DNA binding proteins are important in various cellular processes including transcriptional regulation, recombination, genome rearrangements, and DNA replication, repair, and modification. The interactions between transcription factors and their DNA binding sites are of particular interest because they regulate gene expression required for progression through the cell cycle, through differentiation, and in response to environmental stimuli. However, only a small

From: Methods in Molecular Biology, vol. 338: Gene Mapping, Discovery, and Expression: Methods and Protocols Edited by: M. Bina © Humana Press Inc., Totowa, NJ

handful of sequence-specific transcription factors have been characterized well enough such that all the sequences that they can (and just as importantly cannot) bind are known. This sparseness of binding site sequence data is highly problematic because these sparse datasets are frequently used to search for genomic occurrences of these sites, with many false-positive and false-negative binding sites being predicted. Earlier technologies aimed at characterizing DNA-protein interactions have been time-consuming and not highly scalable, and micro-array readout of chromatin immunoprecipitations (ChIP-chip, or genome-wide location analysis) requires that the given DNA binding protein be bound to its target sites when the cells are fixed (1).

Recent advances in genomics and proteomics have set the stage for rapid, high-throughput characterization of DNA binding proteins. Overexpression and purification of DNA binding proteins of interest is a familiar technique that has been used to allow characterization of these proteins using various traditional biochemical techniques. Now, most researchers also have access to DNA micro-arraying facilities, if not at their own institution, then through another institution that provides microarraying services for a fee. Likewise, DNA microarray scanners and glass slides for printing of the microarrays are readily available.

We recently developed an in vitro DNA microarray technology, which we term protein binding microarrays (PBMs), for characterization of the sequence specificities of DNA-protein interactions. This technology allows the in vitro binding specificities of individual DNA binding proteins to be determined in a single day, by assaying the sequence-specific binding of a given DNA binding protein directly to double-stranded DNA microarrays spotted with a large number of potential DNA binding sites. Specifically, a DNA binding protein of interest is expressed with an epitope tag, purified, and then bound directly to triplicate double-stranded DNA microarrays. The protein-bound microarrays are then washed to remove any nonspecifically bound protein and labeled with a fluo-rophore-conjugated antibody specific for the epitope tag. In order to normalize the PBM data by relative DNA concentration, separate triplicate microarrays from the same print run are stained with the dye SYBR Green I, which is specific for double-stranded DNA (Figs. 1 and 2). The sequences corresponding to the significantly bound spots (Fig. 3A) are analyzed with a motif prediction tool in order to identify the DNA binding site motif for the given DNA binding protein (Fig. 3B).

This PBM technology will likely aid in the annotation of many regulatory proteins whose DNA binding specificities have not been characterized and in the construction of gene regulatory networks. For example, binding site data derived from PBMs on transcription factors from the yeast Saccharomyces cere-visiae, using whole-genome S. cerevisiae intergenic microarrays, corresponded well with binding site specificities determined from ChIP-chip. Furthermore, double-stranded DNA microarrays bind epitope-tagged TF to dsDNA microarrays label with fluorop ho re-tagged anti (epitope) antibody scan triplicate microarrays calculate normalized PBM data Fig. 1. Schema of protein binding microarray experiments. (Reproduced from ref. 2 with permission from Nature Publishing Group.)

comparative sequence analysis of the PBM-derived binding sites indicated that many of the sites identified as bound in PBMs, including some not identified as bound in the ChIP-chip data, are highly conserved in other sensu stricto yeast genomes and thus are likely to be functional in vivo binding sites that may be utilized in a condition-specific manner (2).

0 0

Post a comment