Analysing gene expression data from DNA microarrays to identify candidate genes

mRNA expression profile reveals differentially expressed genes in splenocytes of experimental autoimmune encephalomyelitis model

Experimental autoimmune encephalomyelitis (EAE) is a mouse model that can be used to investigate aetiology, pathogenesis, and treatment approaches for multiple sclerosis (MS). A novel integrated bioinformatics approach was used to understand the involvement of differentially expressed genes (DEGs) in the spleen of EAE mice through data mining of existing microarray and RNA-seq datasets. We screened differentially expressed mRNAs using mRNA expression profile data of EAE spleens taken from Gene Expression Omnibus (GEO). Functional and pathway enrichment analyses of DEGs were performed by Database for Annotation, Visualization, and Integrated Discovery (DAVID). Subsequently, the DEGs-encoded protein-protein interaction (PPI) network was constructed. The 784 DEGs in GSE99300 A.SW PP-EAE mice spleen mRNA profiles, 859 DEGs in GSE151701 EAE mice spleen mRNA profiles, and 646 DEGs in GSE99300 SJL/J PP-EAE mice spleen mRNA profiles were explored. Functional enrichment of 55 common DEGs among 3 sub-datasets revealed several immune-related terms, such as neutrophil extravasation, leucocyte migration, antimicrobial humoral immune response mediated by an antimicrobial peptide, toll-like receptor 4 bindings, IL-17 signalling pathway, and TGF-beta signalling pathway. In the screening of 10 hub genes, including MPO, ELANE, CTSG, LTF, LCN2, SELP, CAMP, S100A9, ITGA2B, and PRTN3, and in choosing and validating the 5 DEGs, including ANK1, MBOAT2, SLC25A21, SLC43A1, and SOX6, the results showed that SLC43A1 and SOX6 were significantly decreased in EAE mice spleen. Thus this study offers a list of genes expressed in the spleen that might play a key role in the pathogenesis of EAE.

A Network-Based Integrative Workflow to Unravel Mechanisms Underlying Disease Progression

Unraveling mechanisms underlying diseases has motivated the development of systems biology approaches. The key challenges for the development of mathematical models and computational tool are (1) the size of molecular networks, (2) the nonlinear nature of spatio-temporal interactions, and (3) feedback loops in the structure of interaction networks. We here propose an integrative workflow that combines structural analyses of networks, high-throughput data, and mechanistic modeling. As an illustration of the workflow, we use prostate cancer as a case study with the aim of identifying key functional components associated with primary to metastasis transitions. Analysis carried out by the workflow revealed that HOXD10, BCL2, and PGR are the most important factors affected in primary prostate samples, whereas, in the metastatic state, STAT3, JUN, and JUNB are playing a central role. The identified key elements of each network are validated using patient survival analysis. The workflow presented here allows experimentalists to use heterogeneous data sources for the identification of diagnostic and prognostic signatures.

Diagnosis using clinical/pathological and molecular information

In diagnosis and classification diseases multiple outcomes, both molecular and clinical/pathological are routinely gathered on patients. In recent years, many approaches have been suggested for integrating gene expression (continuous data) with clinical/pathological data (usually categorical and ordinal data). This new area of research integrates both clinical and genomic data in order to improve our knowledge about diseases, and to capture the information which is lost in independent clinical or genomic studies. The related metric scaling distance is a not well-known, but very valuable distance to integrate clinical/pathological and molecular information. In this article, we present the use of the related metric scaling distance in biomedical research. We describe how this distance works, and we also explain why it may sometimes be preferred. We discuss the choice of the related metric scaling distance and compare it with other proximity measures to include both clinical and genetic information. Furthermore, we comment the choice of the related metric scaling distance when classical clustering or discriminant analysis based on distances are performed and compare the results with more complex cluster or discriminant procedures specially constructed for integrating clinical and molecular information. The use of the related metric scaling distance is illustrated on simulated experimental and four real data sets, a heart disease, and three cancer studies. The results present the flexibility and availability of this distance which gives competitive results.

Detection and preliminary screening of the human gene expression profile for Hirschsprung's disease

The present study investigated a genome microarray of colorectal lesions (spasm segments) in children with Hirschsprung's disease (HSCR), and analyzed the results. In addition, the present study screened for differentially expressed genes in children with HSCR. Microarray technology was used to examine the human gene expression profiles of the colorectal lesions (spasm segments) of six children with HSCR, and three normal colon tissue samples. The data were analyzed be determining P‑values of significance and absolute fold changes. Preliminary screening was performed to identify genes exhibiting significant differential expression in children with HSCR, and these target genes were analyzed in subsequent verification and analytical investigations. Of >20,000 detected human genes, the preliminary screenings demonstrated that 3,850 genes were differentially expressed and upregulated, with P<0.05 and >2‑fold absolute changes in expression. In addition, 645 differentially expressed genes with P<0.05 and >2‑fold absolute changes were downregulated. Of the upregulated genes, 118 were involved in classic signaling pathways, compared with 11 of the downregulated genes (P<0.001; absolute fold change >2‑fold). HSCR etiology is complex and often involves multiple gene changes. Microarray technology can produce large quantities of gene expression data simultaneously, and analyzing this data using various techniques may provide a fast and efficient method for identifying novel gene targets and for investigating the mechanisms underlying HSCR pathogenesis.

Development of novel diagnostic and prognostic molecular markers for sporadic colon cancer

Gene expression studies are informative about changes in colon cancer, increase understanding of the biology of tumorigenesis and aid in developing diagnostic and prognostic markers. In this review, expression techniques used to examine the multistage process of colon cancer are discussed. Many genes have been found to differ in expression between normal and tumorigenic states, as early as the seemingly normal colonic crypts. The clinical usefulness of markers varies with stage, ethnicity and anatomic location of colon cancer. Thus, combinations of markers can be used to develop an approach to molecularly screen and follow the progression of this prevalent cancer.

Array-based methods for diagnosis and prevention of transplant rejection

DNA microarray is a microhybridization-based assay that is used to simultaneously study the expression of thousands of genes, thus providing a global view of gene expression in a tissue sample. This powerful technique has been adopted by many biomedical disciplines and will likely have a profound impact on the diagnosis, treatment and prognosis of human diseases. This review article presents an overview of the application of microarray technology to the field of solid-organ transplantation.

Heterogeneity of gene expression in murine squamous cell carcinoma development-the same tumor by different means

Transformation is a complex process, involving many changes in the cell. In this work, we investigated the transcriptional changes that arose during the development of squamous cell carcinoma (SCC) in mice. Using microarray analysis, we looked at gene expression during different stages in cancer progression in 31 mice. By analyzing tumor progression in each mouse separately, we were able to define the global changes that were common to all 31 mice, as well as significant changes that occurred in fewer individuals. We found that different genes can contribute to the tumorigenic process in different mice, and that there are many ways to acquire the malignant properties defined by Hanahan and Weinberg as "hallmarks of cancer". Eventually, however, all these changes lead to a very similar cancerous phenotype. The finding that gene expression is strongly heterogeneous in tumors that were induced by a standardized protocol in closely related mice underscores the need for molecular characterization of human tumors and personalized therapy.

A Bayesian Approach for Decision Making on the Identification of Genes with Different Expression Levels: An Application toEscherichia coliBacterium Data

A common interest in gene expression data analysis is to identify from a large pool of candidate genes the genes that present significant changes in expression levels between a treatment and a control biological condition. Usually, it is done using a statistic value and a cutoff value that are used to separate the genes differentially and nondifferentially expressed. In this paper, we propose a Bayesian approach to identify genes differentially expressed calculating sequentially credibility intervals from predictive densities which are constructed using the sampled mean treatment effect from all genes in study excluding the treatment effect of genes previously identified with statistical evidence for difference. We compare our Bayesian approach with the standard ones based on the use of the t-test and modified t-tests via a simulation study, using small sample sizes which are common in gene expression data analysis. Results obtained report evidence that the proposed approach performs better than standard ones, especially for cases with mean differences and increases in treatment variance in relation to control variance. We also apply the methodologies to a well-known publicly available data set on Escherichia coli bacterium.

Gene expression data analysis using closed item set mining for labeled data

This article presents an approach to microarray data analysis using discretised expression values in combination with a methodology of closed item set mining for class labeled data (RelSets). A statistical 2 x 2 factorial design analysis was run in parallel. The approach was validated on two independent sets of two-color microarray experiments using potato plants. Our results demonstrate that the two different analytical procedures, applied on the same data, are adequate for solving two different biological questions being asked. Statistical analysis is appropriate if an overview of the consequences of treatments and their interaction terms on the studied system is needed. If, on the other hand, a list of genes whose expression (upregulation or downregulation) differentiates between classes of data is required, the use of the RelSets algorithm is preferred. The used algorithms are freely available upon request to the authors.

Internal standard-based analysis of microarray data. Part 1: analysis of differential gene expressions

Genome-scale microarray experiments for comparative analysis of gene expressions produce massive amounts of information. Traditional statistical approaches fail to achieve the required accuracy in sensitivity and specificity of the analysis. Since the problem can be resolved neither by increasing the number of replicates nor by manipulating thresholds, one needs a novel approach to the analysis. This article describes methods to improve the power of microarray analyses by defining internal standards to characterize features of the biological system being studied and the technological processes underlying the microarray experiments. Applying these methods, internal standards are identified and then the obtained parameters are used to define (i) genes that are distinct in their expression from background; (ii) genes that are differentially expressed; and finally (iii) genes that have similar dynamical behavior.

Targeting stem cells-clinical implications for cancer therapy

Cancer stem cells (CSC), also called tumor initiating cells (TIC), are considered to be the origin of replicating malignant tumor cells in a variety of human cancers. Their presence in the tumor may herald malignancy potential, mediate resistance to conventional chemotherapy or radiotherapy, and confer poor survival outcomes. Thus, CSC may serve as critical cellular targets for treatment. The ability to therapeutically target CSC hinges upon identifying their unique cell surface markers and the underlying survival signaling pathways. While accumulating evidence suggests cell-surface antigens (such as CD44, CD133) as CSC markers for several tumor tissues, emerging clinical needs exist for the identification of new markers to completely separate CSC from normal stem cells. Recent studies have demonstrated the critical role of the tumor suppressor PTEN/PI3 kinase pathway in regulating TIC in leukemia, brain, and intestinal tissues. The successful eradication of tumors by therapies targeting CSC will require an in-depth understanding of the molecular mechanisms governing CSC self renewal, differentiation, and escape from conventional therapy. Here we review recent progress from brain tumor and intestinal stem cell research with a focus on the PTEN-Akt-Wnt pathway, and how the components of CSC pathways may serve as biomarkers for diagnosis, prognosis, and therapeutics.

Gene selection for brain cancer classification

With the introduction of microarray, cancer classification, diagnosis and prediction are made more accurate and effective. However, the final outcome of the data analyses very much depend on the huge number of genes with relatively small number of samples present in each experiment. It is thus crucial to select relevant genes to be used for future specific cancer markers. Many feature selection methods have been proposed but none is able to classify all kinds of microarray data accurately, especially on those multi-class datasets. We propose a one-versus-one comparison method for selecting discriminatory features instead of performing the statistical test in a one-versus-all manner. Brain cancer is chosen as an example. Here, 3 types of statistics are used: signal-to-noise ratio (SNR), t-statistics and Pearson correlation coefficient. Results are verified by performing hierarchical and k-means clustering. Using our one-versus-one comparisons, best performance accuracies of 90.48% and 97.62% can be obtained by hierarchical and k-means clustering respectively. However best performance accuracies of 88.10% and 80.95% can be obtained respectively when using one-versus-all comparison. This shows that one-versus-one comparison is superior.

Optimization of comparative expressed sequence hybridization for genome-wide expression profiling at chromosome level

Comparative expressed sequence hybridization (CESH) has recently been developed for global expression profiling at chromosome level. To improve its specificity and sensitivity, we examined the effects of cDNA amplification and labeling methods on CESH profiles, using a gastric cancer cell line, Kato III, and compared the CESH profiles to cDNA microarray and reverse transcriptase-polymerase chain reaction (RT-PCR) data. CESH results were scarcely affected by the amplification process, either at the RNA level with T7 polymerase or at the cDNA level with degenerate oligonucleotide-primed PCR (DOP-PCR). The labeling method, however, did remarkably affect the CESH results; false positive shifts of the test/reference ratio (T/R) were not detected in self-matched CESH with pre-cDNA labeling and random priming labeling of cDNA but were consistently seen with DOP-PCR labeling in 11 chromosomes. The use of cDNA deriving from mRNA either with pre-cDNA or random priming labeling gave results of higher detection sensitivity for regions of up- or downregulated expression and higher concordance with the microarray and RT-PCR data in the corresponding regions than with conventional CESH. This modification of CESH with random priming labeling was found feasible by its application to Kato III cells with and without 5-aza-2'-deoxycytidine treatment; the regions identified as epigenetically silenced included genes that were reportedly methylated in Kato III.

Bayesian meta-analysis models for microarray data: a comparative study

Background

With the growing abundance of microarray data, statistical methods are increasingly needed to integrate results across studies. Two common approaches for meta-analysis of microarrays include either combining gene expression measures across studies or combining summaries such as p-values, probabilities or ranks. Here, we compare two Bayesian meta-analysis models that are analogous to these methods.

Results

Two Bayesian meta-analysis models for microarray data have recently been introduced. The first model combines standardized gene expression measures across studies into an overall mean, accounting for inter-study variability, while the second combines probabilities of differential expression without combining expression values. Both models produce the gene-specific posterior probability of differential expression, which is the basis for inference. Since the standardized expression integration model includes inter-study variability, it may improve accuracy of results versus the probability integration model. However, due to the small number of studies typical in microarray meta-analyses, the variability between studies is challenging to estimate. The probability integration model eliminates the need to model variability between studies, and thus its implementation is more straightforward. We found in simulations of two and five studies that combining probabilities outperformed combining standardized gene expression measures for three comparison values: the percent of true discovered genes in meta-analysis versus individual studies; the percent of true genes omitted in meta-analysis versus separate studies, and the number of true discovered genes for fixed levels of Bayesian false discovery. We identified similar results when pooling two independent studies of Bacillus subtilis. We assumed that each study was produced from the same microarray platform with only two conditions: a treatment and control, and that the data sets were pre-scaled.

Conclusion

The Bayesian meta-analysis model that combines probabilities across studies does not aggregate gene expression measures, thus an inter-study variability parameter is not included in the model. This results in a simpler modeling approach than aggregating expression measures, which accounts for variability across studies. The probability integration model identified more true discovered genes and fewer true omitted genes than combining expression measures, for our data sets.

Designer genes: Filling the gap in transplantation

Accelerated evolution of the field of functional genomics has been greatly facilitated by high-throughput microarray-based gene function studies, relating to the parallel and serial expression measurements of genomes. Microarray experimentation is being applied for the study of basic research questions, drug target discovery, pharmacology, toxicogenomics, target selectivity, disease biomarker determination, development of prognostic tests, and disease subclass determination. This article will review the current applications of microarray technology in the field of organ transplantation and discuss the potential impact of this technology on transplantation medicine.

The hepatic transcriptome in human liver disease

The transcriptome is the mRNA transcript pool in a cell, organ or tissue with the liver transcriptome being amongst the most complex of any organ. Functional genomics methodologies are now being widely utilized to study transcriptomes including the hepatic transcriptome. This review outlines commonly used methods of transcriptome analysis, especially gene array analysis, focusing on publications utilizing these methods to understand human liver disease. Additionally, we have outlined the relationship between transcript and protein expressions as well as summarizing what is known about the variability of the transcriptome in non-diseased liver tissue. The approaches covered include gene array analysis, serial analysis of gene expression, subtractive hybridization and differential display. The discussion focuses on primate whole organ studies and in-vitro cell culture systems utilized. It is now clear that there are a vast number research opportunities for transcriptome analysis of human liver disease as we attempt to better understand both non-diseased and disease hepatic mRNA expression. We conclude that hepatic transcriptome analysis has already made significant contributions to the understanding of human liver pathobiology.

Genomics and proteomics: emerging technologies in clinical cancer research

Fueled by the complete genomic data acquired from the human genome project and the desperate clinical need of comprehensive analytical tools to study a heterogeneous disease like cancer, genomic and proteomic technologies have evolved rapidly, accelerating the rate and number of discoveries in clinical cancer research. These discoveries include mechanistic understanding of cancer biology as well as the identification of biomarkers supporting early detection, molecular classification of tumors, molecular predictors of metastasis, treatment response, and prognosis. While the technical advances have been significant, clinical researchers and practicing physicians are now confronted with the challenges of understanding technically and statistically complex data sets, translating this complex information to fit clinical contexts and incorporating it into clinical studies. In this review, we will summarize the available technologies and associated bioinformatics, discuss studies that are clinically relevant, and discuss the limitations we are still facing. We will present a framework for future directions of these technologies and how we believe they should be applied in clinical studies.

Expression of protein elongation factor eEF1A2 predicts favorable outcome in breast cancer

Breast cancer is the most common malignancy among North American women. The identification of factors that predict outcome is key to individualized disease management and to our understanding of breast oncogenesis. We have analyzed mRNA expression of protein elongation factor eEF1A2 in two independent breast tumor populations of size n = 345 and n = 88, respectively. We find that eEF1A2 mRNA is expressed at a low level in normal breast epithelium but is detectably expressed in approximately 50-60% of primary human breast tumors. We have derived an eEF1A2-specific antibody and measured eEF1A2 protein expression in a sample of 438 primary breast tumors annotated with 20-year survival data. We find that high levels of eEF1A2 protein are detected in 60% of primary breast tumors independent of HER-2 protein expression, tumor size, lymph node status, and estrogen receptor (ER) expression. Importantly, we find that high eEF1A2 is a significant predictor of outcome. Women whose tumor has high eEF1A2 protein expression have an increased probability of 20-year survival compared to those women whose tumor does not express substantial eEF1A2. In addition, eEF1A2 protein expression predicts increased survival probability in those breast cancer patients whose tumor is HER-2 negative or who have lymph node involvement.

CD24 is a new oncogene, early at the multistep process of colorectal cancer carcinogenesis

BACKGROUND & AIMS

The aim of this study was to identify genes that play a role in colorectal cancer (CRC) carcinogenesis by analysis of differential gene expression of normal and transformed CRC cell lines.

METHODS

Gene expression array analysis ([RG-U34] GeneChip) was performed in normal and transformed rat intestinal epithelial cells before and after exposures to celecoxib. In particular, we were looking for (1) altered gene expression in the transformed cells that reverts to normal following exposure to a selective cyclooxygenase-2 inhibitor, (2) novel genes, and (3) genes encoding membrane receptors or ligands. As a validation of the results and for human patients, immunohistochemistry was performed on 398 biological samples from the gastrointestinal tract (normal, polyps, and adenocarcinomas). Human cancer cell lines were tested for their response to anti-CD24 monoclonal antibodies.

RESULTS

A total of 1081 genes were differently expressed following malignant transformation; 71 genes showed altered expression that reverted to normal following treatment with celecoxib, including the CD24 gene. Immunohistochemistry confirmed that increased expression of CD24 is an early event in CRC carcinogenesis. It was expressed in 90.7% of adenomas and 86.3% of CRCs. Very low expression was seen in normal epithelium (16.6%). Human cancer cell lines showed growth inhibition in response to the antibodies, according to their expression levels of CD24 and in dose- and time-dependent manners. These results were repetitive for 3 different antibodies.

CONCLUSIONS

CD24 is overexpressed in the colonic mucosa, already at an early stage of carcinogenesis. It may be a useful target for early detection and in CRC therapy.

Dose-response relationships in gene expression profiles in rainbow trout, Oncorhyncus mykiss, exposed to ethynylestradiol

Determining how gene expression profiles change with toxicant dose will improve the utility of arrays in identifying biomarkers and modes of toxic action. Isogenic rainbow trout, Oncorhyncus mykiss,were exposed to 10, 50 or 100 ng/L ethynylestradiol (a xeno-estrogen) for 7 days. Following exposure hepatic RNA was extracted. Fluorescently labeled cDNA were generated and hybridized against a commercially available Atlantic Salmon/Trout array (GRASP project, University of Victoria) spotted with 16,000 cDNAs. Transcript expression in treated vs control fish was analyzed via Genespring (Silicon Genetics) to identify genes with altered expression, as well as to determine gene clustering patterns that can be used as "expression signatures". Array results were confirmed via qRT PCR. Our analysis indicates that gene expression profiles varied somewhat with dose. Established biomarkers of exposure to estrogenic chemicals, such as vitellogenin, vitelline envelope proteins, and the estrogen receptor alpha, were induced at every dose. Other genes were dose specific, suggesting that different doses induce distinct physiological responses. These findings demonstrate that cDNA microarrays could be used to identify both toxicant class and relative dose.

Predicting survival within the lung cancer histopathological hierarchy using a multi-scale genomic model of development

BACKGROUND

The histopathologic heterogeneity of lung cancer remains a significant confounding factor in its diagnosis and prognosis-spurring numerous recent efforts to find a molecular classification of the disease that has clinical relevance.

METHODS AND FINDINGS

Molecular profiles of tumors from 186 patients representing four different lung cancer subtypes (and 17 normal lung tissue samples) were compared with a mouse lung development model using principal component analysis in both temporal and genomic domains. An algorithm for the classification of lung cancers using a multi-scale developmental framework was developed. Kaplan-Meier survival analysis was conducted for lung adenocarcinoma patient subgroups identified via their developmental association. We found multi-scale genomic similarities between four human lung cancer subtypes and the developing mouse lung that are prognostically meaningful. Significant association was observed between the localization of human lung cancer cases along the principal mouse lung development trajectory and the corresponding patient survival rate at three distinct levels of classical histopathologic resolution: among different lung cancer subtypes, among patients within the adenocarcinoma subtype, and within the stage I adenocarcinoma subclass. The earlier the genomic association between a human tumor profile and the mouse lung development sequence, the poorer the patient's prognosis. Furthermore, decomposing this principal lung development trajectory identified a gene set that was significantly enriched for pyrimidine metabolism and cell-adhesion functions specific to lung development and oncogenesis.

CONCLUSIONS

From a multi-scale disease modeling perspective, the molecular dynamics of murine lung development provide an effective framework that is not only data driven but also informed by the biology of development for elucidating the mechanisms of human lung cancer biology and its clinical outcome.

Defective ribosomal protein gene expression alters transcription, translation, apoptosis, and oncogenic pathways in Diamond-Blackfan anemia

Diamond-Blackfan anemia (DBA) is a broad developmental disease characterized by anemia, bone marrow (BM) erythroblastopenia, and an increased incidence of malignancy. Mutations in ribosomal protein gene S19 (RPS19) are found in approximately 25% of DBA patients; however, the role of RPS19 in the pathogenesis of DBA remains unknown. Using global gene expression analysis, we compared highly purified multipotential, erythroid, and myeloid BM progenitors from RPS19 mutated and control individuals. We found several ribosomal protein genes downregulated in all DBA progenitors. Apoptosis genes, such as TNFRSF10B and FAS, transcriptional control genes, including the erythropoietic transcription factor MYB (encoding c-myb), and translational genes were greatly dysregulated, mostly in diseased erythroid cells. Cancer-related genes, including RAS family oncogenes and tumor suppressor genes, were significantly dysregulated in all diseased progenitors. In addition, our results provide evidence that RPS19 mutations lead to codownregulation of multiple ribosomal protein genes, as well as downregulation of genes involved in translation in DBA cells. In conclusion, the altered expression of cancer-related genes suggests a molecular basis for malignancy in DBA. Downregulation of c-myb expression, which causes complete failure of fetal liver erythropoiesis in knockout mice, suggests a link between RPS19 mutations and reduced erythropoiesis in DBA.

Gene expression patterns in rainbow trout, Oncorhynchus mykiss, exposed to a suite of model toxicants

The increased availability and use of DNA microarrays has allowed the characterization of gene expression patterns associated with exposure to different toxicants. An important question is whether toxicant induced changes in gene expression in fish are sufficiently diverse to allow for identification of specific modes of action and/or specific contaminants. In theory, each class of toxicant may generate a gene expression profile unique to its mode of toxic action. In this study, isogenic (cloned) rainbow trout Oncorhynchus mykiss were exposed to sublethal levels of a series of model toxicants with varying modes of action, including ethynylestradiol (xeno-estrogen), 2,2,4,4'-tetrabromodiphenyl ether (BDE-47, thyroid active), diquat (oxidant stressor), chromium VI, and benzo[a]pyrene (BaP) for a period of 1-3 weeks. An additional experiment measured trenbolone (anabolic steroid; model androgen) induced gene expression changes in sexually mature female trout. Following exposure, fish were euthanized, livers removed and RNA extracted. Fluorescently labeled cDNA were generated and hybridized against a commercially available Atlantic Salmon/Trout array (GRASP project, University of Victoria) spotted with 16,000 cDNA's. The slides were scanned to measure abundance of a given transcript in each sample relative to controls. Data were analyzed via Genespring (Silicon Genetics) to identify a list of up- and downregulated genes, as well as to determine gene clustering patterns that can be used as "expression signatures". The results indicate each toxicant exposure caused between 64 and 222 genes to be significantly altered in expression. Most genes exhibiting altered expression responded to only one of the toxicants and relatively few were co-expressed in multiple treatments. For example, BaP and Diquat, both of which exert toxicity via oxidative stress, upregulated 28 of the same genes, of over 100 genes altered by either treatment. Other genes associated with steroidogenesis, p450 and estrogen responsive genes appear to be useful for selectively identifying toxicant mode of action in fish, suggesting a link between gene expression profile and mode of toxicity. Our array results showed good agreement with quantitative real time polymerase chain reaction (qRT PCR), which demonstrates that the arrays are an accurate measure of gene expression. The specificity of the gene expression profile in response to a model toxicant, the link between genes with altered expression and mode of toxic action, and the consistency between array and qRT PCR results all suggest that cDNA microarrays have the potential to screen environmental contaminants for biomarkers and mode of toxic action.

Bayesian models for pooling microarray studies with multiple sources of replications

Background

Biologists often conduct multiple but different cDNA microarray studies that all target the same biological system or pathway. Within each study, replicate slides within repeated identical experiments are often produced. Pooling information across studies can help more accurately identify true target genes. Here, we introduce a method to integrate multiple independent studies efficiently.

Results

We introduce a Bayesian hierarchical model to pool cDNA microarray data across multiple independent studies to identify highly expressed genes. Each study has multiple sources of variation, i.e. replicate slides within repeated identical experiments. Our model produces the gene-specific posterior probability of differential expression, which provides a direct method for ranking genes, and provides Bayesian estimates of false discovery rates (FDR). In simulations combining two and five independent studies, with fixed FDR levels, we observed large increases in the number of discovered genes in pooled versus individual analyses. When the number of output genes is fixed (e.g., top 100), the pooled model found appreciably more truly differentially expressed genes than the individual studies. We were also able to identify more differentially expressed genes from pooling two independent studies in Bacillus subtilis than from each individual data set. Finally, we observed that in our simulation studies our Bayesian FDR estimates tracked the true FDRs very well.

Conclusion

Our method provides a cohesive framework for combining multiple but not identical microarray studies with several sources of replication, with data produced from the same platform. We assume that each study contains only two conditions: an experimental and a control sample. We demonstrated our model's suitability for a small number of studies that have been either pre-scaled or have no outliers.

Novel susceptibility genes in inflammatory bowel disease

The inflammatory bowel disease, Crohn’s disease and ulcerative colitis, are polygenic disorders with important environmental interactions. To date, the most widely adopted approach to identifying susceptibility genes in complex diseases has involved genome wide linkage studies followed by studies of positional candidate genes in loci of interest. This review encompasses data from studies into novel candidate genes implicated in the pathogenesis of inflammatory bowel disease. Novel techniques to identify candidate genes-genome wide association studies, yeast-two hybrid screening, microarray gene expression studies and proteomic profiling, are also reviewed and their potential role in unravelling the pathogenesis of inflammatory bowel disease are discussed.

Chipping into the human genome: novel insights for transplantation

High throughput, high density platforms for transcriptional, proteomic, and metabonomic analyses are opening new doors for improving our understanding of the complexity and redundancy of the immune system in the interplay of the innate and allo-immune responses in organ transplantation. New insights are being obtained into the possible discrepancies between the gold standard of tissue pathological diagnosis and clinical graft outcomes, as new transcriptional categories of transplant rejection evolve. The bystander effects of chronic immunosuppression underlying the complexities of graft dysfunction are beginning to be understood. Non-invasive mechanisms to monitor transplants, by following 'footprints' of biomarker sets that reflect the disease phenotype, are being pursued for their clinical application for direct patient care. Utilization of these same biomarker sets may also offer a unique means to titrate immunosuppression and predict specific graft dysfunction events prior to clinical decline, thus bringing in the potential to reduce patient morbidity from infection and malignancy, preserve graft integrity, and limit the progression of chronic graft injury. Bioinformatics support is integral to the unraveling of the mysteries of the human genome, proteome, and metabolome in disease and in health.

Metabolic engineering in the -omics era: elucidating and modulating regulatory networks

The importance of regulatory control in metabolic processes is widely acknowledged, and several enquiries (both local and global) are being made in understanding regulation at various levels of the metabolic hierarchy. The wealth of biological information has enabled identifying the individual components (genes, proteins, and metabolites) of a biological system, and we are now in a position to understand the interactions between these components. Since phenotype is the net result of these interactions, it is immensely important to elucidate them not only for an integrated understanding of physiology, but also for practical applications of using biological systems as cell factories. We present some of the recent "-omics" approaches that have expanded our understanding of regulation at the gene, protein, and metabolite level, followed by analysis of the impact of this progress on the advancement of metabolic engineering. Although this review is by no means exhaustive, we attempt to convey our ideology that combining global information from various levels of metabolic hierarchy is absolutely essential in understanding and subsequently predicting the relationship between changes in gene expression and the resulting phenotype. The ultimate aim of this review is to provide metabolic engineers with an overview of recent advances in complementary aspects of regulation at the gene, protein, and metabolite level and those involved in fundamental research with potential hurdles in the path to implementing their discoveries in practical applications.

Multivariate analysis of microarray data by principal component discriminant analysis: prioritizing relevant transcripts linked to the degradation of different carbohydrates in Pseudomonas putida S12

The value of the multivariate data analysis tools principal component analysis (PCA) and principal component discriminant analysis (PCDA) for prioritizing leads generated by microarrays was evaluated. To this end, Pseudomonas putida S12 was grown in independent triplicate fermentations on four different carbon sources, i.e. fructose, glucose, gluconate and succinate. RNA isolated from these samples was analysed in duplicate on an anonymous clone-based array to avoid bias during data analysis. The relevant transcripts were identified by analysing the loadings of the principal components (PC) and discriminants (D) in PCA and PCDA, respectively. Even more specifically, the relevant transcripts for a specific phenotype could also be ranked from the loadings under an angle (biplot) obtained after PCDA analysis. The leads identified in this way were compared with those identified using the commonly applied fold-difference and hierarchical clustering approaches. The different data analysis methods gave different results. The methods used were complementary and together resulted in a comprehensive picture of the processes important for the different carbon sources studied. For the more subtle, regulatory processes in a cell, the PCDA approach seemed to be the most effective. Except for glucose and gluconate dehydrogenase, all genes involved in the degradation of glucose, gluconate and fructose were identified. Moreover, the transcriptomics approach resulted in potential new insights into the physiology of the degradation of these carbon sources. Indications of iron limitation were observed with cells grown on glucose, gluconate or succinate but not with fructose-grown cells. Moreover, several cytochrome- or quinone-associated genes seemed to be specifically up- or downregulated, indicating that the composition of the electron-transport chain in P. putida S12 might change significantly in fructose-grown cells compared to glucose-, gluconate- or succinate-grown cells.

Clinical applications of genomics in head and neck cancer

Advances in gene expression analyses have allowed global assessment of expressed genes in clinical samples. Gene expression profiles derived from clinical specimens have been used to distinguish differences in tumors that are not obvoius by clinical, radiographic, or histologic characteristics. Despite its common histology and presentation, head and neck squamous cell carcinoma (HNSCC) is associated with widely varying clinical behavior and response to therapy. Currently, clinicians have a dearth of tools to predict response to therapy or to identify patients at high risk of poor outcome. Recently comprehensive analyses of gene expression patterns of individual tumors have shown promise to improve discovery of biomarkers for 1) progression of premalignant lesions, 2) disease presence or absence, 3) prediction of clinical outcome, and 4) identification of targets for therapy. In this review, we will discuss advances, limitations and future directions of genomics as it applies to HNSCC.

In silico analysis of angiogenesis associated gene expression identifies angiogenic stage related profiles

In vitro models have been extensively used to map gene expression in ECs but few studies have used cells from in vivo sources directly. Here, we compare different gene expression surveys on both cultured and fresh tissue derived ECs, and it emerges that gene expression profiles can be paralleled with the angiogenic stage of the cells. ECs stimulated with different growth factors in monolayer cultures exhibit gene expression profiles indicative of an active proliferative state, whereas gene expression in tube forming cells in vitro involves genes implicated in cell adhesion processes. Genes overexpressed in tumor ECs are biased towards extracellular matrix remodeling, a late event in angiogenesis. The elucidation of gene expression profiles under these different conditions will contribute to a better understanding of the molecular mechanisms during angiogenesis in both pathological and physiological circumstances and will have implications for the development of angiogenesis interfering treatment strategies.

Bacillus clausii effect on gene expression pattern in small bowel mucosa using DNA microarray analysis

BACKGROUND

Probiotics are widely used for the cure or prevention of several clinical conditions. However, clinical decisions need to be substantiated by an analysis of the complex bacteria-host interplay in the intestinal lumen.

AIMS

To identify the gene expression pattern induced by Bacillus clausii in the intestinal mucosa of healthy individuals.

METHODS

Six male patients (mean age 38+/-5 years) affected by endoscopically confirmed mild oesophagitis were treated for one month with esomeprazole, and were randomly selected to receive or not B. clausii (groups I and II, respectively). Duodenal biopsies were taken pre and post-treatment to identify the modification of gene expression, using the GeneChip Human U133A array. To validate the microarray analysis, real-time reverse transcriptase-polymerase chain reaction (PCR) of five target genes was performed.

RESULTS

After B. clausii administration, a total of 158 and 265 genes were up and downregulated, respectively. Quantitative PCR confirmed the microarray data. B. clausii mainly affected the expression of genes involved in immune response and inflammation, apoptosis and cell growth, cell differentiation, cell-cell signalling, cell adhesion, signal transcription and transduction.

CONCLUSIONS

Our data represent the first global analysis of B. clausii effects on the gene expression profile in normal intestine, and provide the basis to identify the mechanisms by which these agents interact with the host and exert their beneficial effects. Future studies are needed to clarify the B. clausii-induced gene pattern in the clinical disorders in which probiotics have proved to be effective.

Nonparametric tests for differential gene expression and interaction effects in multi-factorial microarray experiments

Background

Numerous nonparametric approaches have been proposed in literature to detect differential gene expression in the setting of two user-defined groups. However, there is a lack of nonparametric procedures to analyze microarray data with multiple factors attributing to the gene expression. Furthermore, incorporating interaction effects in the analysis of microarray data has long been of great interest to biological scientists, little of which has been investigated in the nonparametric framework.

Results

In this paper, we propose a set of nonparametric tests to detect treatment effects, clinical covariate effects, and interaction effects for multifactorial microarray data. When the distribution of expression data is skewed or heavy-tailed, the rank tests are substantially more powerful than the competing parametric F tests. On the other hand, in the case of light or medium-tailed distributions, the rank tests appear to be marginally less powerful than the parametric competitors.

Conclusion

The proposed rank tests enable us to detect differential gene expression and establish interaction effects for microarray data with various non-normally distributed expression measurements across genome. In the presence of outliers, they are advantageous alternative approaches to the existing parametric F tests due to the robustness feature.

Gene expression arrays in cancer research: methods and applications

During the last 5 years, the number of papers describing data obtained by microarray technology increased exponentially with about 3000 papers in 2003. Undoubtedly, cancer is by far the disease that received most of the attention as far as the amount of data generated. As array technology is rather new and highly dependent on bioinformatics, mathematics and statistics, a clear understanding of the knowledge and information derived from array-based experiments is not widely appreciated. We shall review herein some of the issues related to the construction of DNA arrays, quantities and heterogeneity of probes and targets, the consequences of the physical characteristics of the probes, data extraction and data analysis as well as the applications of array technology. Our goal is to bring to the general audience, some of the basics of array technology and its possible application in oncology. By discussing some of the basic aspects of the methodology, we hope to stimulate criticism concerning the conclusions proposed by authors, especially in the light of the very low degree of reproducibility already proven when commercially available platforms were compared . Regardless of its pitfalls, it is unquestionable that array technology will have a great impact in the management of cancer and its applications will range from the discovery of new drug targets, new molecular tools for diagnosis and prognosis as well as for a tailored treatment that will take into account the molecular determinants of a given tumor. Hence, we shall also highlight some of the already available and promising applications of array technology on the day-to-day practice of oncology.

Comparison of gene expression in trap cells and vegetative hyphae of the nematophagous fungus Monacrosporium haptotylum

Nematode-trapping fungi enter the parasitic stage by developing specific morphological structures called traps. The global patterns of gene expression in traps and mycelium of the fungusMonacrosporium haptotylumwere compared. The trap of this fungus is a unicellular spherical structure called the knob, which develops on the apex of a hyphal branch. RNA was isolated from knobs and mycelium and hybridized to a cDNA array containing probes of 2822 EST clones ofM. haptotylum. Despite the fact that the knobs and mycelium were grown in the same medium, there were substantial differences in the patterns of genes expressed in the two cell types. In total, 23·3 % (657 of 2822) of the putative genes were differentially expressed in knobs versus mycelium. Several of these genes displayed sequence similarities to genes known to be involved in regulating morphogenesis and cell polarity in fungi. Among them were several putative homologues for small GTPases, such asrho1,rac1andras1, and a rho GDP dissociation inhibitor (rdi1). Several homologues to genes involved in stress response, protein synthesis and protein degradation, transcription, and carbon metabolism were also differentially expressed. In the last category, a glycogen phosphorylase (gph1) gene homologue, one of the most upregulated genes in the knobs as compared to mycelium, was characterized. A number of the genes that were differentially expressed in trap cells are also known to be regulated during the development of infection structures in plant-pathogenic fungi. Among them, agas1(mas3) gene homologue (designatedgks1), which is specifically expressed in appressoria of the rice blast fungus, was characterized.

Global gene expression in neuroendocrine tumors from patients with the MEN1 syndrome

Background

Multiple Endocrine Neoplasia type 1 (MEN1, OMIM 131100) is an autosomal dominant disorder characterized by endocrine tumors of the parathyroids, pancreatic islets and pituitary. The disease is caused by the functional loss of the tumor suppressor protein menin, coded by the MEN1 gene. The protein sequence has no significant homology to known consensus motifs. In vitro studies have shown menin binding to JunD, Pem, Smad3, NF-kappaB, nm23H1, and RPA2 proteins. However, none of these binding studies have led to a convincing theory of how loss-of-menin leads to neoplasia.

Results

Global gene expression studies on eight neuroendocrine tumors from MEN1 patients and 4 normal islet controls was performed utilizing Affymetrix U95Av2 chips. Overall hierarchical clustering placed all tumors in one group separate from the group of normal islets. Within the group of tumors, those of the same type were mostly clustered together. The clustering analysis also revealed 19 apoptosis-related genes that were under-expressed in the group of tumors. There were 193 genes that were increased/decreased by at least 2-fold in the tumors relative to the normal islets and that had a t-test significance value of p < = 0.005. Forty-five of these genes were increased and 148 were decreased in the tumors relative to the controls. One hundred and four of the genes could be classified as being involved in cell growth, cell death, or signal transduction. The results from 11 genes were selected for validation by quantitative RT-PCR. The average correlation coefficient was 0.655 (range 0.235–0.964).

Conclusion

This is the first analysis of global gene expression in MEN1-associated neuroendocrine tumors. Many genes were identified which were differentially expressed in neuroendocrine tumors arising in patients with the MEN1 syndrome, as compared with normal human islet cells. The expression of a group of apoptosis-related genes was significantly suppressed, suggesting that these genes may play crucial roles in tumorigenesis in this syndrome. We identified a number of genes which are attractive candidates for further investigation into the mechanisms by which menin loss causes tumors in pancreatic islets. Of particular interest are: FGF9 which may stimulate the growth of prostate cancer, brain cancer and endometrium; and IER3 (IEX-1), PHLDA2 (TSS3), IAPP (amylin), and SST, all of which may play roles in apoptosis.

Mechanical stretch is a highly selective regulator of gene expression in human bladder smooth muscle cells

Application of mechanical stimuli has been shown to alter gene expression in bladder smooth muscle cells (SMC). To date, only a limited number of "stretch-responsive" genes in this cell type have been reported. We employed oligonucleotide arrays to identify stretch-sensitive genes in primary culture human bladder SMC subjected to repetitive mechanical stimulation for 4 h. Differential gene expression between stretched and nonstretched cells was assessed using Significance Analysis of Microarrays (SAM). Expression of 20 out of 11,731 expressed genes ( approximately 0.17%) was altered >2-fold following stretch, with 19 genes induced and one gene (FGF-9) repressed. Using real-time RT-PCR, we tested independently the responsiveness of 15 genes to stretch and to platelet-derived growth factor-BB (PDGF-BB), another hypertrophic stimulus for bladder SMC. In response to both stimuli, expression of 13 genes increased, 1 gene (FGF-9) decreased, and 1 gene was unchanged. Six transcripts (HB-EGF, BMP-2, COX-2, LIF, PAR-2, and FGF-9) were evaluated using an ex vivo rat model of bladder distension. HB-EGF, BMP-2, COX-2, LIF, and PAR-2 increased with bladder stretch ex vivo, whereas FGF-9 decreased, consistent with expression changes observed in vitro. In silico analysis of microarray data using the FIRED algorithm identified c-jun, AP-1, ATF-2, and neurofibromin-1 (NF-1) as potential transcriptional mediators of stretch signals. Furthermore, the promoters of 9 of 13 stretch-responsive genes contained AP-1 binding sites. These observations identify stretch as a highly selective regulator of gene expression in bladder SMC. Moreover, they suggest that mechanical and growth factor signals converge on common transcriptional regulators that include members of the AP-1 family.

Use of cDNA microarrays to monitor transcriptional responses of the chestnut blight fungus Cryphonectria parasitica to infection by virulence-attenuating hypoviruses

Hypoviruses are a family of cytoplasmically replicating RNA viruses of the chestnut blight fungus Cryphonectria parasitica. Members of this mycovirus family persistently alter virulence (hypovirulence) and related fungal developmental processes, including asexual and sexual sporulation. In order to gain a better understanding of the molecular basis for these changes, we have developed a C. parasitica cDNA microarray to monitor global transcriptional responses to hypovirus infection. In this report, a spotted DNA microarray representing approximately 2,200 C. parasitica genes was used to monitor changes in the transcriptional profile after infection by the prototypic hypovirus CHV1-EP713. Altered transcript abundance was identified for 295 clones (13.4% of the 2,200 unique cDNAs) as a result of CHV1-EP713 infection-132 up-regulated and 163 down-regulated. In comparison, less than 20 specific C. parasitica genes were previously identified by Northern analysis and mRNA differential display as being responsive to hypovirus infection. A 93% validation rate was achieved between real-time reverse transcription-PCR results and microarray predictions. Differentially expressed genes represented a broad spectrum of biological functions, including stress responses, carbon metabolism, and transcriptional regulation. These findings are consistent with the view that infection by a 12.7-kbp hypovirus RNA results in a persistent reprogramming of a significant portion of the C. parasitica transcriptome. The potential impact of microarray studies on current and future efforts to establish links between hypovirus-mediated changes in cellular gene expression and phenotypes is discussed.

Microarray analysis using bioinformatics analysis audit trails (BAATs)

Bioinformatics analysis plays an integrative role in genomics and functional genomics. The ability to conduct quality managed, hypothesis-driven bioinformatics analysis with the plethora of data available is mandatory. Biological interpretation of this data is dependent on versions of databases, programs and the parameters used. Thus, tracking and auditing the analyses process is important. This paper outlines what we term Bioinformatics Analysis Audit Trails (BAATs) and describes YABI, a bioinformatics environment that implements BAATs. YABI can incorporate most bioinformatics tools within the same environment, making it a valuable resource.