How to find the mechanism on differential gene expression

[redhawk421001]

We have been doing research on the differential expression of one gene between two cell lines (A and B cell lines), with high expression in A cell line and no expression in B cell line. We found that ~250 bp in the promoter of the gene has its activity in A cell line, but no activity in B cell line. So we think the differential profiling of transcriptional factors between A band B cell lines decide whether the gene is expressed. However, I don’t know to find the transcriptional factors regulating the expression of the gene. Please give me some suggestions.

 

[Monique]

What do you mean with that "250 bp of the promoter of the gene has its activity in A cell line"? How did you determine that?

 

[redhawk421001]

What do you mean with that "250 bp of the promoter of the gene has its activity in A cell line"? How did you determine that?

We cloned different length segments of the promoter, and linked the segments to luciferase reporter gene in expression vector, and transfected the vectors into A and B cell lines. Finally, we found that the segments with the length of >=250 bp have high activity in A cell line and no activity in B cell line.

 

[Ygggdrasil]

You could use a computational approach to narrow your approach and/or identify candidate transcription factors. Many classes of transcription factors have certain sequence binding preferences. You'd probably be able to find a program somewhere where you can input a DNA sequence and it looks for potential transcription factor binding sites (for example, here is one such program: http://the_brain.bwh.harvard.edu/uniprobe/index.php).

 

[Monique]

As Ygggdrasil suggests, you can try the bioinformatics approach to see whether there is a consensus TF binding site in the 250 bp sequence. Just search google or the literature for some good prediction programs.

The biochemical approach would be to crosslink the bound TFs to the DNA, perform a chromatin IP (ChIP) and characterize the proteins that are pulled down. For instance: http://www.ncbi.nlm.nih.gov/pubmed/12054904

 

[redhawk421001]

You could use a computational approach to narrow your approach and/or identify candidate transcription factors. Many classes of transcription factors have certain sequence binding preferences. You'd probably be able to find a program somewhere where you can input a DNA sequence and it looks for potential transcription factor binding sites (for example, here is one such program: http://the_brain.bwh.harvard.edu/uniprobe/index.php).

We have used two programs to predict the transcription factor binding sites with setting the similarity score as 85%, and found that 48 TFs can bind to the DNA sequence. Theremore, we worry whether there are other cofactor not directly binding to the promoter, that affect the expression of the gene.