DISCUSSION
The development of techniques that permit the rapid comparative mapping of DHS between different cell types will greatly facilitate the study of transcriptional regulation in both normal and diseased cells. Recently published high-throughput techniques that map DHS sites using high-throughput sequencing (8,9) and genomic array tiling paths (7,11,13) have clear advantages of scale over more targeted approaches. Major disadvantages of these approaches include cost and lack of focus. This makes them less suitable for many laboratories that want to assess the chromatin accessibility of a number of defined or presumed regulatory elements in a range of cell types. A real-time-PCR-based approach to DHS mapping has, therefore, a number of potential advantages for researchers interested in specific regulatory questions at defined loci. Real-time PCR is relatively inexpensive compared with the large-scale techniques, and permits a rapid, focused DHS analysis of defined regions of DNA from multiple cell types. It also provides flexibility, as any genomic region can be analysed from the DNA library derived from the DNaseI-treated samples by designing further real-time PCR primer sets.
We have previously shown that our basic technique for amplifying a library of DHS generates a template representative of known DHS with excellent sensitivity and specificity (11). Although the experiments presented in this paper were each performed using 5 million cells/digestion condition, we have obtained reproducible data using 5-fold fewer cells as starting material. We feel our technique can deliver acceptable specificity and sensitivity for DHS mapping with small numbers of cells, and will therefore be of use to those researchers working with limited numbers of primary cells. An alternative approach using real-time PCR to define DHS has been previously published (10). The two approaches differ in that Dorscher et al. quantify the DHS through the loss of PCR signal obtained from DHS when DNA is digested, whereas our approach uses real-time PCR to quantify a gain of signal observed from DHS. Dorscher et al. report excellent sensitivity using their approach to map DHS. However, the technique depends on large numbers of comparative quantitative real-time PCR reactions across a region in both digested and undigested material, in order to quantify the loss of enrichment. One advantage of our technique published here is that data can be obtained using far fewer quantitative PCR reactions. The technique is highly reproducible, with relatively little variation in quantifiable enrichments observed between different biological replicates. Moreover, we demonstrate tissue specificity, with variable enrichment at known regulatory elements between different cell types.
The technique published here permits the rapid comparative analysis of DHS between different cell types from relatively small numbers of cells. It will have potential use for researchers across a broad spectrum of biology for the study of transcriptional regulation in both healthy and diseased tissues.