Locating mammalian transcription factor binding sites: A survey of computational and experimental techniques

Abstract

Fields such as genomics and systems biology are built on the synergism between computational and experimental techniques. This type of synergism is especially important in accomplishing goals like identifying all functional transcription factor binding sites in vertebrate genomes. Precise detection of these elements is a prerequisite to deciphering the complex regulatory networks that direct tissue specific and lineage specific patterns of gene expression. This review summarizes approaches for in silico, in vitro, and in vivo identification of transcription factor binding sites. A variety of techniques useful for localized- and high-throughput analyses are discussed here, with emphasis on aspects of data generation and verification.

Remote Control of Gene Expression

“…The genomics revolution has yielded a massive amount of data with a conserved non-coding sequence estimated to be more than twice that of all coding sequences (27). A significant portion of the non-coding sequence comprises CRMs that control gene expression locally or remotely from their target genes.…”

“…Now, with multiple vertebrate genomes in hand, it is preferable to search for CRMs over extended genomic landscapes by comparing orthologous sequences followed by validations with wet-lab assays (26). Numerous internet-available algorithms are available to survey genomes for CRMs (27) and test them either in isolation or in the context of an artificial chromosome transgenic mouse. For example, VISTA (wwwgsd.lbl.gov/vista/) (28) and PIPMaker (bio.cse.psu.edu) (29) were used to compare a 1-megabase YAC containing the human interleukin cluster of genes on 5q31 with the homologous region in the mouse.…”

“…These studies are the basis for a large scale transgenic enhancer activity assay that will systematically test highly conserved blocks of sequence in transgenic mice (see enhancer.lbl.gov) and will provide critical leads for subsequent analysis in artificial chromosomes because the limitations again are that the CRMs are studied out of their native genomic context and may therefore only reveal a portion of information about the full spectrum of regulatory elements underlying a gene's expression profile. Individual cisacting regulatory elements within CRMs may also be identified by computational means though the reliability of finding these requires substantial insight into the transcription factor binding affinities and sequence plasticity across the element of interest (27,37). Limitations of comparative genomics for dis- covery of CRMs or individual cis regulatory elements include the inability to define coregulators that do not contact DNA directly and the insensitivity to detect species differences in gene expression control because of sequence differences in CRMs.…”

Chromatin Immunoprecipitation (ChIP)

“…Since its initial development, the ChIP method has been expanded to include ChIP-chip and ChIP-Seq assays. In ChIP studies designed to identify the locations of transcription factors and histones at a genome-wide scale (Ren et al 2000;Weinmann et al 2002;Hanlon and Lieb 2004;Elnitski et al 2006;Kim and Ren 2006), the precipitated DNA fragments were purified; amplified by whole-genome amplification, ligation-mediated (LM)-PCR, or other methods (see O'Geen et al 2006); and then annealed to tiled DNA arrays www.cshprotocols.org spanning an entire genome, a specific chromosome, or a large collection of putative promoter fragments (i.e., ChIP-chip). Detailed experimental strategies and procedures for these ChIP-chip (or ChIP-on-chip) methods can be found at the websites of companies that synthesize tiling arrays, including Agilent, NimbleGen, and Affymetrix.…”

Genomic tools for dissecting oncogenic transcriptional networks in human leukemia

“…45 Pattern detection methods search for recurring and overrepresented DNA motifs in specific subsets of sequences, such as genomic regions identified by ChIP-chip or ChIP-seq, as bound by a given transcription factor. 42,46 Pattern detection techniques can be used to identify putative DNA-binding motifs for transcription factors without previous knowledge of their DNA-binding properties and for the discovery of neighbor cooperating motifs specifically enriched in coregulated sets of genes.…”