Our Human Variability
A conversation with Stephen Scherer
Introduction
More Things in DNA, Horatio...
Biology fascinates me. But as a non-expert, Iâm forced to think of things in pretty simple terms. So when I hear biologists talk about evolution, adaptability and natural selection, I always find myself asking: Whatâs going on, exactly? What are the physical mechanisms at play?
After all, if a species evolves through mutations of its members, then these mutations must be physically represented somewhere. And where else could that happen other than in our DNA, our own personal âinstruction manualâ of nucleotides and genes that we carry with us in every cell.
If evolution is as strong a force as we are led to believe, then, these sorts of variations must somehow be happening all around us, resulting in a world replete with manifold diversity and uniqueness that is layered upon our common humanity. Which isâto all intents and purposesâpretty well what we see when we look around and see such differences in the people on all sides of us. So far, so comprehensible.
But when the Human Genome Project announced that their DNA sequencing experiment demonstrated that we were all â99.9% identicalâ, things took a decided turn towards the unintelligible for me, and my first reaction was one of sceptical confusion, rapidly followed by one of embarrassed withdrawal.
Like many laymen, the conclusions seemed downright perplexing to me, but who on earth was I to question the scientific consensus of thousands of expert researchers from around the world?
Stephen Scherer, on the other hand, a world-class geneticist who built an internationally renowned research program at Torontoâs Hospital for Sick Children, naturally felt less inclined to be deferential to the prevailing wisdom.
âWhen the rough draft papers came out in 2000, which talked about how weâre 99.9% identical, I remember thinking, âBut weâre not identical. My brothers and I share 50% of our DNA from our parents, and weâre nothing alike.â You could probably pick us out in a crowd, but weâre really quite different.â
Letâs run the numbers. For a human genome of roughly 3 billion nucleotides, that 0.1% difference results in variations of about 3.2 million of the individual nucleotides that make up the âhuman genomeâ. So thatâs one way to look at things.
But, crucially, itâs not the only way.
Many years before the Human Genome Project reached its conclusion, geneticists had also recognized that some 0.4% of the population exhibited large-scale deviations from the normâso-called âcopy number variationââwhere huge chunks of DNA, often millions of nucleotides long, were either missing from their genome or present in extra copies. All of these large-scale changes were associated with serious medical conditions like autism or Down syndrome.
There were, then, it seemed two types of variation: one for the âdiseasedâ and one for âthe rest of usâ. It was a picture that most geneticists and molecular biologists of the time unhesitatingly accepted. But not Stephen.
âI have this figure that I always show the students when I teach. If you plot out the number of different types of genetic variation and divide them into single nucleotide variation and the copy number variations, youâll see that, in fact, 0.4% of the ânormalâ population, the average population, carries big chromosome structural changes. Trisomy 21 is mainly associated with Down syndrome, but there are other big segments of DNA in a very small portion of the population that are different from each other. 0.4% of the population have these big, big changes, and weâve known about that for 50 years.
âOn the one hand, The Human Genome Project talked about those 3.2 million potential single-nucleotide changes that everyone is subjected to,and then on the other hand thereâs 0.4% of the population who experience these large-scale chromosome changes.
âAnd when I was teaching back in 2002, I kept thinking to myself, âBiology favours balance. There have got to be a lot of other variants here. Why is it that we havenât seen them yet?â
âWell, because we didnât have the tools to see them.â
He didnât develop the right tools himself. But as a self-confessed âtechnology guyâ, Stephen had the presence of mind to aggressively seek out better and different techniques to see what others might have missed.
In 2003, he partnered with Craig Venterâs Celera Genomics to study the DNA of chromosome 7, his primary area of expertise. Venter had pioneered a different sort of DNA sequencing technique, called âshotgun cloningâ, that had also been used for the Human Genome Project. Now there was a way of comparing and contrasting the two approaches.
âWe published that in Science in 2003 with Craig Venterâs group. Figure 1 of that paper is probably the most underplayed figure in the field of genetic variation. In Figure 1, we compared the sequence we put together with the Celera group approach with the public Human Genome Project reference sequence.
âIf you look at that figure we show that there were about 167 or so sites along the chromosome that, when we compared the sequences, showed significant differences, including pieces of DNA in one that were not in the other.
âThe reviewers kept saying, âThese are just technical mistakes. You guys screwed up. You made a mistake.â We knew that wasnât the case, because we had used another form of experiments to prove that, indeed, those variations existed. But they still didnât believe us, and the editor wanted it taken out. But I said, âYouâre not getting our paper unless you leave it in. The data support it.â
âThose were the first copy number variations that were identified.â
So âcopy number variationâ again, but this time not necessarily associated with any particular condition or disease. What Stephen and his colleagues had stumbled upon was the groundbreaking possibility that large-scale, DNA copy-number variation might be nothing less than a universal human trait, a key ingredient in allowing evolutionary variabilityâconcrete evidence, in other words, that we were far more distinctive than the Human Genome Project was telling us we were.
More work, though, needed to be doneâand, once again, with cutting-edge tools.
âThe real breakthrough was this technology called microarrays, which allowed us to scan for dimensional differences in the DNA sequence. What we had previously looked for were binary differences: Was it an adenine or a thymine here? Or a cytosine or a guanine there? These are single letter changesâsite by site. There was really no good technology that allowed you to look for what I would call a copy number difference, where instead of having two copies, you might have three copies, or one copy, or in some cases zero copies.
âWe actually used DNA from a child that was autistic as our first set of experiments. I wanted to get the most bang for my buckâI wasnât going to run just anyoneâs DNAâthese experiments cost thousands of dollars. At any rate, we knew that this boy had about a 6-million-base-pair deletion on one of his chromosome 7âs, right near the cystic fibrosis gene. We knew a lot about this.
âWhen we looked along his chromosome 7, starting at the beginning, there are a few blips, then you get to where his deletion is known to be, where he only has the one copy, and it drops down a bit, and afterwards it picks up again and continues along. But along the way there were all these other blips.
âIt looked to be the same site on the chromosome; and we only saw them in some families and not others. That was really copy number variance. There were all these little blips along the chromosomes where there were segments of DNA of the order of 100,000 nucleotides long (an average gene is about 30,0...