Dennis Lo – TRANSCRIPT
When I first learned about the molecule of DNA I became fascinated by how such a seemingly simple molecule could uncode the genetic program of life. However this program, like many other programs could be corrupted and could cause life threatening disease like cancer and genetic diseases. And to diagnose these diseases, a doctor would frequently have to gain access to the cells with such mutations. But those cells could reside deep within one’s body.
Now, for example, to diagnose a cancer a doctor might have to do invasive biopsy of an internal organ. But every time he does this, there is a chance of bleeding. As another example to test for a baby before he was born a doctor would have to do pre-natal testing.
But typically, that method is performed by sampling the fluids from baby from deep within the uterus. And every time we do this, there is a chance that the baby might miscarry. And over the last few years there is a lot of effort to try to develop non-invasive methods which do not have such a risk. In particular researchers are interested in sampling the fluids that circulate in our body, like blood. And this field is generally referred to as liquid biopsy.
I first became interested in this field when I was a medical student studying obstetrics in Oxford. As a young student I wondered why did grown up doctors do dangerous things, such as sticking needles in the abdomen of pregnant women. And I was thinking that maybe we can do something better and safer. Why not just sample the mother’s blood and try to detect any fetal cells which might have entered into a mother’s circulation?
I even managed to persuade a professor in Oxford to let me work in his laboratory for a few months. I had no idea actually how long medical research could take. I actually initially naively thought that I could crack this in a few months’ time. But before I knew it, I spent eight years working on this project without a lot of success. I was lucky I wasn’t fired.
And then in the autumn of 1996 I came across two papers in a journal which said that cancer cells somehow release DNA into the blood plasma of cancer patients. Now plasma is this yellow fluid in which our blood cells swim in. And then I suddenly had a strange idea. I thought that a baby growing inside pregnant mother is actually a little bit like a cancer growing in a patient. And I thought that the placentural baby, which implanted in the uterus is a little bit invasive.
So I thought that if a smallish cancer can release enough DNA for us to see in the plasma, then surely a baby who to us at the end of the pregnancy could be like eight pounds, will surely deliver enough DNA for us to see. But the problem is that at that time I had just returned from Oxford to Hong Kong to start a new career. And I was born in Hong Kong so I had particular affiliation and attachment back there.
So at that time I really didn’t have a lot of resources, so I could only do something which was very readily available and cheap. And then I suddenly recalled when I was back in Oxford, being fed up as a medical student of the college food, I would actually sometimes cook instant noodles, back in my room. I actually got quite good at it. So what I did, is to boil some water to prepare soup base and then put the noodles in and a few minutes later you have a lovely dinner. And I was thinking what could be simpler than if I just boil some plasma and then maybe test a little bit of that boiled soup.
Now surprisingly, even with this crude preparation I was able to detect the male chromosome, that a male baby has released into his mother. And I saw no signal if a mother is carrying a baby girl. So this was the first demonstration of the presence of fetal DNA in maternal plasma. It hasn’t been shown before because previously, when people tried to look for fetal cells in the mother’s blood, plasma is the very first thing they throw away. And then even though the baby is so small compared with the mother, I was very surprised to find that actually the fetal DNA maternal plasma reached a mean of some 15%. And you can detect it as early as seven weeks of pregnancy which is very early.
And then I spent the next few years trying to develop new diagnostic method based on this technology. And the first thing I did – was because the first marker I found was on the male chromosome – so this technology can be used for sexing the baby, which is medically important, because some genetic diseases preferentially affect boys. They are the so called sex linked disease in which the disease gene is present on chromosome x and males have only one copy of this chromosome so we don’t have any back-up.
And one example of a genetic disease which is sex linked is hemophilia, which is a genetic disorder causing clotting problem of the blood. As another example I used this technology to detect the blood group type of the baby. And this could be important because sometimes the mother and baby’s blood group type do not match and a mother could produce antibody which could cause the placenta to attack the red blood cells of the baby.
Now those two applications – sexing and blood group typing – are very accurate. They are over 99% accurate and as a result, it’s now used in many parts of the world. So we next became more ambitious, we decided to actually tackle Down syndrome, which is perhaps the number one reason why many pregnant women go for pre-natal testing. And Down syndrome is caused by the baby having an abnormal number of chromosomes. And normally you detect Down syndrome by counting the number of chromosomes that a cell has within a cell membrane.
But the problem is, I am actually working on cell-free DNA floating around in the plasma. And so I don’t have the luxury of having the chromosomes nicely packaged within cells. So it actually took us ten years of trying different methods before we had to arrive to a method which is very robust and reliable. And this method involved sequencing millions of strands of DNA molecules in the plasma and had to work out the ratios of different chromosomes. And take into account of the fact that a fetal DNA only represents a minor percent of all the DNA in the mother’s plasma.
But the resulting method that we developed was surprisingly accurate – has an accuracy as 99.7% and because it was so accurate that within ten months after we published the first round trial of this technology in 2011, the test was introduced clinically in the U.S. and to date the technology is used by millions of pregnant women in over 90 countries around the world.
Now in China alone, every year 2 million of those tests are performed. And so we were encouraged about this development. So I was wondering how can we push this technology to its limit and what is the limit? The limit would be: can we use it to sequence the entire fetal genome?
Now that is a tough nut to crack because a human genome consists of 6 billion base pairs of genetic code. And initially I had no idea how we can do this. Until, one day, I went to see a Harry Potter movie with my wife, in a cinema like this, and it was in 3D. So I put on my 3D glasses and was waiting for the movie to start and then suddenly I saw the Harry Potter title flying towards me in 3D. And then suddenly my eyes were caught on to the H on the left hand side. And actually it looked to me at that time, that the H had two vertical strokes. Actually it looks like a pair of chromosomes.
So immediately I told my wife, who was sitting next to me, “I think I have an idea.” And then I spent the next two hours thinking about that idea. Actually, I never probably hoped a film would finish so quickly than that one. So within the movie finishing I just rushed home and wrote an email to my team, saying, “This is the idea, lets work on it.” And within a few months we managed to deduce the entire fetal genome from the mother’s blood, using this method.