A new method significantly increases our ability to explore how genes are organized in their spatial environments. Three studies have pioneered this approach to show that many interactions between genes take place in the nucleus. They also provide new insights into how such interactions contribute to gene regulation.
Chromosome conformation capture (3C) is widely used to tell whether genes are closely positioned in the nucleus. It uses gene-specific PCR primers to generate a product if two genes are close enough to be ligated following chromatin crosslinking. In the new approach, which two of the studies have named 4C, adding a circularization step followed by inverse PCR removes the need for two gene-specific primers. Any sequence that comes into close proximity with a gene of interest can be detected, whether or not such an interaction was previously suspected.
Würtele and Chartrand used 4C to probe the spatial environment of the HoxB1 locus in mouse embryonic stem cells by looking at the genomic positions of the 4C PCR products. Loci that lie closest to HoxB1 along the chromosome colocalize with this gene most frequently in the nuclear space. When the cells were induced to express HoxB1, interactions with loci on other chromosomes became more common. This is consistent with models in which inactive genes are constrained within a relatively small space, but explore a larger area when their expression is induced, which can also involve a repositioning of the gene in the nucleus.
Simonis and colleagues took a similar approach, but used microarrays to identify interacting loci following PCR. They studied interactions with two loci, the tissue-specific 
-globin gene and the housekeeping gene Rad23a. Both were found to make many contacts in 3D space, with intrachromosomal interactions being more frequent than interchromosomal contacts. Interestingly, in brain tissue, in which the -globin gene is silenced, it tends to interact with inactive genes, whereas in blood cells, in which -globin is expressed, it interacts with active genes that are not necessarily tissue-specific. Rad23a makes intra- and interchromosomal contacts with active genes and these contacts seem to be largely conserved between tissues.
A third study also used 4C combined with microarrays, this time to explore the epigenetic control of physical interactions between loci. The parental origin of an imprinting control region (ICR) upstream of H19 is known to control the formation of a chromatin loop that allows the ICR to contact and silence a neighbouring gene, Igf2. Zhao and colleagues used 4C to investigate whether H19 also contacts other loci. They uncovered an extensive network of interchromosomal and intrachromosomal interactions, with both imprinted and non-imprinted genes, that was linked to either activation or silencing of the interacting genes in different cases. Importantly, many of these interactions were epigenetically regulated, relying on the methylation status of the H19 ICR.
Although 4C has been applied to just four genes, a complex set of interactions has already been uncovered. A thorough exploration of interactions for many more genes, and a careful analysis of how they relate to gene expression, will be key challenges for the future.