Sam Behjati: The placenta is a genetic dumpster fire
Sam Behjati, Image courtesy of Sam Behjati
Click here to listen to the full podcast episode
Kat: As Ros discussed, the placenta is a transient organ: it grows from cells that split off from the embryo early in development, it hangs around for nine months in a human and then it’s gone.
And as Sam Behjati, a group leader at the Wellcome Sanger Institute, recently discovered with his collaborators Professors Steve Charnock-Jones and Gordon Smith, it’s an absolute genetic dumpster fire in there. The pattern of genetic alterations in the placenta is different to any other human organ and resembles that of a tumour, harbouring many of the same genetic mutations found in childhood cancers. As Sam explains, it’s a finding that he actually never set out to make.
Sam: The reason why we did this study has actually not that much to do with what came out of the study if that makes sense. So if you want me to go all the way back, that sort of takes us back about seven years when I was a Ph.D. student and did some other research. And it's a little complicated to explain. But the bottom line is that the end of that piece of research, we ended up with this observation that if you look at the fertilised egg, so it was really about what does a fertilised egg give rise to in terms of the two cells? And what part of the embryo of the pregnancy do those two cells make up? And we did that study years ago, which seemed to suggest in a very indirect way that one cell, one derivative of the first cell division, mainly makes the embryo, the baby, and the other cell mainly makes the placenta.
Kat: So what were you doing in those experiments? What did you set out to try and examine?
Sam: The initial question was, can we say what the first two cells that form from the embryo give rise to? So we've got the fertilised egg that divides into two cells and can we then study whether one cell mainly makes the placenta and the other maybe makes the embryo, or whether that is not true? And the way we do that is by looking at mutational postcodes. The cells as they divide acquire mutations, they thus leave behind a postcode of mutations. And that then enables us to sort of data wrangling to reconstruct where cells came from.
Kat: So if you've got a piece of placenta tissue and you've got some tissue from the baby, you should sort of be able to compare and almost track back and say, okay, because these genetic changes are just in the placenta, they have come down that lineage, maybe from that first cell or small group of cells. And because these changes are only in the baby and not in the placenta, they must have come from another early group of cells. So there has been some kind of split here.
Sam: Correct. It's a bit like if you look at an apple tree and you can look at an apple that's on the left side of the tree, another apple that's on the right side of the tree. You can then follow the branches and understand at what point did they diverge, do they come from one big branch together, or are they literally at opposite ends of the tree? That's the sort of basic idea.
Kat: So one question, obviously, is what tissues were you using and how do you get hold of them? Because obviously, getting hold of human samples and particularly human samples early in development can't be that easy.
Sam: Yeah. So this is where Gordon and Steve come into the equation. So Gordon and Steve are very accomplished placenta or pregnancy researchers. They have spent the past 20 years trying to understand or predict which pregnancies go wrong and then try to find ways of mitigating that. And the way they have been doing this is by essentially collecting lots and lots and lots and lots of placentas through studies, which are called the pregnancy outcome studies. I believe, yes. And their various iterations say one, two, three and so forth. And what they've done is just systematically collect, at birth placentas, different biopsies from different regions, placenta, and this is a key bit, a little bit of umbilical cord which, in embryological and developmental terms, represents a baby.
Kat: So you've got these samples and then you can extract DNA from them and then you can look at the genetic changes and compare them. So what sort of scale of data and how did you start comparing all this data to figure out what was going on?
Sam: So what we do is something called whole genome sequencing. So every cell or every piece of tissue has got a unique genetic code. That code is three billion letters in length. So we read out all three billion letters of the different pieces of placenta and the piece of umbilical cord. And then we just compare them to each other. That's literally, it's an A here, It's not an A over there, it's a T here, it's not a T over there. Mostly they will be the same because they come from the same biological unit, from the same fertilised egg. But there will be very little tiny changes in between things. Those are mutations and those are the mutations that represent that mutational postcode that I talked about at the beginning.
Kat: So when you did compare the embryo tissue, the baby with the placenta and these mutations, then what did you find that surprised you?
Sam: So the umbilical cord tissue was not particularly exciting. The umbilical cord tissue representing the baby. We found a few mutations and we were able to say, well in some pregnancies, going back to our original question, actually, the two daughter cells are completely separate. So one cell really does make all of the baby and the other cell probably makes all of the placenta, so they are completely separate things. In other cases, that distinction was not so clear and mostly it was a two third one third thing, as in one cell makes two thirds of the baby and the other cell makes two thirds of the placenta and vice versa with the one third that is missing. That's a bit complicated now. But that's what we found, and that sort of answered our question. But the really exciting bit which we did not expect to find, which completely blew us off our socks, is that when we looked at each of the mutations of each piece of placenta, they really looked like tumours.
Kat: Wow. So that's weird because obviously I've spent a lot of time researching and writing about cancer. And the one thing we know about cancer is that it is full of mutations and many more mutations than we'd expect to find in normal healthy tissue. And you're telling me that placentas, which I understand to be normal healthy tissue, are just full of mutations? What's going on?
Sam: So when we looked at the placental tissues, what we found is that, first of all, the number of mutations was incredibly high. If you sort of normalise it over per day of life because the placentas only lived for nine months, we ended up with mutations burdens that are very, very similar, in fact higher than most childhood cancers, which is quite extraordinary.
Sam: And the second thing that we found, which we didn't expect to find, we can look at what kind of mutations they are. So they are just the normal mutation that one gets when the cell doesn't divide quite right. And then the other mutations that have got a particular sort of signature, as we call it, and that signature also mutations in the placenta, again, very much was a childhood cancer signature, which was quite extraordinary.
Kat: So what does this mean?
Sam: It's interesting. So one could look at this and say the placenta just doesn't care. The placenta just hangs around for nine months and then dies. It's a very unusual organ. It comes out of nothing and then just goes away. So maybe therefore, Mother Nature doesn't actually care what happens to the genetic integrity of the placenta, and therefore it doesn't bother to maintain it. It puts its resources somewhere else. So that might be one way of looking at it. And therefore you end up essentially getting the Wild West in the placenta, and then it's over after nine months anyway.
Sam: Another explanation might be that perhaps just because the placenta has to divide so massively fast. I mean, if you think about it, you could argue, well it's extraordinary that we end up making a baby out of one cell, which is very, very true. But the placenta has to grow faster than the baby because it has to provide nutrition for the baby, particularly initially. So we end up in this period of quite ridiculous growth. And maybe during that period of ridiculous growth, it becomes impossible to maintain the integrity of the placenta.
Kat: I mean, it does make sense that if the baby is the thing that's going to hang around and contribute to the next generation and the placenta isn't, that we would have evolved to invest time and resource in repairing any mistakes in development in the baby and not really care so much about the placenta because it's on its way out, right?
Sam: Yes. And we found a little bit of evidence in that. So it's been known for a very, very long time that very occasionally the placenta, the whole of the placenta, can be genetically very abnormal. Yet the baby is just an utterly normal, beautiful, little healthy baby that goes on and becomes a healthy adult. And we were actually able to see that in our data.
Sam: There was one case where there was a baby that had three copies of a particular chromosome and should have only had two. And then what we found is that the additional copy was sort of shuffled into the placenta, almost pushed out of the baby into the placenta. And then in the baby, one copy was lost and you were back to what the baby should have, which is two copies of the chromosome to be able to survive. So it almost paints a sort of picture of the placenta as a dumping ground of human development.
Kat: I sort of love this idea of, let's just shuffle all the nonsense into the placenta because we're going to get rid of it at the end of it. And it does remind me, thinking about the evolution of this, the kind of evolutionary strategies when you compare things like long lived animals, like humans or elephants, with really short lived animals like mice. Because we know that short lived small animals, they tend to have really high burdens of cancer, but they live fast, they die young, they reproduce and then they're off to the next generation. Whereas larger animals, longer lived animals, we invest more time and resources in maintaining our bodies because they've got a last for a long time.
Sam: Absolutely. And actually, what you just said raises a really interesting question. What does the placenta of an elephant look like in terms of mutations? What does the placenta of a mouse look like in terms of mutations?
Kat: I think it's probably going to be easier to get mouse placentas than elephant placentas,
Sam: Possibly, unless you have got very good friends at London Zoo.
Kat: That's an experiment for you! write the grant now, we'll do it? I find this stuff just mind blowing that the more we have the tools and technologies to really dig into DNA and mutations in smaller and smaller amounts of tissue, just the more complexity we uncover it really does blow my mind. So where do you want to go next with this? What are the next questions that you're really interested in answering?
Sam: So to me, it throws up two questions. The big elephant in the room really is do these genetic abnormalities in the placenta associate or predict pregnancies that go wrong? You know, that's a critical question about maternal and fetal health and about pregnancy. Which we haven't been able to answer. Simply, we haven't got the power in that initial study to do it. But we now know how to do this properly, and the way to do that properly is, because each piece of placenta was different from the same placenta, we will have to look at lots and lots of pieces from the same placenta.
Sam: And the other thing that we have learnt is that we really have to use that very detailed sequencing, the three billion letter reading whole genome sequencing, to really capture all changes. And I think that gives us a basis to roll out this approach to a very large cohort. Again in collaboration with Gordon and Steve, who've done this wonderful collecting of placentas and everything we need to know about the pregnancy to be able to answer that question.
Sam: On the cancer angle, is the childhood cancer research, which is what drives you most of the time. The fundamental question that I'm interested in is exactly what he articulated. How can a piece of tissue be genetically a cancer? Yet it is entirely normal, and it is not a cancer. Beyond the placenta, that's something that we've seen in other situations in children as well. That normal tissues of children harbour really, you know, quite classical cancer causing mutations, Yet they're just normal tissues.
Kat: It turns out that life is a lot weirder than we ever thought, isn't it?
Sam: It is. And this is the beauty of sequencing that we can sort of begin to unravel the genetic code of life, of Mother Nature and all her peculiarities. All this stuff was completely unknown before we are able to read the code, and this is really where the power of this technology lies.
Kat: Sam Behjati from the Wellcome Sanger Institute. And if you’re interested in learning more about mutations and cancer, the extent of mutations in normal, healthy bodies, and the risks of cancer in different animal species then check out my recent book, Rebel Cell: Cancer, evolution and the science of life, available in paperback, ebook and audiobook from all good - and all evil - bookshops.
References:
