Genomewide analysis of aryl hydrocarbon receptor binding targets reveals an extensive array of gene clusters that control morphogenetic and developmental programs.

BACKGROUND: The vertebrate aryl hydrocarbon receptor (AHR) is a ligand-activated transcription factor that regulates cellular responses to environmental polycyclic and halogenated compounds. The naive receptor is believed to reside in an inactive cytosolic complex that translocates to the nucleus and induces transcription of xenobiotic detoxification genes after activation by ligand. OBJECTIVES: We conducted an integrative genomewide analysis of AHR gene targets in mouse hepatoma cells and determined whether AHR regulatory functions may take place in the absence of an exogenous ligand. METHODS: The network of AHR-binding targets in the mouse genome was mapped through a multipronged approach involving chromatin immunoprecipitation/chip and global gene expression signatures. The findings were integrated into a prior functional knowledge base from Gene Ontology, interaction networks, Kyoto Encyclopedia of Genes and Genomes pathways, sequence motif analysis, and literature molecular concepts. RESULTS: We found the naive receptor in unstimulated cells bound to an extensive array of gene clusters with functions in regulation of gene expression, differentiation, and pattern specification, connecting multiple morphogenetic and developmental programs. Activation by the ligand displaced the receptor from some of these targets toward sites in the promoters of xenobiotic metabolism genes. CONCLUSIONS: The vertebrate AHR appears to possess unsuspected regulatory functions that may be potential targets of environmental injury.

Gene expression profiling of blood to predict the onset of leukemia.

No studies have tested the hypothesis that the onset of a disease can be predicted by gene expression profiling. The AKR/J mouse strain, which spontaneously develops acute T cell lymphatic leukemia, was used to implement a novel strategy to generate global gene expression profiles of WBCs at different time points. The experimental approach was bias free because it was unknown as to which individuals in the mouse population would eventually develop the disease. Our results suggest that profiling WBC gene expression may be an effective means for the very early diagnosis of disease in humans.

Genome-wide analyses show that nuclear and cytoplasmic RNA levels are differentially affected by dioxin.

The aryl hydrocarbon receptor (AHR) mounts the body's main molecular defense against environmental toxicants by inducing a battery of genes encoding xenobiotic metabolizing proteins. The AHR is activated by polycyclic aromatic hydrocarbon toxicants, including the pervasive teratogen and carcinogen 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD or dioxin). The TCDD-activated AHR significantly changes the cytoplasmic mRNA levels of hundreds of genes, but little is known of the mechanism by which the activated AHR causes such a strong effect on global gene expression. We used high-density microarrays to compare nuclear and cytoplasmic RNA levels from untreated and TCDD-treated mouse embryonic fibroblasts (MEF) to test the hypotheses that (1) TCDD has a large impact on nuclear RNA levels and (2) that cytoplasmic RNA levels are dependent on nuclear RNA levels. We found that nuclear RNA levels are strongly affected by TCDD, and that nuclear and cytoplasmic RNA levels are only weakly correlated, indicating that other regulatory mechanisms are controlling cytoplasmic RNA levels. The nuclear RNAs most affected by TCDD encode proteins involved in nuclear RNA processing and transcription. We conclude that although the AHR regulates key xenobiotic metabolizing genes at the transcriptional level, a larger impact of the TCDD-activated AHR may be at post-transcriptional levels.

A new method to remove hybridization bias for interspecies comparison of global gene expression profiles uncovers an association between mRNA sequence divergence and differential gene expression in Xenopus.

The recent sequencing of a large number of Xenopus tropicalis expressed sequences has allowed development of a high-throughput approach to study Xenopus global RNA gene expression. We examined the global gene expression similarities and differences between the historically significant Xenopus laevis model system and the increasingly used X.tropicalis model system and assessed whether an X.tropicalis microarray platform can be used for X.laevis. These closely related species were also used to investigate a more general question: is there an association between mRNA sequence divergence and differences in gene expression levels? We carried out a comprehensive comparison of global gene expression profiles using microarrays of different tissues and developmental stages of X.laevis and X.tropicalis. We (i) show that the X.tropicalis probes provide an efficacious microarray platform for X.laevis, (ii) describe methods to compare interspecies mRNA profiles that correct differences in hybridization efficiency and (iii) show independently of hybridization bias that as mRNA sequence divergence increases between X.laevis and X.tropicalis differences in mRNA expression levels also increase.

Microarray results improve significantly as hybridization approaches equilibrium.

Dual-channel long oligonucleotide microarrays are in widespread use. Although much attention has been given to proper experimental design and analysis regarding long oligonucleotide microarrays, relatively little information is available concerning the optimization of protocols. We carried out a series of microarray experiments designed to investigate the effects of different levels of target concentration and hybridization times using a long oligonucleotide library. Based on principles developed from nucleic acid renaturation kinetics studies, we show that increasing the time of hybridization from 18 h to 42 h and 66 h, especially when lower than optimal concentrations of target were used, significantly improved the quality of the microarray results. Longer hybridization times significantly increased the number of spots detected, signal-to-noise ratios, and the number of differentially expressed genes and correlations among replicate arrays. We conclude that at 18 h of incubation, target-to-probe hybridization has not reached equilibrium and that a relatively high proportion of nonspecific hybridization occurs. This result is striking, given that most, if not all, published microarray protocols stipulate 8-24 h for hybridization. Using shorter than optimal hybridization times (i.e., not allowing hybridization to reach equilibrium) has the consequence of underestimating the fold change of differentially expressed genes and of missing less represented sequences.