Global gene expression shift during the transition from early neural development to late neuronal differentiation in...

Regulation of transcription is one of the mechanisms involved in animal development, directing changes in patterning and cell fate specification. Large temporal data series, based on microarrays across the life cycle of the fly Drosophila melanogaster, revealed the existence of groups of genes which expression increases or decreases temporally correlated during the life cycle. These groups of genes are enriched in different biological functions. Here, instead of searching for temporal coincidence in gene expression using the entire genome expression data, we searched for temporal coincidence in gene expression only within predefined catalogues of functionally related genes and investigated whether a catalogue's expression profile can be used to generate larger catalogues, enriched in genes necessary for the same function. We analyzed the expression profiles from genes already associated with early neurodevelopment and late neurodifferentiation, at embryonic stages 16 and 17 of Drosophila life cycle. We hypothesized that during this interval we would find global downregulation of genes important for early neuronal development together with global upregulation of genes necessary for the final differentiation of neurons. Our results were consistent with this hypothesis. We then investigated if the expression profile of gene catalogues representing particular processes of neural development matched the temporal sequence along which these processes occur. The profiles of genes involved in patterning, neurogenesis, axogenesis or synaptic transmission matched the prediction, with largest transcript values at the time when the corresponding biological process takes place in the embryo. Furthermore, we obtained catalogues enriched in genes involved in temporally matching functions by performing a genome-wide systematic search for genes with their highest expression levels at the corresponding embryonic intervals. These findings imply the use of gene expression data in combination with known biological information to predict the involvement of functionally uncharacterized genes in particular biological events.