known genes and 43% are unknown or novel ESTs, with M15E having the highest percentage of novel ESTs. Of these, 2,361 ESTs correspond to 747 unique genes and the remaining 6,272 are represented only once. Phototransduction genes are preferentially identified in MRA, whereas transcripts for cell structure and regulatory proteins are highly expressed in the developing eye. Map locations of human orthologs of known genes uncovered a high density of ocular genes on chromosome 17, and identified 277 genes in the critical regions of 37 retinal disease loci. In silico expression profiling identified 210 genes and/or ESTs over-expressed in the eye; of these, more than 26 are known to have vital retinal function. Comparisons between libraries provided a list of temporally regulated genes and/or ESTs. A few of these were validated by qRT-PCR analysis.

CONCLUSIONS

Our studies present a large number of potentially interesting genes for biological investigation, and the annotated EST set provides a useful resource for microarray and functional genomic studies.

New methods for finding disease-susceptibility genes: impact and potential

Improved techniques for defining disease-gene location and evaluating the biological candidacy of regional transcripts will hasten disease-gene discovery.

Improved techniques for defining disease-gene location and evaluating the biological candidacy of regional transcripts will hasten disease-gene discovery.

Identification and characterization of C1orf36, a transcript highly expressed in photoreceptor cells, and mutation analysis in retinitis pigmentosa

By means of computational methods, we identified an uncharacterized human transcript, Chromosome 1 open reading frame 36 (C1orf36), that is expressed in the retina and that maps to 1q32.3. The cDNA contains an open reading frame of 585bp that encodes a 195-aminoacid protein with a predicted mass of 22.7kDa. An alternatively spliced transcript in a retinoblastoma cell line, encoding for a truncated peptide, was also identified. PCR experiments performed using human cDNA from several sources indicate that C1orf36 has a preferential expression in the retina. Accordingly, in situ hybridization experiments, performed using as probe a murine C1orf36 cDNA fragment, detected a hybridization signal on mouse retinal adult sections. The C1orf36 protein shares homology with putative proteins in Mus musculus and Fugu rubripes, suggesting evolutionary conservation of its function. Additional sequence analysis of the C1orf36 gene product predicts its subcellular mitochondrial localization and the presence of both evolutionary conserved phosphorylation sites and regions adopting a coiled-coil conformation. We also defined the genomic structure of the gene. This enabled us to perform a mutational analysis of the C1orf36 coding region of about 300 patients affected by retinitis pigmentosa. No pathological mutations were detected in this analysis.

Identification of expressed sequence tags preferentially expressed in human placentas by in silico subtraction

OBJECTIVES

To identify expressed sequence tag (EST) clusters preferentially expressed in placentas.

METHODS

The National Center for Biotechnology's online UniGene database contains 14 placenta libraries. In silico (computer-based) subtraction compared placenta libraries against the remaining libraries to identify transcripts preferentially expressed in placentas. For known genes, placental expression or their use in prenatal diagnosis was then explored online using LocusLink and PubMed.

RESULTS

Placentas preferentially expressed 475 EST clusters. Of these, 18 EST clusters with no known function were expressed exclusively in placentas. Of the remaining 457 EST clusters, 90 showed preferential placental expression by >/=25 times. Of these 90, literature searches on the 45 EST clusters with known functions showed 44 linked to placental physiology or proposed as markers for prenatal diagnosis [i.e. beta-hCG, pregnancy-specific glycoproteins, human placental lactogens, pregnancy-associated plasma protein A (PAPP-A)]. Selected genes with known function in pregnancy but whose preferential placental expression fell below the factor of 25 threshold were also identified.

CONCLUSION

In silico subtraction identified 44 previously studied genes involved in placental physiology as well as 63 EST clusters preferentially expressed in placental tissue, which may serve as targets for future studies seeking novel markers for prenatal diagnosis or to better understand placental genetics.