Scientist Stories: David Haig, Evolution of Imprinting

So thank you, Tara. You left out the most important part of my biography. My mother was a science teacher, and that probably has something to do with me being here today. So after Charles Dickens, I am calling this talk A Tale of Two Mothers, perhaps portentous. It may be the best of lectures or the worst of— lectures, but I'm going to start off with talking about two mothers who have contrasting reproductive histories. The first mother had 10 children in 17 years, and 7 of them survived her, and her inter-birth intervals were 2 years, 1, 1 year, 2, 2, 1, 2, 1, and 4 years. She was a very well-off mother. She was using a wet nurse, and this probably contributed to the short inter-birth intervals. Her age at first birth was 31 years. Age at last birth, 48 years. And then she lived to a ripe old age of 88 years. The second mother had 9 children. In 31 years. 7 of them survived her, and because they were more spread out, the inter-birth intervals were 5 years, 5, 4, 4, 3, 4, 2, and 4. Her age at first birth, to fit them all in, was 13 years. Age at last birth was 44 years, and she died 2 years later at age 46. So there's some striking similarities and differences between these two mothers. The first mother is this woman, Charles Darwin's wife, Emma Wedgwood, or Emma Darwin. So we were looking at the reproductive history of the Darwins. And the second mother is Fifi, daughter of Flo. So Fifi is by far the most successful female chimpanzee in terms of number of surviving offspring they have data from. At Gombe. This is— Fifi was also a privileged mother. Jane Goodall was handing out bananas in the center of her mother's Flos territory. So she was probably a very, very well-fed chimpanzee. And she sort of converted that into a lot of surviving offspring. This photo's interesting. This is a typical chimpanzee family. Here she is with her infant who would be suckling. She would suckle that infant for 5 years, roughly, until she becomes pregnant again. At that time, her previous infant, well, that infant then becomes an independent older brother or sister who can wander around feeding itself. And mother then has a newborn offspring. OK, so this is just comparing their life histories. Picking the age of 40 years, Emma Darwin at age 40 had 6 dependent offspring, plus Charles, who took a lot of taking care of. Fifi at age 40 had 7 living offspring. 3 of them were already parents. 2 of them were semi-independent juveniles, and then Flirt was was her dependent offspring that she was suckling at that time. But she also had 4 living grand offspring at that time. Notice here that one individual, Fred, is both her offspring and her grand offspring. So with DNA fingerprinting, we know that this is the product of an incestuous mating with Frodo, her rather fearsome one of her old sons. It would have been too good to be true if it had been Freud, her oldest son. But Freud, as far as we know, has had no offspring. Okay, so Fifi had 4 offspring after becoming a grandmother. So she was a grand— she was still becoming pregnant, having children when her daughters were having children.. And so that's rather unusual for human beings. So pregnant grandmothers are relatively rare among women. Women usually stop having their children at about the time their daughters start having children, whereas in chimpanzees there's much more overlap in the reproductive histories. So this is just a comparison of Emma and Fifi, pointing out some of the differences between humans and chimpanzees in their reproductive histories. One of them is that Emma took a long time until she started reproducing, not until her 30s. That was sort of an unusually long period, whereas chimpanzee mothers start to reproduce often at younger ages. In Fifi's case, at age 13. Then, when Emma started having offspring, she was having them much closer together than a chimpanzee can manage. Typically, a chimpanzee can only manage one surviving offspring every 5 years, whereas you think about the women you know, once they start having children, they're often having a child every couple of years. So human mothers are able to fit their offspring closer together. And I won't talk about it, then there's this very unusual long post-reproductive period that we see in human females after menopause, where they're still vibrant, fit, but not reproductive. And there's no evidence of that sort of postmenopausal period in the other great apes. So Emma started late, but then she produced her babies rapidly. And just pointing out that Fifi is an outlier among female chimpanzees in having 7 surviving offspring. Here is some comparative data looking at age at weaning. So this is the cessation of breastfeeding in our closest relatives. And so in the great apes and in natural fertility human populations, the typical reason for weaning is that the mother has become pregnant again with the next child. So in orangutans, they can only produce a second child after suckling for 7 to 8 years. So very long inter-birth intervals. Gorillas, wean at about 4 years, chimpanzees at 5 years. Part of Fifi's reproductive success was that she was returning to fertility sometimes a bit shorter than that. And humans typically in natural fertility populations are weaning in about the 3rd year of life. So this says that a derived part of human life history is early weaning compared to our ape relatives and the inferred ancestors because we're most closely related to chimps. Our ancestors were actually suckling their offspring for longer than human females do now. And that's just not in, you know, modern developed countries, but that's our records for hunter-gatherers, for agricultural populations and the like. If we look at age at first birth, An orangutan mother produces her first offspring at about 15 years. So think of what she's got to do if she were to have 7 surviving offspring. She has the first one at 15. 7, 8 years later, at age 22, she has a second offspring. To produce 7 offspring, she would need to be reproducing into her 60s. This is part of the conservation problem with orangutans. That they have very slow reproduction, but we also have records for orang mothers out in the field who are having offspring into their 50s. Gorillas at 10 to 12 years. Chimpanzees usually start reproducing around 13 or 14 years. Humans, of course, are very variable, both when they become fertile and when they first start having children. The age I've got here is from the Ache, which are a hunter-gatherer group from Paraguay, while they were out in the forest. And there, the median age at first child for women was 19 years of age. So, we wean our offspring more rapidly than the great apes, but we tend to start reproducing at a later age. So humans wean early, but mature late. Early weaning is associated with providing offspring with supplemental foods. So a chimpanzee, when it is weaned, already has fully erupted dentition, teeth. It can go out and feed itself. Whereas when we wean human babies, they're not able to feed themselves. We've got to produce specially processed food, different from the adult diet, to feed them. So overlapping, so the distinctive thing that human mothers do compared to the great apes is that they're sort of looking after more than one offspring at the same time. You know, a chimp mother does not become pregnant again until her previous offspring is more or less able to fend for itself and look after itself. Whereas human mothers, as we know, are often having 2, 3, 4 or perhaps more dependent offspring at the same time. So anthropologists believe that overlapping care was only possible because mothers had help and were not provisioning offspring by themselves. And so over the decades, who's providing that extra help to mother has different fashions as who it is. Back in the '60s and '70s, it was obvious that it was me, man the hunter, going off slaying great chunks of wildebeest and bringing them back to my nuclear family. Into the '80s, sisterhood was powerful. You know, it was other females helping each other reproduce. And now the current fashion is grandmothers, those old women who are not reproducing, helping their daughters to reproduce. So it's probably a bit of all of those stories. The important thing is that human mothers, because of the social groups they sit in, are able to produce offspring at a faster rate than the great apes. Okay, so I'm— this is all a prelude to talk about genomic imprinting, which is a subject that I've devoted too much of my life to. And this— I'm a theoretician, so this is the way that I look at mothers and offspring. You have a mother here who has two sets of genes and an offspring who has two sets of genes.. And one of those sets is shared between them. These are the genes that the offspring inherits from its mother, but the offspring's also got a different set of genes that it gets from its father. And so a lot of my work has been looking at evolutionary conflicts that exist within the genome between genes of maternal and paternal origin, particularly with respect to acquiring resources from mothers. Because maternally derived genes are also sitting in the mother, they're selected to, you know, ration the rate of demand on the mother's time and attention that she is giving to that child. But within the same child, we're getting genes of paternal origin, which are selected to maximize the survival of that particular offspring. And I'm going to start off talking about the evidence for this sort of internal conflict in prenatal development, and then I'll start looking at possible effects of it during childhood. So the prediction is that paternally derived genes— so this is for prenatal development— in fetuses will favor greater demands on mothers than will maternally derived genes. And genes of paternal origin will be favoring greater growth of the placenta, and of the attached fetus. And so I'm going to talk about one of the best studied imprinted regions in the human genome, and I'll tell you what imprinting is in a moment, and look at some of the disorders associated with it. So this is part of human chromosome 11, technically 11p15.5. And what I'm representing here is this is a maternal chromosome, and this is the paternal chromosome. So we each get, for each chromosome, one copy from mother and one from the father. The most famous gene on this chromosome is this gene here, which is insulin. But sitting next to insulin is a gene called IGF-2, for insulin-like growth factor 2, and this is a fetal growth promoter. You can think of it like an accelerator on fetal growth, and it shows an interesting pattern of expression. The copy of this gene that you get from your mother is silent, is not being transcribed and is not being expressed, whereas the copy you get from your father is being expressed. So here we have the chromosome coming from the father is producing a factor that is promoting fetal growth, but that factor's not being produced from the maternal chromosome. Did you have a question? Yeah. Is this only in the fetus, or does this continue throughout life? This imprinting here does continue into postnatal development. But at the moment, I'm concentrating on fetal growth. Just to point out here that this is because this gene was sitting in a female body in the previous generation, it's not being expressed in this generation. So history matters. If this were me and I passed on my silent maternal copy to my children, it's going to be active in my children because they're getting it from the father. So this activity-inactivity state isn't anything to do with the DNA sequence, but it's some sort of imprint that is put on the DNA that is inheritable from parent to offspring. Affects the expression of the gene, but it's an imprint that can be re— erased and reset. Further down here, we've got this gene CDKN1C. This is a cyclin-dependent kinase inhibitor, so it's inhibiting progression through the cell cycle. So this is a growth inhibitor, and it shows the opposite pattern of expression. The copy that you get from your mother is being expressed, is slowing down growth, whereas the copy that's coming from the silent— from your father is silent. So just sort of stripping away all of the other genes, we've got a situation where on a paternal chromosome IGF-2 is being expressed. So that's a paternal foot on the accelerator of growth, whereas on the maternal chromosome we've got CDKN1C being produced— and this is a maternal foot on the brake. And so simultaneously, you've got these conflicting forces on the amount of fetal growth. Okay, so what are the consequences when this system goes wrong? So I'm going to start off with the example of Beckwith-Wiedemann syndrome. So this is a fetal overgrowth syndrome. I got sent this baby photo by a PhD student. Who had Beckwith-Wiedemann syndrome. So this is her as a newborn baby. And so by the nurses' and doctors' hands, you can get an idea that she's a rather large child at birth. Some prominent features, she has a very long, large tongue. So the tongue muscles are hyperdeveloped. She also has an umbilical hernia so that some of the internal organs are protruding and that needs to be corrected. So she was a full-term delivery by Caesarean section, and at birth she weighed 11 pounds 5 ounces, or 5 kilograms, so she was a heavy baby. She had macroglossia, which just means a large tongue, and an umbilical hernia. And here is a baby photo of her as a somewhat older and happy child. So Beckwith-Wiedemann syndrome is associated with fetal overgrowth. These grow up to be large individuals, but not enormous individuals. So the effects are primary— primarily on prenatal growth. Okay, so what causes Beckwith-Wiedemann syndrome? So some individuals with this syndrome have what is called paternal uniparental disomy. So they've got 2 copies of this chromosome region, but they're both paternally derived. Both come from the father. So in this case, they've got two paternal feet on the accelerator. They're producing two doses of insulin-like growth factor 2. This is associated with a high risk of a variety of pediatric childhood cancers. Some other individuals have disorders of imprinting, so they have a maternal and paternal chromosome, but the maternal copy of IGF-2 is inappropriately reactivated, and so once again they have, if you like, two feet on the accelerator of growth. There are familial cases which are caused by mutations of the maternal copy of the CDKN1C gene. So this is the growth inhibitor. On the maternal chromosome, it's inactivated by a mutation.. So they're not producing any functional product from this chromosome, and the paternal copy is inactivated by the imprinting process. So these have an interesting genetics. If you inherit the mutation from your mother, you have Beckwith-Wiedemann syndrome. If you inherit the same mutation from your father, you have a normal phenotype because the mutant gene is inactivated by imprinting anyway, and you have a normal copy coming from the mother. Okay, so now I'll look at the opposite sort of phenotype. This is Silver-Russell syndrome, which is an intrauterine growth retardation syndrome. And so here we have a picture of two girls who are pretty well the same age. This girl has Silver-Russell syndrome. It's a diagnosis that covers a variety of genetic causes, but I'm just going to focus on a subset of them. So it's associated with intrauterine growth retardation, so small babies at birth. Also, postnatal growth retardation. So these grow up to be tiny adults if not treated. Triangular face, variety of things. Some individuals with Silver-Russell syndrome have maternal uniparental disomy of this chromosome region. So they have 2 copies of the maternal chromosome and no paternal copy. So both copies of IGF-2 are silent, so they don't have any foot on the accelerator, and they've got two feet on the brake, and that's associated with growth retardation. And then there are a variety of other causes that I won't talk about. I now want to start to become more speculative and look at the effects of imprinted gene expression in postnatal arm growth and looking at this typical human family here. This is unlike a chimpanzee family in that we have these three dependent offspring here, Bart, Lisa, and Maggie, that are dependent on their mother with perhaps some assistance coming from this guy here. So let's think about Marge's reproductive history in a natural fertility population. So Marge, at age 19, we're going to start there, she has Bart. And 3 years later, she weans Bart when she's pregnant with Lisa. And then another 3 years later, she's weaning Lisa because she's having Maggie. And this is about the time which Bart is going through a process called adrenarche. The adrenal glands start to produce a variety of androgens in both girls and boys. This is a period called the 5 to 7 year transition. This is when a child— we're sending them off to school at about that age. Children— I think of that transition as when your child no longer wants to hold your hand in the street. They want to appear to be independent. They sort of focus more on peers rather than parents. So I think that what human mothers are doing is sort of in this period in which a chimpanzee mother would be only having one offspring, they've been able to fit in two offspring into that period. So the distinctive human traits are the prolonged childhood and juvenile period. So we have this very prolonged childhood period compared to other primates. Then this is associated with a pubertal growth spurt. So that we're growing at a really, really slow rate. So our growth is not limited by nutrients. And then suddenly when we're coming into puberty, we shoot up in height. And that is a unique human attribute. Chimpanzees don't show a pubertal growth spurt. In an increase in height at the time of becoming sexually mature. This suggests to me that we've actually evolved to stay small, small sort of infants, until the time that we're entering the reproductive period. We have delayed first reproduction, and then once we start to reproduce, we— I'm putting myself with the women in the room here— have short birth intervals, We have early weaning relative to the great apes. We produce what I think of as staggered litters, having multiple offspring of different ages but dependent at the same time. And then the post, prolonged post-reproductive period. Okay, so my hypothesis is that the evolution of this life cycle was driven in part between different conflicting forces acting on parents, particularly on mothers and their offspring. So I'm going to argue the prolonged childhood juvenile period was an adaptation of offspring to learn about the world in comparative safety while being looked after by their mothers. You know, they're outsourcing some of the costs of keeping themselves out of trouble to have a childhood where they can accumulate experience to become more effective adults. So the pro— getting longer childhood is enhancing offspring fitness, but it's reducing the total number of offspring that a mother can have. So the maternal counter-adaptation to this is to wean her offspring early and return to fertility sooner. And this is a maternal counter-adaptation to the extended juvenile period. And to find evidence for this hypothesis, what I'm going to do is I'm going to look for conflicting actions of genes in the child on the age at weaning and also potentially at the age of entering into puberty. Okay, so my predictions are that paternally derived genes in a child will favor delayed maturation, so growing at a slower rate and late weaning, whereas maternally derived genes in the child will favor earlier maturation and earlier weaning. Okay, and so to provide some evidence for this hypothesis, I'm going to look at this particular syndrome. So this boy has what is called Prader-Willi syndrome. He has a deletion of part of chromosome 15, and specifically, he has a deletion of the paternal copy of that chromosome. This boy has a syndrome called Angelman syndrome, which is caused by the same deletion, except in this case it is a deletion of his maternally derived chromosome. So you can have the same chromosomal deletion causing different phenotypes depending on which parent it is inherited from. Now, I'm not going to say much about Angelman syndrome in the main talk, but there may be time if we have it in the questions afterwards. It used to be called Happy Puppet Syndrome. These children are noticed by a smiling, laughing demeanour. It's the most distinctive feature that they're always laughing, smiling. It's associated with profound mental retardation and a complete absence of speech and gestural communication. Prader-Willi Syndrome, as we will see, is associated with a very different phenotype. So just looking at the causes of Prader-Willi Syndrome, So it's associated with a particular region on human chromosome 15q. This region has some genes, for example, this one here, SNRPN, that is expressed off the paternal chromosome but not off the maternal chromosome. It's also associated with genes like this one here, UBE3A, that is expressed off the maternal chromosome but not the paternal chromosome. Chromosome. Just out of interest, sitting here in the region, this is where the single nucleotide polymorphism that explains the difference between blue eyes and brown eyes sits in European populations. Okay, so Prader-Willi syndrome is associated in many cases with a paternal deletion of this region. So they only have a maternal copy of the chromosome So any genes that are only expressed off the paternal chromosome, they don't have any copies of those genes. Some other individuals with Prader-Willi syndrome have maternal uniparental disomy, two copies of the maternal chromosome and no paternal chromosomes. So once again, they're lacking any products from genes that are only expressed off of the chromosome when you get it from your father. Interestingly, almost all of these individuals with this particular diagnosis develop psychosis into their teen years or early 20s. And so there's a parental origin effect there that's not present in the deletions. Okay, so Prader-Willi syndrome is associated with the absence of expression of paternally derived genes. Therefore, evolutionary theory suggesting that Prader— Willis syndrome should exhibit an exaggeration of traits that reduce demands on mothers. So the paternally derived genes are enhancing demands on mothers. You take them away and you're going to see behaviors that reduce those sort of costs to mothers. So the phenotype before birth and immediately after birth in Prader-Willis syndrome, it's associated with reduced fetal movements. Neonatal hypotonia. So these tend to be floppy babies with low muscle tone. And it's associated with poor suck. So these children generally have to be gavage-fed with a tube into their stomach to give them adequate nutrition. So they have an absent or very, very weak suckling reflex. They show excessive sleepiness. They sleep a lot. So the inferences, which are somewhat unsurprising, is that maternal costs are reduced if babies suckle less. So we're seeing a complete absence of suckling response in these children, and also if a baby sleeps more. By contrast, in Angelman syndrome, which is the opposite deletion, those babies show excessive wakefulness. In case reports, you can find babies that are awake 21 hours out of 24. 4 hours. They're quite different, more difficult children to deal with. So, so far so good for the theory. There's a problem, but the Prader-Willi syndrome phenotype shows an interesting change. So, so far I've been saying that these children are anorexic. They have no appetite. You've sort of got to force-feed them. But that changes during childhood. So they start to develop hyperphagia, Perhaps from the second year. There is some debate about this. This change is potentially occurring at about the time the child would have been weaned in the past. They show a non-fastidious appetite, which is unlike any children I know. They will eat anything that you put in front of them. And they will go out in search of food. There are anecdotes of them eating the dog food and eating food from garbages and stuff. This is unlike typical children. They develop obsession with food and foraging behaviors, so part of the treatment is putting locks on cupboards and refrigerators to cut down the caloric intake. They develop massive obesity, which is superimposed on growth retardation, so they have a low lean muscle mass, they have short stature, but all of the food that they're taking is being deposited in fat. They also have very, very low energy requirements. So on a diet which an obese person would be losing weight, a Prader-Willi syndrome individual is maintaining their heavy weight. And a lot of the health problems in Prader-Willi syndrome are associated with this excessive obesity. It's associated with incomplete puberty and a variety of other symptoms. Interestingly, the hyperphagia, so that's the obsessive eating, is not present in any of the mouse models of Prader-Willi syndrome. So they produce mouse models where all of those genes are deleted. They recapitulate the poor suckling and the sleepiness and things like that. But the obsessive appetite and overeating is not a feature of non-human models of this syndrome. And so I'm suggesting that the curious switch from anorexia to hyperphagia that we see in Prader-Willi syndrome reflects genetic conflicts associated with the weaning process. While the infant was dependent for nutrition on the mother's breast, paternal genes are promoting suckling and maternal genes inhibiting it. But at the time of weaning, when the mother is attempting to switch the child over to other sources of food, paternal genes are inhibiting taking those supplemental foods. The child is uninterested. You remove those genes and you take the foot off the, foot off the brake and you get this hyperphagic response. So my conjecture is that paternally derived genes promote suckling and that paternally derived genes also inhibit the appetite for for supplemental foods. And so when you lose them in the deletion, you're getting this obsessive appetite. And the evolutionary reason for this is the children benefited from delayed weaning at cost to the mother's fitness because milk was a superior but more costly food for mothers to supply. And also because suckling prolonged the inter-birth interval, prolonged the period of time that the previous child had before it had to compete for maternal attention with a younger sibling. And I think I will leave it there and answer any questions that you might have. Yeah.

I'm a little bit confused. So if my child didn't suckle well and slept all the time, while that might be an advantage to me on one level, that would be an enormous disadvantage to me on another level, my stress level would be incredibly high.

And I'm sure that that would have tremendous effects on me. Yeah, so these— so what these genetic disorders are showing— so in normal— in a normal child, you've got a maternal and paternal chromosome sitting there, and they're pushing— let's take sleepiness for an example, because Angelman and Prader-Willi syndrome go in opposite directions. And so the maternally derived chromosome is trying to put the baby off to sleep a little bit sooner, whereas the paternally derived one is wanting to wake that baby up during the night to suckle more or whatever, demanding more attention. And the baby's sleepiness actually expressed is determined by the balance of those two forces. I've compared it to a tug of war that you've got two sort of sides pulling in opposite direction. What you have in Prader-Willi syndrome is one side stops pulling. You've taken out the paternal chromosomes, you get excessive sleepiness. Or in Angelman syndrome, you take out the maternal chromosome and you get this hyper-wakeful child. Now, that phenotype is not going to be adaptive for anybody in that situation, but it's giving you clues about the genetic architecture of sleepiness in the child and the evolutionary forces that were acting. Yeah. So would you assume that like for this evolution to happen, that the mother would stay with one father or would you have multiple, would the different siblings have different parents? Yeah. So the theory of parent offspring conflict that I'm talking about in explaining the origin of this genomic imprinting is presupposing that sometimes mothers are having offspring by multiple fathers. If all offspring were always one man and one woman, the relationship to the other children of the mother for genes of maternal and paternal origin is the same. So this depends upon some siblings who are competing for maternal resources having different fathers. And we just know that to be the case. Relationships break up. People have affairs. Fathers die and mothers remarry. For all of those sort of reasons in In the evolutionary past, women were sometimes having offspring with multiple fathers, and so therefore genes of paternal origin were being selected to make the most out of extracting resources from the mother while in a sense they haven't access to that maternal care.

Yup, and then there. I just wonder how do you get the maternal uniparental disomy? This has to be what, a failure in nondisjunction meiosis or something?

How do we get that genotype? Yeah, it's a really good question. And we don't know in every case. But what— well, we do know in some cases that what has in fact happened is that there was nondisjunction in maternal meiosis. You had a trisomic conceptus, and then one of the three chromosomes was eliminated to bring the child down to being disomic, but it was the paternal chromosome that was eliminated. So there, you only need, technically, you only need one non-disjunction event at meiosis, but then you need a chromosomal loss early in development. And so sometimes you will find the trisomy, for example, present in the placenta, but not present in the child. There's accumulating evidence that human embryos have a certain ability to compensate and correct for extra chromosomes. You know, so if the complement is sort of unbalanced, you have these processes of losing chromosomes to get back to a balanced complement, and that could be how it comes about. I should emphasize, you know, uniparental disomy is not a common event. You know, it's picked up in these genetic syndromes. You had a question.

You had said that if it was your genes that were passed down, they're paternal. However, if you received them from your mother, they were maternal.

Is there something that changes in the genetic structure that now says yours are paternal? What actually is— Yeah. So for molecular biologists, you know, how is it done? What is the molecular mechanism of genomic imprinting? It's not fully understood. Clearly, a process called DNA methylation has something to do with it. Because genes that interfere with the DNA methylation process can interfere with the imprinting process. And so in DNA methylation, cytosine residues can exist in two forms, methylated cytosine or unmethylated cytosine. So there are enzymes that can put those methyl groups on. There are also enzymes that can take them off. And there's a process when DNA replicates that can allow the methyl— the pattern of methylated cytosines to be copied through cell divisions. So the pattern of methylation can be inherited, passed from parent to offspring, but it can also be reset, the methyl groups removed When my maternal chromosomes are transmitted to my children, they're going to get the paternal imprint on them. So in my germ cells, the maternal imprint has got to be erased and paternal imprint put on. Having said that, there are examples of imprinting in the human genome where DNA methylation does not appear to be involved. So there must be other mechanisms that can do this as well. Yes. Actually, could you say something about what percentage of genes are thought to have Sexual imprinting? So what proportion of the genome is subject to imprinting? That's not well known at the moment. In humans, it's around in the order of 100 definitely known imprinted genes. There's some evidence that the proportion of genes may be substantially greater than that. But we're talking of a genome of over 30,000 genes in it, and the majority of genes are not imprinted, which makes sense because classical genetics, which assumed that parental origin didn't make a difference, was really, really successful for quite a long time. So for most of the— you know, if we think about blue eyes, if you've got a gene for brown eyes from your mother and blue from your father, you have the same phenotype if you switch the parental origin. So it's only a small subset. One other thing that you can see there is that these imprinted genes do occur in clusters. The best way to find another imprinted gene is to look at the gene next to something that you know is already imprinted.

You had a follow-up, yes?

And could you say something about species that imprinting occurs in? Yeah. So the— for example, in both of these clusters, the imprinting is conserved between mice and humans. Okay, so it goes back that far. We know for IGF-2 that it's imprinted in marsupials, but it's not imprinted in monotremes, in platypuses and echidnas. So at the moment, there's, you know, the evidence is weak, but we definitely know it's present in marsupials, which are giving live birth. We don't know if it's present in monotremes at the moment, which are egg-laying mammals.

There was a question down here.

Yes, ma'am.

Just to follow up on the ones that we do know, are they all related to human development, fetal or after fetal?

I mean, the ones we do know? The current— there are two tissues where we know of substantial effects of imprinted genes, you know, that recur. One of them is the placenta. So there are a lot of imprinted genes in the placenta, that's clear. The second is in the brain, and so there are a number of imprinted genes in the brain. And the Prader-Willi syndrome genes, their imprinting is brain-specific. And in some cases, we really don't know what is going on. Just another interesting case. So we talked about uniparental disomies. Maternal uniparental disomy for chromosome 14, all individuals who have that go through precocious puberty. So that is suggesting that there's some paternal inhibitor of pubertal progression that's absent in those individuals that is bringing on early onset of puberty.

You had a question. Yes, so this gene in the mothers that causes a increased growth in the fetus is somehow related to an insulin gene. Diabetic mothers often have large babies.

Do they have something funny with their maternal— Yeah. So first of all, it's the paternal copy that's promoting growth, not the—

of IGF-2. And if in a diabetic mother this is not turned on instead of inhibiting growth, like would a diabetic mother be diabetic because she has a gene somewhere for insulin that should be turned the way it isn't supposed to be?

[Speaker:JOSHUA ELKINGTON] So let's come back here. Should have been a simpler way of doing this, but have I got the right one? Yeah. Okay, so sitting here, so here is the insulin gene and IGF-2 is related to insulin. There are some parent of origin effects in diabetes that map to this region. So there is some hint that there are parental origin effects caused by genes in this region that are predisposing individuals to developing diabetes, and that depending on which parent you inherit the chromosome from, it can have some sort of effect on diabetes. The overgrowth that you see in diabetic pregnancies, there appear to be two factors going there. One of them is because of the mother's diabetes She has high blood glucose levels, so in a sense, the fetus has a nutrient-rich food, and so that's particularly, you know, high weight in the baby. But there seems to be something intrinsic to the baby that is promoting greater growth. Insulin itself can bind to the same receptors as IGF-2, so it may be having some sort of effect, but it's a bit— It's still a bit unclear, but there's clearly something in diabetic pregnancies going on here in the mother metabolically, but there are also some effects occurring in the fetus at the same time. Sorry?

That's why babies crash when they're born.

The fetus supplies the insulin. Yeah. So that's a different, you know, I'm interested in You know, gestational diabetes. So women often develop diabetes during pregnancy. That's being caused by hormones coming from the placenta going into the mother's blood. They're increasing maternal insulin resistance. If that went unopposed, maternal blood glucose would go higher, providing greater nutrients for the fetus. Mothers respond to that by increasing their insulin production during pregnancy. So you have this antagonistic effect. You've got a hormone coming from the placenta that's creating insulin resistance, and the mother's pancreas ramps up its insulin production to counteract that. Now, some women, their pancreas can't increase insulin sufficiently. They develop gestational diabetes mellitus. They have high blood glucose during sugar, but as soon as the placenta is delivered, the diabetes resolves.

So it's a related sort of issue going on there. Yes. So follow-up, I've always wondered for children born of diabetic mothers, for the rest of their lives, are their insulin production, once they've been delivered, going to normal? And I understand her crash comment that the fetus crash, or once the baby's born, they crash because they're using a certain amount, a high level of insulin to take care of all that glucose, and then if you don't do a line right away, they—

But what's long-term sequelae of the grown-up child of a diabetic mother? I'm not an expert on that. The— What I do know is that the women who develop gestational diabetes, so that's diabetes that was not pre-existing at the time of pregnancy, It's probably that they have beta cells in the pancreas that aren't able to mount a good insulin response. They then do have a greater risk of developing Type 2 diabetes later on in life. So in a sense, you can think that the pregnancy is revealing the compromise within, within the pancreas. But we don't know if that's passed on at this point. You know, all of these, you know, there is familial tendencies in all of the diabetes, but, you know, disentangling what way the correlation goes, you know, I'm just not an expert in that subject.

I wondered if this— these areas you're looking at, are there multiple alleles for this? So we see a range of values in both? Or is this sort of you've got an all or nothing kind of a gene and it must be sort of a—

it's obviously codominant in their relationship. Well, let's talk about— let's talk about this sort of situation just in terms of dominance and recessive. This has odd sort of inheritance. So that if, so suppose you've got some, if you inactivate, if you had an inactivating mutation in IGF-2, you know, that's associated with reduced growth. And so if you inherited, if your father had that gene, all of the offspring who inherited it would have the reduced growth because only the paternal gene is expressed and it's got the mutation in it. So it would look like it had dominance inheritance. But then his daughters who were affected, when they pass it on, none of their children are going to be affected because the mutation is on the copy that is inactivated. And so in looking at pedigrees for these imprinted genes, you know, in parts of the pedigrees they look like they're dominant,. And in parts they look like it's recessive. And you only start to make sense of it when you start looking at what is the sex of the parent transmitting it. So that's one thing I would say about dominance and recessive. On the question you ask about polymorphism, you know, one of the ways you maintain polymorphism is through some sort of heterozygote advantage. And in imprinting you don't get that because only one of the alleles is expressed. And so that actually creates some difficulties in maintaining polymorphism. But that's an area of theory that I'm working on at the moment with some of my students. Have I exhausted you? As you can tell, I can talk and— At Macquarie, we used to get— mature age students would come in from all around Australia And so we gave them really, really intense teaching. So we'd start on Friday at 9 o'clock, go through to 9 o'clock at night, and then do that again on Saturday, and then again on Sunday. So that was my training in teaching. And one thing I learned was that I could keep talking when the students were just saying, no more, no more, we've had enough. Thank you.

Oh, a question, yeah.

Area 11p15.5, how big is it in terms of base pairs and in terms of the number of genes on it? I can't give you the actual— so yeah, this is sort of— so I focused on these two genes, but if you'll notice, there are other genes sitting in here. You know, this one's expressed off the paternal and not on the maternal. So it's a substantial cluster of genes. You know, in the order of 10, 12 genes. You know, that's not a large part of the chromosome. The P, for those who don't know, refers to the short arm of the chromosome. So this is on the short arm of chromosome 11. 15.5 is just sitting near— this is sitting up really quite close to the telomere on that chromosome. So it's a sizable chunk, you know, a few megabases, but I can't do any better than that. Other questions? Okay.

Thank you.