A new interaction map of the Drosophila melanogaster genome shows that it is sharply divided into domains. Whether or not those domains interact with one another depends largely on whether the regions are expressed or not expressed.
Giacomo Cavalli, a researcher at the Institute of Human Genetics in France, and his collaborators adapted the chromosome conformation capture approach, making it work genome-wide and at a higher resolution, to study how different parts of the fly genome interact with each other.
From this, Cavalli and his team found that the fly genome is highly compartmentalized. The entire genome contains numerous domains that strongly interact within themselves, and the ends of those domains are marked by frontier regions that contain insulator binding sites. "What we did see was basically that the genome is made of these domains," he says.
The domains also interact with their neighbors — less so with those domains that are farther away. In addition, active parts of the genome tend to interact with other active parts while inactive parts of the genome interact with other inactive parts. "It's a hierarchical organization," Cavalli says. "First, you have the domain, then you have interactions between or among these surrounding domains. Whenever you look at active portions of genome, they have a strong interaction within a domain, but they tend to interact with other active portions of the genome — they can be far away, even on [other] chromosomes.
The inactive domains stick together and Cavalli says they might form the backbone of the chromosome. But not all inactive domains are the same. One type seems to have little genetic activity, he says, while another other appears to be heterochromatin, and yet another is marked by polycomb proteins that often silence signaling or other pathways.
Cavalli and his team plan to focus their next efforts on better understanding these interactions using mutant flies.
"Basically, the idea would be to go and mutate many of these components to see what components define or determine the interactions that you have within a domain or between neighboring domains or between long-range interactions between different domains, and possible different chromosomes," he says. "I think that is going to be keeping us busy for the next year."