For a place designed to explore what makes us who we are, the Allen Institute for Brain Science could hardly be in a more fitting location.
Overlooking a canal in Seattle's artsy Fremont district, the institute is a stone's throw from outdoor patios and coffee shops that hum with social interaction. Everywhere, the glorious mingling of human brains – their pleasure at getting inside each other – is on display. The way our brains do this, in a manner so conspicuously different from that of other animals, has intrigued philosophers and scientists for centuries.
What the institute brings to the table, and what may finally offer some insight into the essence of what makes humans unique, is big data.
Founded in 2003 with a $100-million donation from ex-Microsoft mogul Paul Allen, it has been on a mission to transform neuroscience by mapping the brain the way Google (coincidentally, a next-door neighbour) has been mapping the planet.
"The whole idea is to put this data in the hands of the right people," says Ed Lein, a senior researcher at the institute. "It all goes out as we generate it."
Dr. Lein leads a project called the Brainspan Atlas. It amounts to a three-dimensional reference for how the brain is put together and how it changes through time, revealing precisely which genes are active in different brain regions as cells grow, migrate and create the circuitry that is essential for thought and behaviour.
The scale of the project is … well, mind-blowing.
By linking the functions of genes to specific locations and moments in brain development, the atlas has effectively revealed the brain's construction crew in action. This feature has been eagerly taken up by researchers beyond the institute who are looking for clues to a range of complex mental disorders – cases where the construction crew has taken an unexpected turn.
"Three years ago, this didn't exist," says Mohammed Uddin, a postdoctoral researcher at Toronto's Hospital for Sick Children. He developed a novel way to sort the Brainspan data and compare it with other data sets, and uncovered, in one fell swoop, thousands of potential genetic links to autism spectrum disorder, a debilitating condition characterized by problems with communication and socialization.
The results, which sparked much excitement in the autism community when announced last month, suggest a new framework for thinking about the genetic basis of autism. They help to explain why the disorder manifests in some people and not others, such as siblings who share similar genetic characteristics. In a clinical setting, they may offer a surer route to identifying which children will end up on the spectrum long before behavioural symptoms appear.
But the Hospital for Sick Children study also may point to a broader discovery that includes all humans, with or without autism spectrum disorder.
Because those with autism are challenged in the way they exhibit one of the more distinct human traits – the way we interact – they also may allow us to discover where the recipe for that trait lies in our genes. In sifting out the genetic vulnerabilities underlying ASD, the study may have zeroed in on the evolutionary toolkit that makes us different from other species, compels us to mingle like the patrons of Fremont coffee shops and, tens of thousands of years ago, equipped us to take over the planet.
"I think what we've done, in essence, is find the minimal set of genes involved in human cognition," says Stephen Scherer, director of the Centre for Applied Genomics at the Hospital for Sick Children.
"Natural selection has sculpted these genes by tweaking their expression in the brain."
Scientific double threat
Dr. Uddin himself is a personal study in how the right mix of traits can go a long way.
Born in Dhaka, he still remembers the computer his sister gave him when he was in high school. The pleasure of learning to program its Pentium I processor had him dreaming of a career in Silicon Valley.
The best chance of getting there, he reasoned, lay with Memorial University in Newfoundland – the place where his family's budget intersected with the cost of a degree in computer science.
Arriving in St. John's in January, 2001, the boy from Bangladesh found himself smack in the middle of "the worst winter ever." He survived it and over the next few years stuck to his plan, earning a bachelor's, then his master's degree.
But then the plan changed. Inspired by Todd Wareham, his graduate supervisor at Memorial and a scientist with wide-ranging interests, Dr. Uddin was drawn to computational biology – using computers to analyze living systems. A few years more, and he had a doctorate in genetics.
Scientifically speaking, his double specialty is a killer combo. While computers have become essential to dealing with vast quantities of genetic information, these days those who program the computers are often hired hands, not the investigators pursuing the scientific questions that drive the research. Dr. Uddin is computer scientist and geneticist in one package. "It gives you a fresh look at the problem," he says.
Offered a fellowship at Dr. Scherer's lab in Toronto, renowned as an autism research centre, he arrived in 2012 just as the Brainspan Atlas in Seattle was starting to make ripples.
It had been created by sectioning the donated brains of four stillborn human fetuses 15 to 21 weeks old, which Dr. Lien calls "a critical phase of cortical development, where both genes and the environment can adversely affect formation and therefore function later in life."
In each walnut-sized specimen, he and his team at the Allen Institute painstakingly identified and extracted tissue from about 300 distinct brain regions using a technique called laser microdissection. The method is ideal for cutting small complex shapes, allowing researchers to follow the brain's anatomy as they sampled the various regions.
They then used the cut-out samples to identify which bits of DNA are at work in each region. This provides an important new piece of information, Dr. Lein says, because it reveals "where these regions of the genome are active across the brain and different stages of development."
The BrainSpan data also can be directly compared to similar data available in mouse development, he adds, showing which genes are used in the same way by different mammals and which are used in unique ways in humans.
The exon revelation
What makes the Brainspan Atlas especially useful to researchers investigating autism is that it shows genetic activity in the brain not just at the level of individual genes, but at the level of even smaller pieces of DNA.
Each of the roughly 20,000 genes in the human genome consists of a series of segments known as exons and introns. Whenever a gene is used to create a protein that a cell may need, the information from the exons is assembled in a particular order, while that from the introns is spliced out. Importantly, this mechanism also helps to drive evolution since different assemblies of exons can allow a gene to be repurposed for different uses.
Without knowing what is happening at the exon level, it is hard to understand what effect a change in a particular gene may be having, Dr. Scherer says. If the change doesn't affect a key exon, it may not matter at all.
"Instead of the usual idea of not being able to see the forest for the trees, here we have to look at the characteristics of each tree – the exons – before we can see the forest emerge," he explains.
The high quality of the Brainspan Atlas data got Dr. Uddin thinking about another newly available resource, the Exome Variant Server, also based in Seattle – at the University of Washington. Currently the largest database of its kind in the world (an exome is a catalogue of all of a person's exons), it consists of the sequenced exons from thousands of apparently healthy individuals. The variations among them therefore represent a normal spread of genetic differences in a typical population.
Looking at the two sets of data – Brainspan and Exome – Dr. Uddin wondered, "What is the relationship between them?" Both contained information down to the individual exon. Perhaps together they could reveal something relevant to ASD.
He set about writing computer code that would allow him to make a comparison. Part of the task involved checking all the exons in the Exome Variant Server to see which have higher and lower rates of mutation in the healthy population. He and Dr. Scherer worked on the assumption that exons with very low mutation rates must be important, and so were being carefully preserved by the body's built-in genetic defence system.
He then turned to the Brainspan Atlas to see where and when those same preserved exons were busy in the prenatal brain. This led to a whopping-large group, about 30,000 exons, that were rarely mutated and very engaged – what geneticists call "highly expressed" – during brain development.
Such a large number is not a surprise. The Brainspan data show that about 95 per cent of all human genes perform some function somewhere in the developing brain, but most are also used in other tissues and organs, too, fulfilling multiple roles in the growing body.
Dr. Uddin's next step was to narrow his list of exons to exclude those that are highly expressed in cells outside the brain. Such exons are unlikely to be related to autism since, if shut off or changed, they might cause problems all through the body, whereas the symptoms of ASD are largely cognitive.
After filtering out all but the brain-specific candidates, he was left with about 4,000 exons distributed among 1,700 genes. When he and Dr. Scherer looked to see where those genes had come up elsewhere in the scientific literature, they were stunned. Most of the hundreds of genes already suspected of having a connection to ASD were in the group – about one-third of the 1,700.
When they checked other cognitive problems suspected of having multiple genetic causes, including schizophrenia and intellectual development disorder, they got the same result. The genes that were turning up in studies of people with these disorders were frequently the ones that fell into the category of very low variation in the typical population and very high expression in the prenatal brain.
"It's kind of a simple concept, in retrospect," Dr. Scherer says. But until now no one has found a sensible way of relating all the various genes that have turned up in connection with high-order cognitive disability.
Our Neanderthal link
The discovery raised a striking possibility to Dr. Scherer. Might these 1,700 genes somehow represent the building blocks of the uniquely human mind?
Although individuals with ASD can show extraordinary mental capacity, it is particularly telling that the disorder manifests itself as a deficit in communication.
In considering the evidence for how humans evolved, anthropologists have long noted their lack of any obvious physical advantages over their nearest relatives, the Neanderthals and the Denisovans, who populated Europe and Asia for tens of thousands of years before modern humans effectively replaced them.
Clearly, they were a lot like us – so much so that genetic analyses of Neanderthal and Denisovan tissue extracted from fossil teeth show that humans were able to breed with them. It's estimated that 2 per cent of modern Europeans' DNA is Neanderthal.
That is a startling number, but one might just as easily ask why is the overlap so small? If humans were cross-fertile with other branches of their family tree, why did these branches not simply blend together once in contact with one another? Why don't we share more DNA?
"The early modern humans did something very unusual: They won every time," says Ajit Varki, a professor of cellular and molecular medicine at the University of California San Diego and co-director of a research centre devoted to exploring and explaining the origins of what makes us human.
Dr. Varki thinks he knows why humans were so successful. While they may have been physically similar to their cousins, behaviourally they were light years apart. He suspects that somehow a small group of individuals in Africa developed a suite of gene variations that enabled something like the nuanced social communication that humans practice today. These new variations would be keyed to the social environment that young humans typically grow up in, including a prolonged period of immaturity and learning after gestation.
If so, says Dr. Varki, the children of a mating between humans and Neanderthals may have been physically healthy but "cognitively sterile." The disadvantage of losing the uniquely human genetic package, even to a small degree, would have conveyed a tremendous cost in terms of social interaction and reduced their chances of reproductive success.
It's hard to prove or refute this idea. Since behaviour doesn't leave fossils, there is no way to be sure how our ancestors compared with their cousins. Archeologists have found some signs that Neanderthals cared for the injured and were expert tool-makers, but they do not appear to have been especially innovative.
There is marked explosion in the kinds of artifacts that show up once modern humans are on the scene. In fact, long before written history, our ancestors were essentially creating the paleo equivalents of Athens, Florence – and Fremont.
Could the 1,700 genes that turned up in Mohammed Uddin's digital sieve somehow be involved?
The word is out
Just weeks after publication of the Hospital for Sick Children work, its impact is being felt in the world of ASD research. Researchers have begun to contact Dr. Scherer with various genes that have turned up in studies of people with ASD to see if they appear in the set that Dr. Uddin identified. One of the immediate applications of the work is weeding out genetic variants that are clearly not worth pursuing because they aren't highly preserved or highly expressed in the brain.
Meanwhile, with the 4,000 exons that have been identified, there is much more work to be done to establish what proteins they make and how altering their production may impinge on brain development.
In time, the work could help to identify the many forms of autism that make up the spectrum. This is crucial information when those with ASD are administered drugs for a clinical trial. If a gene variant occurs only in one subset of people with ASD but a drug that compensates for that variant is tested across a broader spectrum, then the drug "would appear to be a failure even if it responded to everyone it was meant to respond to," says Greg Farber, director of technology, development and co-ordination at the U.S. National Institute of Mental Health in Bethesda, Md.
Another implication of the Hospital for Sick Children work is that it may soon be possible to calculate a risk for ASD based on a whole exome sequence. This could, in turn, help to prepare caregivers and parents before symptoms start to show up in a toddler.
But the work also has Dr. Scherer thinking about deeper questions. For example, how does the set of exons they found compare with those same exons in non-human primates?
"That's my perfect experiment," he says. "If we had the Allen Institute equivalent for primates, my guess is that you would see a smaller set."
Dr. Lein agrees that the idea has merit and that the Brainspan Atlas may offer a window into human uniqueness. His institute is already working to map the structures and genetic relationships within some non-human primates – and such work has a purpose beyond intellectual curiosity.
"At some level I think we are all interested in what makes human beings unique," he says, "but it also has important ramifications for treating human disease."
Because disease research is mainly conducted with non-human model organisms, such as mice, Dr. Lien adds, researchers need to understand what kind of genetic signals are common among different species and what is unique and can be understood only by looking at humans.
For Dr. Uddin, who has grown increasingly aware of the burden carried by those with ASD since going to the Hospital for Sick Children, the payoff comes from watching his ideas and his number-crunching materialize into something that may improve lives.
"These kids often don't talk to anyone, or have any way to express their concerns," he says. "As a researcher, I want to help if I can. It's us, as a society, that has to come up with our own understanding of how to approach this."
Along the way, the bonus prize may be the ability to see what it is that compels us to connect – to get inside each other's heads – in a way that no other creatures seem to do.