How Cancer Thrives via Genetic ‘Wormholes’

​In astrophysics, a wormhole occurs—or is theorized to occur—at a point where gravity effectively rips a hole through the undulating sheet of space-time, potentially linking two distant space-time locations in the process. It's a tunnel or shortcut linking here with anywhere else we might imagine: local causes, distant effects.

In a new paper, scientists at the the Institute of Cancer Research, London offer an analogous explanation for, of all things, cancer. The wormholes in this case are genetic, however, but through a process known as DNA looping it appears that single-letter alterations in junk genetic code might have effects within genes far, far away.

It's an answer to a classic cancer mystery: How do genetic mutations in seemingly unrelated, unimportant portions of the genome (genomic "deserts") lead to cancer? There are portions of our genetic codes that tend to not do very much, consisting of what's often termed "junk DNA." While these genes are mostly non-functional while healthy and intact—with ​"little specificity and convey[ing] little or no selective advantage to the organism"—they offer an outsize influence on cancer-causing mutations elsewhere.

But why?

The most general answer offered by the new study—which is based on a newly-developed technique for probing genomic interactions called Hi-C—is that isolated genes aren't actually all that isolated. Instead, they're linked back into the genomic mainstream by DNA looping, sometimes across great (molecular) distances.

"A lot of the genetic variants already linked to cancer occur in gene deserts—often very long and quite mysterious DNA sequences that don't actually contain 'genes,' but which are involved in causing cancer in ways we do not yet fully understand," said Paul Workman, the Institute's chief executive, in a statement. "DNA looping is notoriously difficult to study but this research has taken an important step to understanding what genetic variations in DNA deserts might do to drive the development of bowel cancer."

DNA looping is easy enough to understand. The activity of genes is regulated by the presence of different proteins, which act as sensors and messengers telling different genetic zones that the right conditions are in place to start the process of DNA transcription. And transcription is itself the first stage in the process of gene expression, e.g. when genetic information is translated into IRL action.

For example, gene expression might lead to cellular differentiation, where a bone marrow stem cell might become a white or red blood cell. Or gene expression might issue some instruction involving morphogenesis, which is the collection of processes by which an organism literally takes shape along the path from cell to tissue to organ to organism.

"For example, transcription of a gene could require the simultaneous presence of four different conditions, the status of each of which must be transmitted by a protein sensor to the initiation complex," explained Johns Hopkins University biochemist Robert Schleif in a survey published in the Annual Review of Biochemistry. "In order that the protein sensors, which we normally call regulatory proteins, confine their activities to the correct genes, they bind to specific sequences located near the genes to be regulated."

Looping in the most basic sense is where one protein is able to bind simultaneously to two distant sites on some DNA. This has all kinds of benefits, from allowing proteins to effectively do double duty to helping bridge or tether together different DNA sites. A cooperative short-circuit, perhaps. This can work the other way too, where there are two different proteins brought together into a single loop by virtue of associating with different but nearby DNA sites. It's a clever, complex series of interactions.

But it has a downside, as the current study reveals. By encouraging cooperation across genomes, looping allows for distant mutations to have local effects. The researchers examined 14 different single-letter genomic variations associated with bowel cancer and found that for every one there was an associated long-distance looping relationship.

It's a mystery solved, but the implication is also a bit ominous. If cancer could come from anywhere on a genome, even a junk DNA wasteland, beating it genetically may be that much more difficult—an ugly game of mutation whack-a-mole.