The chromatin immunoprecipitation (ChIP) assay has recently been exploited as a powerful and versatile technique for probing protein-DNA interactions within the chromatin environment. In this method, intact cells are fixed with a reversible DNA-protein cross-linking agent (formaldehyde), and associated DNA is enriched by immunoprecipitating a target DNA binding protein. The bound DNA in the immune complexes is then used to identify that specific DNA binding protein's endogenous genomic targets. Nuclear factor κB (NF-κB) is a highly inducible transcription factor that controls genetic networks important for pathogen- or cytokine-induced inflammation, immune response, and cellular survival. In our studies of the genetic network under control of the inducible NF- κB transcription factor, we found that the conventional ChIP technique using a single formaldehyde cross-linking step did not reproducibly cross-link it to DNA. As a result, we have developed a novel ChIP assay using a two-step cross-linking procedure, incorporating N-hydroxysuccinimide (NHS)-ester-mediated protein-protein cross-linking prior to conventional DNA-protein cross-linking. We demonstrate that this technique is highly efficient, cross-linking virtually all NF-κB/Rel A into covalent complexes, resulting in quantitative and robust identification of inducible NF-κB family binding to a variety of validated NF-κB-dependent genomic targets. To demonstrate the general utility of this two-step cross-linking procedure, we performed enhanced capture of cytokine-inducible signal transducer and activator of transcription-3 (STAT3) binding to one of its known target genes. Our method represents a significant improvement in the efficiency of ChIP analysis in the study of endogenous targets for rare transcription factors.
Coordinate changes in gene expression are temporally and spatially controlled to produce a variety of phenotypic changes important in normal development and disease (1). In eukaryotic cells, genetic elements are maintained in dynamic chromatin structures. The first order of chromatin structure involves DNA compacted by wrapping around core nucleosomes, which in turn are organized into higher order domains through the action of H1 linker histones, architectural proteins, acetylases, and other scaffolding proteins (1). Recently, it has been appreciated that this chromatin structure exerts significant effects over coordinated gene expression by controlling promoter access by inducible sequence-specific DNA binding proteins (2,3). Of relevance, many hormonal or developmentally controlled stimuli activate signal transduction networks controlling families of low abundance, sequence-specific DNA binding proteins (4,5,6). Because of their influence on cellular phenotype, the binding and targets of inducible transcription factors have been intensely investigated using a variety of biochemical and genomic approaches.
The techniques that have been classically used to this end are electro-mobility shift assays (EMSAs) and DNase I footprinting assays. The EMSA involves binding a radiolabeled oligonucleotide to either a stimulated nuclear protein extract or a recombinant transcription factor and resolving the resulting complexes by native polyacrylamide gel electrophoresis (PAGE) (7). The problems now recognized with this approach are that (i) the DNA-protein complex does not have the complexity that is seen in cellulo due to the lack of relevant chromatin structure; (ii) the DNA is often of insufficient length to measure complex interactions (e.g., DNA bending or looping); and (iii) DNA binding sites identified by EMSA poorly predict the presence of actual binding sites in cellulo (8). DNase I footprinting is used to identify the region of DNA bound by a transcription factor by assessing nucleotides resistant to nuclease attack. Although this technique can be adapted to understanding protein binding within its native chromatin context (9,10), the method is not sensitive to weak or partial DNA binding, and the precise identity of the protecting complex cannot be elucidated.
The finding that the access of some sequence-specific transcription factors is tightly controlled by the chromatin environment of the target gene has stimulated development of other approaches to the analysis of protein-DNA interactions (2,3). Avoiding some of the shortfalls of EMS A and DNase I footprinting, the chromatin immunoprecipitation (ChIP) assay has been used to assay the binding of architectural DNA binding proteins, transcription factors, and members of the polymerase complex within native chromatin contexts (11,12). In this technique, intact cells are treated with formaldehyde (FA) to covalently link protein to DNA, the nucleoprotein complexes are then mechanically sheared and the cross-linked DNA-protein complexes enriched by immunoprecipitation. The retrieved complexes are then analyzed by PCR amplification to detect and quantify specific DNA targets.
The ChIP assay has been adopted as a powerful method for the analysis of proteins interacting within a native chromatin environment and is versatile enough for adaptation for a variety of purposes (13). This assay has been utilized in yeast (14), drosophila (15), tetrahymena (16), various mammalian cell lines, and even on whole mouse embryos (17,18) for the analysis of low abundance transcription factor binding. In addition to focused study of a single or group of genes, some have used this methodology for systematic promoter cloning (19,20) or identification of gene targets using promoter microarrays (14,21). ChIP has been used to determine the allele-specific transcription factor binding patterns (22) or measure long-range enhancer binding (23). One group has combined the ChIP technique with DNA footprinting methods (24), while another has devised ways of using ChIP for the analysis of RNA-protein interaction (25).
We sought to apply the ChIP methodology for network analysis of genes under control of the nuclear factor κB (NF-κB) transcription factor to study the precise timing of its binding in response to activating stimuli in a native chromatin context (6). NF-κB is a family of highly tumor necrosis factor (TNF)-inducible cytoplasmic transcription factors that controls genetic networks important in the hepatic acute phase response (26), immune response (27), atherosclerosis (28), and cellular survival pathways (29). Upon cytokine stimulation, the prototypical NF-κB complex, composed of 65 kDa Rel A • 50 kDa NF-κB1 heterodimers, enters the nucleus and binds to target gene promoters containing specific DNA binding sites. Although the mechanisms for NF-κB activation have been intensively investigated, relatively little is known about the global genomic targets of this transcription factor. Earlier we exploited a high-throughput analysis of expressing a tightly tetracycline-regulated dominant negative inhibitor to identify gene networks directly under NF-κB control (5,6,30). In spite of the dependence on NF-κB translocation for expression of interleukin 8 (IL-8) (31), IκBα (32), CXCL-20/Exodus-1 (33), TNFAIP-3/Naf-1 (6), and NF-κB2 (34), and the presence of high-affinity binding sites in the promoters, we were surprisingly unable to identify NF-κB bound to these targets using conventional ChIP assay.