One of the main beliefs in biology today is that there is a relationship between genes that share common promoter motifs and their expression. It makes sense that if genes share common regulatory elements that they would be turned on and off under similar conditions – and thus would have similar microarray expression profiles. A recent Bioinformatics paper published in October 2009 suggests that the relationship might be a little more complex than one might think initially. Antonio Marco and his colleagues at the Biodesign Institute at Arizona State University found that while genes with shared transcription factor binding sites (TFBS) tend to share common expression profiles, genes with common expression profiles do not necessarily seem to share common TFBS. Their research focused on Drosophila melanogaster and Saccharomyces cerevisiae.
They analyzed data from the REDfly TFBS database <http://redfly.ccr.buffalo.edu/> as well as data for six specific transcription factors involved in Drosophila development. In the analysis of 2485 gene pairs constructed from 71 REDfly genes, they found that genes sharing at least one TFBS showed significantly higher expression profile correlation than genes that did not share a common TFBS (P << 0.01). They found similar results in yeast – as the number of shared upstream motifs increased for gene sets, there was a greater coexpression correlation.
The intriguing finding in the paper, though, was when they turned the question around and asked if genes that show expression correlation share common promoter motifs. In flies, although gene pairs with similar coexpression showed slightly more likelihood of having common motifs compared to randomly chosen gene pairs, this was not statistically significant. They conclude that “Therefore, the use of expression correlation to find genes via comparative genomics of shared motifs may not be very productive.” In yeast, there is a different story. If yeast genes have a common expression profile, they are likely to share at least one common TFBS.
So what is going on here? Certainly one significant difference is that Drosophila is a multicellular organism with many different tissue types, while yeast is a single celled eukaryote. Unlike yeast, Drosophila will have many genes that are related to development and also many genes that are expressed in a tissue specific manner. Thus, even though some genes might share common promoter motifs, the genes themselves might not be available for transcription due to tissue specific differences in chromatin structure. In fact, when they examined expression profiles for the first 6 hours of Drosophila development, genes that did have expression correlation indeed were enriched for common TFBS. During this time of Drosophila development, there is no tissue differentiation.
The authors also demonstrated that genes which share the same Gene Ontology <http://www.geneontology.org/> biological process are likely to share common TFBS. These processes are often expressed in tissue specific ways and are spatially distinct within the fly. Thus, the Gene Ontology incorporates spatial organization and tissue specific information which contributes to correlation in gene expression.
The rather obvious take home lesson from this paper is that gene expression in multicellular organisms is complex. One must look at tissue specific expression patterns in order to begin to unravel the ways in which genes are regulated. It will be interesting to see how the relationship between gene expression and promoter motifs is studied in other species as a result of this paper.
REFERENCE
“Relationship between gene co-expression and sharing of transcription factor binding sites in Drosophila melanogaster” by Antonio Marco, Charlotte Konikoff, Timothy L. Karr and Sudhir Kumar; Bioinformatics 25(19):2473-2477 (2009)
Gene Expression and Transcription Regulation
One of the main beliefs in biology today is that there is a relationship between genes that share common promoter motifs and their expression. It makes sense that if genes share common regulatory elements that they would be turned on and off under similar conditions – and thus would have similar microarray expression profiles. A recent Bioinformatics paper published in October 2009 suggests that the relationship might be a little more complex than one might think initially. Antonio Marco and his colleagues at the Biodesign Institute at Arizona State University found that while genes with shared transcription factor binding sites (TFBS) tend to share common expression profiles, genes with common expression profiles do not necessarily seem to share common TFBS. Their research focused on Drosophila melanogaster and Saccharomyces cerevisiae.
They analyzed data from the REDfly TFBS database <http://redfly.ccr.buffalo.edu/> as well as data for six specific transcription factors involved in Drosophila development. In the analysis of 2485 gene pairs constructed from 71 REDfly genes, they found that genes sharing at least one TFBS showed significantly higher expression profile correlation than genes that did not share a common TFBS (P << 0.01). They found similar results in yeast – as the number of shared upstream motifs increased for gene sets, there was a greater coexpression correlation.
The intriguing finding in the paper, though, was when they turned the question around and asked if genes that show expression correlation share common promoter motifs. In flies, although gene pairs with similar coexpression showed slightly more likelihood of having common motifs compared to randomly chosen gene pairs, this was not statistically significant. They conclude that “Therefore, the use of expression correlation to find genes via comparative genomics of shared motifs may not be very productive.” In yeast, there is a different story. If yeast genes have a common expression profile, they are likely to share at least one common TFBS.
So what is going on here? Certainly one significant difference is that Drosophila is a multicellular organism with many different tissue types, while yeast is a single celled eukaryote. Unlike yeast, Drosophila will have many genes that are related to development and also many genes that are expressed in a tissue specific manner. Thus, even though some genes might share common promoter motifs, the genes themselves might not be available for transcription due to tissue specific differences in chromatin structure. In fact, when they examined expression profiles for the first 6 hours of Drosophila development, genes that did have expression correlation indeed were enriched for common TFBS. During this time of Drosophila development, there is no tissue differentiation.
The authors also demonstrated that genes which share the same Gene Ontology <http://www.geneontology.org/> biological process are likely to share common TFBS. These processes are often expressed in tissue specific ways and are spatially distinct within the fly. Thus, the Gene Ontology incorporates spatial organization and tissue specific information which contributes to correlation in gene expression.
The rather obvious take home lesson from this paper is that gene expression in multicellular organisms is complex. One must look at tissue specific expression patterns in order to begin to unravel the ways in which genes are regulated. It will be interesting to see how the relationship between gene expression and promoter motifs is studied in other species as a result of this paper.
REFERENCE
“Relationship between gene co-expression and sharing of transcription factor binding sites in Drosophila melanogaster” by Antonio Marco, Charlotte Konikoff, Timothy L. Karr and Sudhir Kumar; Bioinformatics 25(19):2473-2477 (2009)