Experimental considerations

Several considerations are important in the experimental design. First, due to the low yield of DNA in the ChIP it is important to determine the optimal number of cells to generate robust signals on the microarray. This can also vary depending on the labeling technique, as some protocols allow for less starting material by adding an amplification step to the labeling reaction (10, 11). Others choose to avoid amplification biases by increasing the number of cells (9, 12).

Another important aspect to consider is the use of antibodies or epitope tags in the immunoprecipitation step. In organisms where inserting an epitope directly at the gene locus is readily feasible, such as the case in yeast, this approach is often ideal since a standard well-proven protocol is available. This allows researchers to directly compare tagged versus untagged samples to determine transcription factor binding sites (see below). Generally the signals generated by this approach are robust as several copies of the tag are inserted into the genome and antibodies directed toward the tag are highly specific. One drawback, however, is that the protein is modified so the binding profile is not of a native protein. Thus, it is important to ensure that the modified factor is functional using genetic and/or biochemical tests.

In most organisms it is not possible to epitope tag a protein expressed at endogenous levels. Although it is often possible to transfect tagged constructs in these systems, this runs the caveat that the levels of protein will affect site occupancy and thus the final target lists. Thus, for these systems it is most desirable to use antibodies specific for the protein of interest. Antibodies for many factors are available, although their quality and/or characterization are often poor. Thus, it is important to ensure that antibodies can specifically immunoprecipitate a factor of interest. In addition, if a bona fide target of the transcription factor is known, then a standard ChIP using PCR confirmation should be performed to ensure the antibodies function in a ChIP assay.

One final important consideration is the selection of a reference channel so that enrichment for transcription factor binding in the ChIP can be determined. This reference or background sample is commonly either genomic DNA or a mock immunoprecipitation (IP). However, it can also be IP DNA from a non-induced sample that is compared with the induced state. The latter is obviously attractive when studying an inducible transcription factor where binding only occurs in one state. Transcription factors that translocate to the nucleus upon activation such as NF-kB and STAT1 are suitable for this type of approach (9). Whichever reference is chosen, the main consideration is an adequate signal-to-noise ratio so that the highest quality data is generated.

0 0

Post a comment