The HSPome. Credit: Novarino et al, Science 2014.
The HSPome. Credit: Novarino et al, Science 2014.

Now This Is How You Find Disease Genes

When you read stories about scientists identifying a new link between Gene X and Disease Y, the underlying studies vary a lot in quality. At one extreme, you get papers which show that a variant of Gene X is common in a small group of people with Disease Y and not in healthy controls… and that’s it. You don’t really know if X is really responsible for Y, or even if the result is genuine and not a false alarm produced by small numbers.

At the other extreme, you have this—a study that used a smorgasbord of experiments to identify 18 new genes behind hereditary spastic paraplegias (HSPs). This diverse group of genetic disorders all involve damage to the long neurons running between the brain and spinal cord, leading to stiffness and involuntary contractions in leg muscles.

Scientists have already linked 22 genes to HSPs, but these only explain around 20 to 30 percent of cases. “Many of the children with these conditions can’t receive a proper diagnosis and there’s no treatment available,” says Joseph Gleeson at the University of California, San Diego. “We wanted to understand more about the causes, and hopefully see some new treatments come out of that.”

To find more HSP genes, Gleeson’s team forged contacts with scientists in countries where HSP is more common and where genetic studies are rare, including Egypt, Pakistan and Iran. They found 55 families with the disorders and sequenced every gene in 93 of their members. They identified several genes that seemed to cause HSPs in these people, and they bred mutant fish to check that getting rid of these genes actually does produce relevant symptoms. They created a network to show what these genes do, and how they interact with each other. And they used that network to find even more HSP genes.

The scope of the work, led by team members Gaia Novarino, Ali Fenstermaker and Maha Zaki, is incredible. “We’ve been working on it for close to 10 years,” says Gleeson. “We just didn’t feel comfortable publishing it until it was all done. Hopefully, people will look to our paper as a roadmap for studying genetically diverse conditions.”

Between them, the 18 new genes and the 22 old ones explain around 70 percent of the HSP cases among the team’s recruits. “That is of enormous value, not only for biological understanding, but also for providing a definite diagnosis in families and for accelerating research into possible treatment of these progressive disorders,” says Joris Veltman, a geneticist at Radboud University Nijmegen Medical Centre.

Just finding the families was hard enough. “It’s not easy for an American to get into Iran,” says Gleeson. But it was worth it because in these parts of the world, the practice of marrying relatives means that family members share an unusually high proportion of their DNA. This makes it easier to find recessive genes that only cause HSP when people inherit two copies.

The team sequenced every volunteer’s complete exome—the 1 percent of their genome that codes for proteins. By comparing the exomes of family members with or without HSPs, they showed that a third of the cases were due to genes that had already been implicated in the disorders. But another 40 percent were possibly caused by mutations in 15 new genes.

Next, they verified this list by engineering baby zebrafish that lacked each of these candidate genes. None of the mutants could swim properly. Some, for example, had tails that were permanently curved to the side, much like the stiff limbs of children with HSP. “We felt compelled to do that,” says Gleeson. “For a lot of the genes, we only had a single family with the mutation.” Without the fish experiments, he wouldn’t have felt comfortable claiming that these genes were really related to HSP.

Exome sequencing is quickly becoming the frontline technique for gene detectives, who no longer have to narrow down their search to specific parts of the genome. They can just sequence every gene and see what jumps out. “The current study clearly takes this approach to the next level by applying it to a very large cohort and performing systematic functional follow-up studies,” says Veltman.

Even that wasn’t enough. “In some diseases, one is left with a hodgepodge of genes and no clear path forward,” says Gleeson. “We tried to weave commonalities between our genes and understand what they were telling us.” They did that by mapping all the interactions between their HSP genes and the proteins they make, creating a tangled network that they call the “HSPome”.

The genes clustered in different groups based on what they did. “It was like lifting the veil,” says Gleeson. “We could see how all the factors that were identified fit together.” Some are involved in folding proteins correctly, others help to make building blocks of DNA, and yet others help neurons to grow and move to the right places. These clusters tell us about “points of molecular vulnerability” in the brain-to-spine neurons that are damaged in HSPs, says John Fink from the University of Michigan.

The team then extended the network to look at other genes that interacted with the ones they identified—the “friends of friends”. By scanning this extended list, they found more three more new HSP genes, which underlie the disorders in three more families. That brought the total up to 18.

The network also overlapped a lot with other sets of genes that have been implicated in Alzheimer’s disease, Parkinson’s disease and Lou Gehrig’s disease. This suggests that these disparate brain diseases may have some common ground, and that drugs which target these overlapping genes could help to treat several conditions.

“This is important, because drug development is very costly, and the larger the potential market, the more interested pharmaceutical companies will be to pursue these leads,” says Craig Blackstone from the National Institutes of Health. Indeed, by linking HSPs to better-studied (and better-funded) conditions, Gleeson hopes to spur interest in these often-overlooked conditions.

“These are exciting times for research not only into the causes and treatments for HSP but for other neurodegenerative disorders as well,” says Fink.

Reference: Novarino, Fenstermaker, Zaki  Hofree, Silhavy, Heiberg, Abdellateef, Rosti, Scott, Mansour, Masri, Kayserili, Al-Aama, Abdel-Salam, Karminejad, Kara, Kara, Bozorgmehri, Ben-Omran, Mojahedi, Gamal El Din Mahmoud, Bouslam, Bouhouche, Benomar, Hanein, Raymond, Forlani, Mascaro, Selim, Shehata, Al-Allawi, Bindu, Azam, Gunel, Caglayan, Bilguvar, Tolun, Issa, Schroth, Spencer, Rosti, Akizu, Vaux, Johansen, Koh, Megahed, Durr, Brice, Stevanin, Gabriel, Ideker, and Gleeson. 2013. Exome Sequencing Links Corticospinal Motor Neuron Disease to Common Neurodegenerative Disorders. Science