ChIP-on-chip technology unites chromatin immunoprecipitation with microarray platforms in order to reveal protein/DNA interactions. By allowing the identification of protein-binding sites on DNA on a global genomic basis, it offers a new path to pharmaceutical discovery.
ChIP-on-chip pairs chromatin immunoprecipitation (ChIP) with glass slide microarrays (chip) to analyze how regulatory proteins interact with the genome of living cells. Still largely an academic protocol, it consists of immunoprecipitation of chromatin, using a specific antibody directed against a regulatory protein or peptide. This results in an enrichment of the DNA in the regions surrounding the target protein. The DNA that is brought down with the antibody can then be queried with a DNA microarray.
Several biotech firms and academic groups are pushing the envelope to connect this form of gene regulation to disease processes. These include Aviva Systems Biology; The Institute for Diabetes, Obesity, and Metabolism at Pennsylvania School of Medicine; The London Regional Cancer Program; Genetics and Genome Biology Program, the Toronto Hospital for Sick Children; The Bioinformatics Program at the University of Memphis; and The Genome Center at the University of California.
Finally, several companies offer chip technologies adaptable to this function.
Activation of Regulatory Genes
“There are a number of different ways to look at changes in the genome,” according to Bekim Sadikovic, Ph.D., postdoctoral fellow at the University of Toronto. He explains that ChIP-on-chip technology lends itself particularly well to the analysis of epigenetic changes, that is, alteration of molecular instructions that regulate both gene expression and chromatin organization. The pathway through which these events come about requires histone modification and heritable DNA-methylation changes.
Dr. Sadikovic describes how the dynamic epigenetic patterning of DNA methylation plays a role in differential gene expression, chromatin organization, X-chromosome inactivation, and genome imprinting, as has been known for many years.
Furthermore, it has long been recognized that environmental toxicants such as benzo(a)pyrene can profoundly affect gene expression, but until recently the mechanism by which these cancer-causing changes occurred was not well understood.
Dr. Sadikovic and his mentors—Maria Zelinska, Ph.D., and Jeremy Squire, Ph.D., at the Ontario Cancer Research Institute and the Toronto Hospital for Sick Children—have performed a number of studies in which they investigated these changes using high-resolution microarray platforms. In order to target regions of the genome involved in epigenetic changes, methylated genomic DNA is enriched using immunoprecipitation with a powerful antibody directed against 5-methyl cytosine.
This DNA is hybridized to genomic microarray platforms allowing precise mapping of genome-wide epigenetic profiles. In some cases stretches as short as 35 nucleotides are involved in regulation of an osteosarcoma-producing phenotype.
“It’s possible to localize these cancer-related regions of the genome using a restriction analysis, but this depends on serendipitously having the appropriate sequences next to the gene you are looking for; with ChIP-on-chip technology this isn’t necessary, this gives us a much greater level of confidence and specificity,” Dr. Sadikovic states. “Moreover, the level of resolution and the subsequent yield of information is different by orders of magnitude.”
Dr. Sadikovic and his coworkers have also performed epi-toxicogenomic studies in which antibodies against acetylated histones were used to immunoprecipitate chromatin extracted from cells exposed to benzo(a)pyrene. Then Affymetrix promoter tiling microarrays were probed to investigate the relationship between such environmental exposure and global gene-specific changes in histone acetylation.
“It’s important to carry out replicate studies in order to lower noise and the chance for erroneous signals,” Dr. Sadikovic notes. “Repeatability has bedeviled the discovery process in genomics, which may account for the failure of genome analysis to uncover new biomarkers.”
The group’s studies have turned up a number of genes altered in their expression by polycyclic hydrocarbons, including those affecting proliferation, apoptosis, and cell-cycle regulation that may serve as targets for future anticancer drugs.
“Aviva Systems Biology has developed a transcription network mapping system,” explains Jeff Falk, Ph.D., director of technology and business applications at Aviva. One of the functions of this platform is to provide transcriptional fingerprinting of hepatotoxicity regulatory networks.
For the last six years, Aviva has been building ChIP-based technologies for mapping genome-wide gene transcription regulation, and it has applied these technologies to profiling multiple tissues and key disease-related systems as well as mapping the response of liver specific genes to toxic substances.
“Now the initial pilot studies are generating a complete map of the human liver transcriptional networks,” Dr. Falk says. “This will provide a framework for identifying the critical pathways and biomarkers of toxicity.”
A major component of the company’s strategy is a multiepitope polyclonal antibody collection, which includes a complete set of antibodies for all human transcription factors. The company has developed a high-throughput, automated ChIP-on-chip screening capability that allows prevalidation of the antibodies directed against various regulatory proteins.
Dr. Falk argues that a stable of powerful polyclonal antibodies directed against putative marker proteins and peptides is absolutely essential for a successful ChIP- on-chip program, a view that he shares with other scientists.
Dr. Falk compares the Aviva platform with conventional toxicology testing in which compounds are initially screened and moved through preclinical and clinical evaluation, with painfully high attrition rates, resulting in a spectacular loss of time and great expense.
Aviva’s DNA/protein interaction mapping technologies focus on three variations of the traditional ChIP-on-chip platform, which uses antitranscription factor antibodies to identify genome-wide promoter/enhancer interactions, epigenetic modifications, and DNA-methylation sites. These include ChIP-DSL (DNA selection and ligation), a promoter array technology that is faster and much more sensitive; ChIP-Seq, known as sequencing-based profiling of transcriptional fingerprinting interactions; and ChIP-qPCR, for validation of the identified binding interactions at specific targets.
“DSL refers to the improvements made to the ChIP-on-chip procedure. In typical ChIP-on-chip reactions, the immunoprecipitated DNA products undergo general blunt-end ligation of linkers and subsequent PCR amplification,” Dr. Falk explains. “This is an inefficient process since blunt-end ligation only attaches linkers to a small percentage of the actual immunoprecipitated DNA, so the background is very high.”
With DSL, there is a specific amplification process. The immunoprecipitated DNA is first hybridized to a pool of 20,000 oligo pairs that correspond directly to the 20,000 arrayed promoter oligos. The hybridized oligo pair is then ligated into a 40 base template using Taq DNA Ligase, which only occurs if there is a 100% direct match. As a result, the specificity is dramatically increased, since only the immunoprecipitated sequences present in the promoter array are amplified. In addition, the reactions are more efficient so that DSL requires less starting material.
Among the many possible applications for products in the clinic, the Aviva chip technologies are being applied to a variety of cancer prognostics as well as hepatotoxicity biomarkers.
Identification of Binding Sites
“We’re using the ChIP-on-chip approach to analyze regulatory elements in cancer cell lines,” says Victor Jin, Ph.D., of the bioinformatics program at the University of Memphis. “We have developed a computational genomics approach (ChIPModules) and a motif discovery approach (ChIPMotifs) to mine the ChIP-on-chip data.”
Dr. Jin focuses on data-mining algorithm development and stresses the synergism that he and his collaborators have built by combining the experimental and computational approaches in the elucidation of cis-acting transcriptional regulatory elements. These include promoters, enhancers, and repressor elements. At present the ENCODE consortium (of which Dr. Jin and his colleagues are participants) seeks to develop a detailed map of all the transcribed regions of the human genome. This work calls for the use of microarray-based analysis of human RNAs. They stress that the experimental data, covering a spectrum of information, must be interpreted through the interplay of computational strategies, with the ultimate goal of validation of biological mechanisms.
It is important to recognize the advances that have been made in the past decade using ChIP-on-chip technology applied to the understanding of regulatory control. As the technology has improved, the number of searchable target-binding sites on microarray chips has grown more complex in nature, and now constitutes a more detailed and realistic model of the actual organism and its web of interactions.
Dr. Jin and his colleagues assert that the collection and visualization of data sets from multiple independent research groups will comprise a robust treasure trove, driving a deeper understanding of the regulatory process. For instance, many studies on the OCT4 regulatory have shown that it is the key to maintaining pluripotency and self-renewal of human stem cells, germ cells, and tumor cells. Moreover, it has many targets in common with SRY, an important masculinizing gene on the Y chromosome. In the future, these genes may constitute important targets for drug development.
“For example, we have used our ChIPModules approach to identify transcriptional regulatory modules, one of which was validated by using ChIP-on-chip with arrays containing about 14,000 human promoters,” Dr. Jin notes.
Epigenetic Control of Genes
Foxa2 transcription factor is a well-known regulator of bile homeostasis in the liver and an important regulator of gluconeogenesis in this organ. Klaus Kaestner, Ph.D., professor of genetics at the University of Pennsylvania School of Medicine, and his collaborators have demonstrated these important facts using ChIP-on-chip analysis and cell-type-specific gene-ablation studies. According to Dr. Kaestner, earlier studies focused on individual binding and potential interaction between known hepatic regulators but did not use computational tools to identify additional transcriptional factors.
“We start with liver tissues in which we crosslink the transcriptional regulators to the chromatin DNA and employ a specific antibody against the transcription factor of interest to precipitate the bound DNA sequences,” Dr. Kaestner explains. “We then hybridize the purified DNA to a microarray that contains 35,000 mouse promoters and enhancers.”
With the aid of cell-type-specific gene ablation, Dr. Kaestner’s team has shown that Foxa2 regulates whole families of genes involved in hepatic bile homeostasis. Previously, studies using models in which the gene was overexpressed yielded conflicting results.
Dr. Kaestner’s studies demonstrate the power of ChIP-on-chip technology including its high resolution and ability to provide extremely specific information regarding the properties of the regulatory genes and their surrounding cis-active companions. As these studies move forward, they are providing a more detailed picture of differential gene expression and also the liver’s response to environmental toxins. A striking observation made by Dr. Kaestner and his coworkers concerns the ability of Foxa2 to bind to DNA even when the consensus sequence is not a strong match. Genes containing a cis-regulatory module with a weak consensus site are much more liver specific than those genes with a strong consensus site.
“Examining the chromatin of pancreatic islets isolated from normal or diabetic pancreases allows side-by-side comparisons of histone modifications with the eventual aim of determining how lifestyle changes predispose the individual to type 2 diabetes,” Dr. Kaestney adds.
This technology represents an improvement on the early fishing expeditions of the genome and proteome for the exploration of potential therapeutic targets. Its sensitivity, specificity, and ability to uncover previously unknown regulatory networks suggest that within the near future specific models for drug development will come to light.
On the other hand, one needs to be realistic concerning the time frames and the difficulty of applying new technologies to a process that becomes ever more challenging. The draconian elimination process for new drugs and the long road through the pipeline will continue to be a constant challenge for biotech and pharma companies as they attempt to apply this platform.