Current Genomics - Volume 3, Issue 1, 2002

Large scale analysis of gene expression using DNA arrays shows great promise in enhancing our understanding of how the genome constructs the brain. This technology has been employed to analyze gene expression variations in the brain resulting from genetic polymorphisms, drug abuse, aging, and the neuropsychiatric disorders such as schizophrenia and Alzheimer's disease. Although the data from DNA array analysis is necessarily noisy, thoughtfully designed experiments can give useful insights. In addition, as the use of DNA array technology grows, replicated experiments will provide even greater confidence in drawn conclusions, giving optimism that these methods will yield profound insights into the genomics of the brain.

We review a series of statistical issues that arise in the design of a gene-expression array experiments in the statistical assessment of which genes are differentially expressed and in the use of array data to understand transcription regulation and prediction of cell characteristics.

In the post-genomic era, microarray technology will have a major impact on our understanding of complex gene expression patterns and circuit function in the brain. Due to the phenotypic and transcript complexity of the brain, transcriptome profiling data is multifaceted, and is best interpreted in the context of the cellular diversity of the studied brain region. In the near future, massive microarray datasets will be placed into a biological context. To separate biologically significant transcriptome differences from experimental noise, verification of microarray data and anatomical localization of expression changes to neuronal subpopulations will continue to be an integral part of microarray experiments.

An estimated 9 to 22 % of the pediatric age group are affected by the extremely broad range of childhood neurological disorders, which includes autism and other pervasive developmental disorders (PDD). As with most all of the neurodevelopmental disorders, the PDD spectrum disorders demonstrate a tremendous degree of clinical, biochemical, and genetic heterogeneity. In the great majority of cases, the underlying molecular defects are not known and both diagnosis and appropriate treatment remain challenging. Gene expression profiling using DNA microarrays is a useful tool to characterize cellular samples from patients with childhood neurological disorders. There are two principal goals in the characterization of transcriptional changes in these disorders: 1) Accurate molecular diagnosis of disorders in the absence of clear clinical or biochemical diagnostic markers. Additionally, analysis of transcriptional patterns can be used to generate sub-classifications of individual disorders. 2) Identification of cellular systems that are disrupted in the disorder. Changes in the expression of individual genes as well as functionally related groups of genes can delineate biochemical pathways that are perturbed in a disorder. Both of these goals are ultimately directed at the development of effective therapeutic strategies. This review describes recent advances in these areas, discussing issues relevant to the study of gene expression in pediatric neurodevelopmental disorders.

Cancer is the result of cumulative multiple genetic mutations, which result in the activation of oncogenes and / or the inactivation of tumor suppressor genes. Current theories of malignant transformation postulate that the development of primary brain tumors is the consequence of the sequential accumulation of multiple genetic alterations in brain cells, each of which contributes to the induction of a progressively more malignant phenotype. Traditionally, our understanding of brain tumor biology has come from the study of single specific genes or chromosome regions at a time. Although several genetic aberrations and gene expression changes have been identified using such focused techniques, the traditional methods of cancer research are tediously slow and provide limited insight into the global gene expression patterns that occur during the malignant transformation, development, and progression of primary brain cancers. Because hundreds of genes may be simultaneously involved in the mechanisms of carcinogenesis, new genomic high-throughput technologies have recently come into the forefront of cancer research. These technologies, which include cDNA and oligonucleotide microarrays (gene chips) and tissue microarray (tissue chip) techniques, may considerably facilitate the molecular profiling of human tumors. Such molecular fingerprinting of malignant gliomas, for example, may lead to advances in diagnosis, prognosis, and design of novel therapeutic approaches that could improve the clinical outcome of patients suffering from this rapidly fatal disease. This review will describe the principles of microarray technology as applied to human brain tumor research, summarize its use and limitations in identifying brain tumor-associated genes thus far, and speculate on the future applications and clinical ramifications of such genomic large-scale screening techniques in neuro-oncology.

In order to identify a transcription component of programmed cell death (PCD), this review compiles and cross-compares genome-scale gene expression analyses of several models of apoptosis. Large-scale transcription profiling studies to date have utilized diverse technologies such as cDNA arrays and serial analysis of gene expression (SAGE). Despite the varying technology platforms, commonly regulated genes were identified and fell into three major categories those related to the regulation of cell cycle, inflammatory responses, and oxidative stress. The observation that common markers were regulated across different models of apoptosis demonstrates the utility of genome-scale gene expression analysis for cell death pathway mining.