• Reach Magazine
  • Genetic Networks

Decoding Autism

by Cynthia Macdonald Apr 5 / 15
Decoding Autism
Stephen Scherer in his lab (credit: Aaron Wynia)

When speculation about possible Nobel Prize winners started up last fall, geneticist Stephen W. Scherer was at the top of the list.

His main research interest — decoding the genetic origins of autism spectrum disorder — was attracting ever more interest from A-list donors and scientists around the world, and major projects he was involved in had recently secured backing from Google.

Given all this success, it’s surprising that one of Scherer’s favourite topics is, well, failure. “If you’re doing cutting-edge science, you should be pushing the limits. And most of your experiments should fail,” he says.

Regardless of prizes, 51-year-old Scherer has fomented a revolution, and he can’t rest now. “I have a million ideas,” he says. “All the time.”

The Double Helix

Did he always want to be a scientist? “Uh, no,” he laughs, sitting back in a chair that takes up a good part of his tiny 13th-floor office. “I was just a goofball kid like anyone.” The second of four sons born to a plumber and a homemaker in Windsor, Ontario, Scherer brainily skipped a grade but tended to be more interested in nature and sports — until high school, when he chanced upon a copy of The Double Helix, James Watson’s account of the discovery of the structure of DNA.

“That got me really excited about DNA, genetics and the discovery process,” he says. Still, the idea of doing such a thing for a living seemed far-fetched. “I am one of only a few students in my class who got a professional degree; many kids went to work in the factories. That’s what you did there.”

In fact, Scherer worked briefly in a sheet metal factory before enrolling at the University of Waterloo and then going on to graduate school at the University of Toronto (U of T).

Scherer’s labIn Scherer’s lab, a robotic machine that takes extremely small samples for DNA sequencing. (Credit: Aaron Wynia)

One day at U of T, he wandered into a talk given by Ron Worton, a Canadian geneticist who had enjoyed recent acclaim for his part in discovering the gene linked to Duchenne muscular dystrophy. The celebrated scientist had just come back from a planning meeting for what was to be the world’s biggest group biology project: identifying and mapping the DNA sequence of all 25,000 genes in the human body. Scherer was hooked.

A key purpose of the Human Genome Project was to identify genes known to play a role in the development of disease. In the 1980s, development of a new technique called positional cloning allowed scientists to locate the position of a gene on a chromosome. Defective genes linked to ailments such as retinoblastoma, Huntington’s disease and muscular dystrophy had already been mapped (that is, pinned to a particular chromosome) or found.

The race to find others was gathering steam, and Scherer wanted in. As luck would have it, the one place he really wanted to go was mere blocks away. The Hospital for Sick Children had recently made important advances in its work on diseases like retinoblastoma and Tay- Sachs disease. “Sick Kids was the place in the world to be for human genetics,” Scherer says.

Scherer also found a mentor to match in Lap-Chee Tsui, a Sick Kids geneticist he began to work with. Tsui was the kind of deeply creative scientist Scherer wanted to emulate. (Tsui had been interested in architecture and had studied design; Scherer is an avid art collector.) Tsui’s small and crowded lab was rapidly closing in on CFTR, the gene that (when mutated) causes cystic fibrosis. 

Canada's Chromosome

Tsui had already mapped CFTR to chromosome 7. He sent Scherer on a kind of scavenger hunt to help find it, something the younger man likened to “solving a puzzle with 158 million pieces.” By the summer of 1989, Tsui had found the cystic fibrosis gene.

In Tsui’s lab, Scherer hit on a technique that could clone up to a million nucleotide pieces at a time. He became an expert on chromosome 7, now known as “Canada’s chromosome” in tribute to the disease findings Canadian scientists have linked it with. These have included genes for colon cancer, leukemia and another confounding disorder that was soon to change Scherer’s life.

A centrifuge used to spin DNA. (Credit: Aaron Wynia)

Sometime around 1996, Wendy Roberts, codirector of the Autism Research Unit at Sick Kids, walked into Scherer’s office, which was festooned with maps of chromosome 7. “I told him, ‘There’s got to be something to do with autism on chromosome 7. We need you in our world,’” Roberts says.

Scherer didn’t know much about autism, a neuropsychiatric condition invariably marked by diminished social skills. People with autism engage in repetitive behaviours and often have language deficits. The condition may confer severe disability, savant-like genius or anything in between. “I did a little digging and found we were seeing lots of kids in our pediatric hospital with this.… Also, previous papers showed that genetics was a major factor,” recalls Scherer. “I figured I could use my expertise in mapping to find some of the genes.”

CFTR had been hard to find; locating the cause of autism was going to be even harder. Whereas cystic fibrosis is caused by a nucleotide swap in a single gene, autism emanates from possible defects in well over a hundred genes. How these mutated genes interact — with themselves, and with the environment — remains a key question in Scherer’s work.

A Hunch

The completion of the Human Genome Project in 2003 brought him closer to the answer. The newly available information helped usher in the era of genomics, and microarray scanning made it possible for scientists to study huge numbers of genes at once and discover heretofore unseen characteristics. Scherer had noticed some odd structural changes on chromosome 7, and he had a hunch. What if the biological gospel that all humans were 99.9 per cent identical turned out to be wrong?

To find out, he had to run a lot of expensive microarray sequences. “I burned through $50,000 every month, and most of those experiments failed,” he says. CIFAR, he says, was among the donors willing to support his ideas at the time

Steadily, however, more data emerged, and Scherer decided to pool his results with Charles Lee, a Korean-born Canadian at Harvard University. Scherer and Lee realized that they were seeing large-scale variations in the genome. According to received wisdom, we all inherit two copies of every gene, one from each parent, but Scherer was seeing that, in many cases, this wasn’t true. Sometimes, only one copy existed; sometimes there were three; other times, genetic material got swapped around.

Microtitre plates
Microtitre plates, each capable of holding DNA samples from 96 individuals, prior to being loaded into a sequencing machine. (Credit: Aaron Wynia)

Scherer had long known that rare changes in copy number — copy number variations, or CNVs — led to disorders such as Down syndrome. He also knew that differences at the base-pair level accounted for other diseases, and for things like hair and eye colour. By comparing the genomes of people with autism and controls, Scherer found that certain CNVs tended to occur in people with autism.

But his work went further. What no one knew until that moment was this: every single healthy person on earth may harbour a dozen or more genetic deletions or duplications, just as people with congenital diseases do. Maybe those changes foretold diseases yet to develop, provided information about how we processed drugs and food, or did nothing at all.

The Importance of Collaboration

Since that day, Scherer’s group has published close to 100 papers describing disease associations with CNVs. Autism research — which involves multiple gene mutations, including but not restricted to CNVs — is now his main focus. It’s one that requires teamwork.

This is where Scherer excels. In the old days, says Wendy Roberts, “people weren’t very inclusive in terms of collaborating. Everybody wanted to make their own big discovery and didn’t want to share. It became clear that if anything would work, it was collaboration. And that’s one of the things Steve’s really good at.”

Through its Genetic Networks program, CIFAR has given Scherer the opportunity to work with scientists who wouldn’t ordinarily cross his path. Says Brenda Andrews, co-director of the program: “Dr. Scherer’s remarkable work embodies an overarching goal of the program — to catalyze new interdisciplinary collaborations with the goal of discovering how genes interact in complex human genetic diseases.”

“Stephen Scherer’s work in Genetic Networks has been a unique contribution. The bulk of the work in the program uses simple models such as yeast. His work challenges the entire program to think about the complexities of the human model,” says Pekka Sinervo, CIFAR’s senior vice president, research.

The Autism Formula

An example of this is Scherer’s ongoing collaboration with Brendan Frey, a senior fellow in CIFAR programs in both Genetic Networks and Neural Computation & Adaptive Perception. Frey is a University of Toronto computer scientist who studies gene regulation, which examines the processes governing how genes are expressed.

In some people, multiple genes might interact; other people may exhibit only a single mutation; still others might have different mutations on the same genes. “The combinations are exponentially large,” Frey says.

Realizing they couldn’t throw a lasso over all possible combinations (and mindful that new ones can always arise), Frey and Scherer have still been able to both identify a core group of genes involved in cognition and develop an algorithm to calculate the probability of whether certain genes will lead to autism. Their paper on this “autism formula” was published last May in Nature Genetics and another related study in Science in 2015.

“Steve’s able to decode more genomes, and I’m able to infer causal explanations for autism using computational analysis,” says Frey. “That’s what brought us together in the first place.”

Indeed, the acquisition of more genomes will allow Scherer and his collaborators to see ever more recurring patterns. Scherer says he can now link specific genes to 20 per cent of autism cases, up from zero per cent a decade ago. He thinks 50 per cent of cases will ultimately be directly attributable to genetic factors, with the environment possibly playing a significant role in the rest.

Ultimately, though, Scherer is seeking genomes of all types. The Database of Genomic Variants, which he established in 2004 with funds from CIFAR, is the leading CNV database in the world and is used by thousands of clinical laboratories worldwide.

Scherer is also behind the Personal Genome Project Canada. He is recruiting healthy volunteers willing to assist medical science by donating their entire genome for study.

Early Intervention

Scherer also knows that there are ethical questions attached to all of this. If your genomic information is made public, will you be protected against discrimination by employers or insurers? When prenatal testing detects CNVs linked to autism — as it now can — will that usher in an era of genetic engineering.

Neither Scherer nor Roberts wants this. “Genetic modification is not our goal,” Roberts says emphatically. What they do want is a better understanding of where autism comes from, so that no parent believes that it results from bad parenting. They also want medicines specific to autism symptomatology. Most of all, they want to encourage early intervention.

Mike Lake, the father of a child with autism and a Conservative member of Parliament from the Edmonton area, seconds this. His son Jaden was diagnosed with autism several years after he was born. Lake’s advocacy and love for his son are evident. Echoing his teenage daughter, Lake says: “If Jaden could be cured of autism, we wouldn’t have the Jaden we have now.”

But Lake knows that in future other families might well take advantage of Scherer’s findings to benefit from intervention or treatment at the infant stage.

“This is critical,” Lake says. “If we can find a way to identify kids at age one, or even earlier than that, we can work on finding interventions that will have maximum impact.” Roberts agrees and cites a new study showing that autism can actually be turned around if caught at a very early stage.

Interacting with people like Mike Lake is a big part of what Scherer does. “On an annual basis, we invite pretty much all the families who are enrolled in our genetic studies to a meeting so we can update them and get their input,” he says.

“One of Steve’s great strengths is his ability to communicate his research in ways that people can understand,” Lake says. Wendy Roberts adds that, because Scherer is so friendly and approachable, he is contacted by parents from all over the world who worry that their child might have autism. “This goes way beyond what most scientists do,” she says.

Scherer justifiably prides himself on a great “bench-side” manner and wants to talk that concept up to other scientists. “CIFAR has been great,” Scherer says. “We had a workshop last year where we brought in some of the different CIFAR groups to talk about the social aspects of what we do, because how we communicate our results is so important.”

Being in the Moment

When in rare moments Scherer does get time alone — when he’s not with his son, daughter and wife Jo-Anne Herbrick, a biologist who manages the Centre for Applied Genomics facility, or with the vast array of families, students, funders, journalists and colleagues all clamouring for his time — well, that is when he thinks. His best ideas, he says, “always come to me in strange places. Airplanes are ideal, because nobody can bug you as you contemplate.” Gardening, which he’s been known to do at midnight is good for that, too.

Of the Group of Seven originals that Scherer owns, one in particular is worth mentioning. It’s a rare glimpse of Tom Thomson hard at work, painted by Arthur Lismer. In the painting, Thomson appears to be in what psychologists call “flow,” and what Scherer calls “being in the moment”: working with such exhilarated diligence toward a goal that failure, motivating though it may be, is simply no longer a possibility.

It could well be a painting of Scherer himself.