By Jennifer Welsh
Plastic balls, plaid, Ramen noodles, and Romeo and Juliet. Erez Lieberman-Aiden sees the genome very differently than the rest of the world.
He gained this view while working on an immunology-based project. He started wondering if the genome folding that creates antibodies might be occurring elsewhere, and if these interactions might affect gene expression.
"There are some big gaps in our understanding," said Lieberman-Aiden, currently a research fellow at Harvard University. "Between the scale of about 100 bases and about 100 million bases we have a very limited knowledge of how it is the genome is folded."
The genome is essentially a polymer -- a long string of repeated bases. At the base pair level two strings of bases wind into a double helix, which folds around structural proteins. We also know how chromosomes, huge chucks of hundreds of millions of base pairs that make up our genome, are arranged, but the levels between are a scientific grey area.
Figuring out genomic interactions across these distances could explain how the genome can create such a wide variety of cells in our body. "It has something to do with the regulation of this information -- how it is accessed -- what is turned on, what is turned off," said Lieberman-Aiden. "Our work has actually shown that that [regulation] is intimately related to how the genome is folded."
To understand this folding, Lieberman-Aiden and his teammate, Nynke van Berkum, approached the genome as if it were the text of Romeo and Juliet written out one letter at a time on a huge noodle. If you swirl that noodle in a bowl, like chromosomes in a nucleus, many different sections come into close contact.
To identify these touching sections, they froze the strand (be it noodle or genome) where it was, shattered it, glued it back together and then “read” all of the pieces, looking for the out-of-order sections. For example you might see a strand where "Two households, both alike in dignity" connected with "and Juliet is the sun!" instead of, “In fair Verona, where we lay our scene.” This would mean those two sections were interacting across long stretches, because they were out of place when reconnected.
When the researchers analyzed which sequences within a chromosome were interacting, they could see they formed plaid-like patterns of compartments. The red compartments had more interactions and more activity, while the blue compartments had fewer interactions and few markers of activity.
This interaction data showed something interesting; at the million-base level the data didn’t fit accepted polymer folding models. The traditional theory, a tangled mess called the equilibrium globule, just wasn’t possible. The team found that a structure called the fractal globule, which had never been observed before, fit their data better. It’s just as dense as the equilibrium globule, but unknotted for easy access.
Future work includes studying even smaller regions of interaction within the genome. They are also hoping to study different cells during differentiation to see how the genome changes conformation.