A live conversation between
Stylianos Antonarakis, iGE3 Institute of Genetics and Genomics of Geneva
and
Jesús Argente, Universidad Autónoma de MadridIssues
With contributions by George Chrousus, Ana Claudia Latronico, Stefano Cianfarani and Christos Yapijakis.
Editor: Ivo Jorge Prado Arnhold
Transcribed from the live conversation by Esther Marten
Jesús Argente: We, pediatric endocrinologists, take care of patients and follow all of the progress in genetics and all of its new brand techniques. This allows identification of new diagnosis and new therapies. Therefore, we need to deal with this new knowledge and need to study hard to better understand what genetics can offer to clinicians.
So in a very informal manner we would like to have a conversation, a discussion in relaxed manner about the relevance of the progress that genetics has made in pediatric endocrine diagnosis. To talk also about the necessity of organized genetic laboratories, how many genetic laboratories we need in one country, even in the states, how many laboratories do we really need to make any advance in the area of one specific genetic diagnosis. Also about the analysis of the different techniques, where to apply what technique. So how the clinician should decide with every patient with a genetic component, what technique could be the best, the cheapest, maybe the most important one to make a proper diagnosis, so that we analyze the sample of the patients, and the parents later. Therefore, those are all of the questions that we would like to discuss. Of course, the most important thing for us are the patients, and when we are in front of a patient, we deal with a specific diagnosis and we deal with a potential approach to this kid, in order to offer our best. So Professor Antonarakis just to break the ice, what kind of reflections do you have with this small introduction?
Stylianos Antonarakis: I will answer the best I can all of your questions but I will leave them for the question period. And I will be very brief actually because I will talk about facts and not theories. I will give you an anecdote about briefness; my mentor was Dr. Victor McKusick at Hopkins for many years, and when I became an assistant Professor, he asked me to give a talk on the Mutations in Thalassemia (1981). Then he called me into his office and said, “young man I have this talk for you, and you talk about the variance in Thalassemia”. And I was delighted and asked Dr. McKusick, how much time do I have, he said, “15 minutes.” I turned this in my head and said It will take me one hour to say what I have to say. I said 15 minutes is not that much. He looked at me and said, “young man, if you cure cancer, it will take you five minutes to say it and the whole world will change.” then I learned my lesson and said thank you very much and went out and when I was at the door, I opened the door and then he said to me, “young man, I’ll tell you something else, when you propose to your future wife, how much time does it take to say yes or no, so I will try to be brief as crystals.”
So, I talk to you about facts and I start with the most important fact in genetics, the Genome. So what is the Genome? There’s a book of instruction, of information for each one of us of our future, of our destiny, of our development and our susceptibility to disorders. The Genome is big, tremendously big. It is 3.2 billion nucleotides from haploid Genome. So each one of us has a Genome, a finite Genome of approximately 6.4 billion nucleotides. Half from the father and half from the mother. 6.2 bases of gigabytes of information. Those in this iPhone I think are about 16 gigabytes of information. So in that iPhone, if I erase everything, all your WhatsApp that are on, I can put my Genome, plus the Genome of my Dad and part of the Genome of my Mom. So to give you an idea of how big the Genome is, but also how finite the Genome is, I represent to you in a kind of a book.
As you know the Genome is divided into several Chromosomes. The Chromosomes are like buses that transport the Genome from one cell to another, and from one generation to another. And its dividing Chromosomes, the smallest Chromosome, Chromosome 21, is about 1.2% of our Genome. Let’s say 1% for the sake of argument. And I convinced the Swiss Institute of Bioinformatics (SIB), to print Chromosome 21 for me, because I work with Chromosome 21 trisomy etc. So they did it. I give you the book, so you can take a look at this and appreciate. So this is a rare copy, the only copy that exists on earth printed. Its Chromosome 21 and it has 1470 pages and it has 248 nucleotides per line, letter, per line and has 31000 nucleotides per page and the total length is 46 million nucleotides and as I said its 1.2% of the human Genome. Try to appreciate, our Genome is about 80 times bigger. 80 books, you can put it in your library to enjoy it. And I argue that this is the most important text, the most important book for us humans, written. Perhaps a little bit more important than you can see the repetitive sequences etc. The letters are a 5 or 6 size in New York times text. This book is important. However, if you compare your Genome to that book, your Genome is a bit different from this. This is what we call the reference Genome that we had put after the sequence of the Genome in the database. This does not mean that this is normal Genome, actually there is no “normal” Genome.
If you compare your Genome to this, you will be a bit different. How much is the difference? The difference is that one in every thousand letters, thousand nucleotides is different between two randomly chosen Genomes in the population. So one in a thousand. So that means that for a 3+ Billion letters, you differ from another randomly chosen Genome about 3 million letters. Between 3, 3.5, 3.8 depending on your origin. If you’re an African, you have about 4 million differences. If you’re a European, about 3 million differences etc. Because the Africans are earlier, the first stalk of humanity, and therefore they have more time to accumulate variance. And not only some letters change, but also pieces of the text change. In other words, you may have from the copy you receive from your Dad, you may have a piece of DNA from one copy and from your Mom two copies, or three copies or four or five or ten, and that’s a normal variation and so we call this copy number variation. If we add the copy number variation with a single nucleotide variation, the letter changes, our Genomes are different form each other, for 0.9%.
So we are all Homo sapiens sapiens, because we are 99.1% identical, but we are 0.9% different. And that difference is fantastic. We are all different here, different intelligence, different orientations, different likings, different appreciation of the environment, and different susceptibility to disorders. So we are all different because when we copy the DNA, from one cell to another, or from a generation to another generation, the copy mechanism is extremely accurate. Fantastic actually, it is amazing how efficient and correct the copying of one DNA to another. I am talking to you for about five minutes perhaps and millions of my cells in the blood and the gut and elsewhere have been renewed, so I have copied all this DNA right now, and I pass it from one cell to its daughter cells. The copying mechanism is accurate, but not without mistakes. Every time we copy the DNA we make some mistakes. How many? One letter change in a hundred million letters. minus nucleotide, minus division. Every hundred million letters we make one change. So a new child that’s born from two parental Genomes, he or she has between forty and a hundred new variants, new mutations. So every time we do the genetic experiment, we add to the gene pool forty or a hundred variants. And that in long term, there are about two hundred thousand years of existence on this earth, has made us very different from each other and has given us the variability. And the variability is important because we can adapt to the ever changing environment and can evolve. If we were monomorphic, if the environment changes in a bad direction for us we would all disappear. But because we are polymorphic, then some of us, because of our genetic variation that we have, we survive.
That’s a gift of nature that the organisms that survive are polymorphic. And we should celebrate this, this is a fantastic gift. However, as you know it very well, there’s no free lunch. We pay for this. And the price that we pay for this very ability, for this possibility to evolve and adapt, is that sometimes the variation that we have is pathogenic. It gives us all the genetic disorders. And as Theodosius Dobzhansky said in a fantastic book that he wrote in 1973, Dobzhansky was a major figure in population genetics at Cal Tech. He wrote that “nothing in biology makes sense except in the light of evolution” and you can think a little bit further, “nothing in medicine makes sense, except in the light of evolution”. So because of this very ability, we can see medicine now a bit different, in a different way.
Now let me challenge you, I argue that the causes of disorders are twofold. The tree of medicine, the trunk is divided into two main branches. The branch of genomic variation, and the branch of environmental variation. Or, in the leaf of the trees, in the further branching, some of the branches go together again, and there is an interaction in the Genome and the environment. And I challenge you, you are in medicine for many years now, I challenge all of you to find one disorder or one phenotype or one trait, that is not due to genomic variation and is not due to environmental insult variation, or the interaction between the two. I tried very hard to find one and have not found any yet. So because of this, I and many others think medicine as either genetic medicine, or environmental medicine. And I’m not an expert in environmental medicine, so I’ll talk to you about genomic medicine.
What is genomic medicine? It is medicine based on the individual genetic variation. Now if you think about this individual genetic variation, some people call it personalized medicine. A tremendous misnomer in my view, because personalized medicine is not only based on the genomic variation, it could be based on other variations. For example, could give you a different personalized medicine than your genetic variation. So in my view, let’s not use this. Second, Francis Collins at the NIH convinced president Obama to use precision medicine in my view another misnomer, we always try to practice precision medicine. What has changed is the individual genomic variation that we can now appreciate and measure, and therefore, the evolution of medicine is because we understand better our Genome. The variation is extensive, and could be common, that many of us have the same variants, or could be rare. And the rare variation is the one that is of the low hanging fruit because we can identify the causes of millions of Mendelian phenotypes, and we can find what we call high impact variants, that if you have one, you have a disorder, or we can identify the low impact variants, for which you need several of those and several combinations of those in order to develop one of the late onset common human disorders.
Now the late onset of common human disorders is extremely important in medicine. Extremely important. Because people that are not pediatricians, and other specialties and the population are mostly interested in late onset disorders that they have. However, in evolutionary terms, the late onset disorders are not important. Because our species and every species, every bisexual species dies with menopause. So what happens after the age of menopause in our species is not important in evolution. That’s why we’re not protected from the common, low impact variants that predispose us or cause the common complex phenotype of adult age. And most of the pediatric endocrinologists are dealing with the rare high impact variants. Now, if you give me your sequence today and I can sequence your DNA, quite cheaply actually, if you like for me to sequence your protein coding genes, I will ask you to pay 300 dollars of reagents and wait for a day or two. If you want me to sequence your entire genome, you probably need to pay about 1200 dollars in reagents and wait about a week. But even if I had all this sequence, for how much of this sequence I can make an intelligent, medical verdict. I can say something that is useful for the patient and satisfactory for the physician because we identify the cause of a trait. 0.3% about 40 million base pairs, 40 mega base of sequence. 0.3% because if you look at the databases today, you’ll find that we know 4171 genes that that could cause if mutated in a bad manner a Mendelian phenotype. And for the other 16000 protein coding genes that we have, we don’t know, so they are not today medically reimbursed. And it’s not only that, but you and I have about 19000 non coding RNA’s and long ones and about 2000 short non coding RNA’s and we have about 70,000 promoters and have about 600,000 enhancers and transcription factor binding sites. Very functional parts of the Genome that we have a smell of, we know that they exist. However, we don’t know that their variation causes different phenotypes. So all the genetic medicine today, the sophistication that you may think we have is based on this 0.3% and so the medical genome that I call today is 0.3% and tomorrow we expect that the medical genome will cover about half of our genome (50%). Why half? half is what the encode project and several other projects, mainly funded by the NIH but also Europeans contributed to it, have identified that perhaps half of the genome may be functional. And the other half is repetitive and perhaps its important. Some people doubt that it’s important and in the next thirty years I don’t think that we will touch the repetitive elements because we expect to find more etiologies in disorders in the functional element. The 50%. The repetitive elements you could think of as of an accumulation of DNA’s over evolution.
I’ll give you an example; let’s say that I buy a house when I’m thirty and then at eighty I will die and then I accumulate things, many things in the basement, and then I give the house to my child and the child does not throw anything out because she thinks that its of value, and then she also in her lifetime accumulates another basement and then she gives it to her children, and the children continue to accumulate, and then the Genome that we have today as 50% so called “junk”. Evolutionary it’s important not to throw anything away because when we throw away, because you are blind in evolution, you may throw out something important among these things. So we accumulate, but we don’t throw out. So you see the first goal, first research goal of today, the medical goal of the Genome, is to go from 0.3% to 50%.
Of course our genome is also a history book. We can say something about our ancestry, where we come from, what is our line in time. And we can find that I have a part of my chromosome eleven, like five mega bases of Ashkenazi Jewish origin. And if I look at my ancestry I won’t find one, but if I do the calculation, the mutation rate I can predict, theorize that in 7030, there was a Jewish genome that went to my genome. To my lineage of genome and then I probably have another 2% of my genome Asian. Then in 1450, I can say because of the mutation rate and the variants in that particular genome, that I had an influence in somebody with that particular genome now entered the lineage. This is called recreational genetics. We don’t care about this in medicine, except that because of the sectorization of the populations that they’re inbreeding, several groups of people develop their own disorders sometimes.
Now, let’s go to medical issues. When you have a client, a patient, a patient is somebody that you see that has a disorder. But in genomic medicine, let’s think differently. Everybody is pre sick. Everybody has a pathogenic variation or pathogenic variation that will make her or him susceptible to develop a specific disorder. So let’s not talk about patients, let’s talk about humans. So if you have a human with a specific phenotype that you recognize, and then you give the DNA for sequencing, and we recognize the variants, if they are genomes, we cannot interpret them except for 0.3%, and in that 0.3% there’s a variant, and I use an example; 375 variants to sustain in humpty dumpty gene, and this variant is one among 23,000 variants that are defined in the genes, in the exome as we call it. The exome means the sequence of a protein coding fraction of the genome. So 1 in 23,000 is probably causative in the phenotype that you are interested in. And all the other ones are not. So the difficulty variant to find that one variant that is pathogenic among the sea of other variants that are not. And for this we rely on prior knowledge, prior probability, and prior knowledge for this kind of exercise. The databases.
We look at the databases and we find that this variant is found in a hundred different individuals that do not have the phenotype that you are interested in. Then you do the simple statistics, and then you say that this is not the causative variant. And you can use the same Aristotelian logic if you find nobody else or if you find it on everybody that has the same phenotype, you conclude that this is the causative variant. So the point is that the argument is mainly statistics. It’s a numbers argument. So for the majority of variants that we declare pathogenic, the argument is 99% statistical. Of course there are other arguments, if you have this variant in a model organism, mouse, or yeast, or zebra fish, or any other model organism, then you can study this variant and say whether it is pathogenic or not. But this argument, the model argument, the animal model argument is not as strong as the statistical argument. Somebody did the experiment in the BRCA1 Gene that causes if mutated deleteriously breast cancer in females and in some countries in males, and ovarian cancer in females too, somebody from Seattle, made all the variants that could exist in that particular gene, all of them. And then he used a viability of haploid cells and the finding was that about 30% of the amino acid changes are deleterious in that particular cell setting. However, a fraction of them are found in normal people, without any phenotypic significance. So the model organisms are fantastic but we have to use the information with caution. And then in everyday life people use prediction programs, programs that look at the amino acid substitution isoleucine leucine and valine if you interchange them then there’s not much of a difference because the side change is almost the same but if you have an arginine to a histidine a to a small one negative charge let’s say, the mutation is very serious if you have an arginine to leucine, both of them are positively charged the polar big amino acids, there’s not much of a difference so most of the programs of the prediction programs look at the amino acid substitution, they look at the frequency in the population, they look at the domain that they occur and they also look at the evolutionary conservation. An amino acid that is important is likely to be conserved. And a place of an amino acid that is changing very frequently, then is likely not to be important.
All these programs work that way, they give you an output that this variant is likely pathogenic, pathogenic, likely benign and something in between that is the nightmare to the geneticist. A real nightmare that’s called VUS (Variants of unknown significance), which are the majority of them today. The databases that you can consult for this prior knowledge, they exist, there’s a database of about 150,000 genomes that you can access, it’s called gnomad from the Boston Poly theorem the Broad Institute, there are other databases like the Craig Venter’s database is a database of about 10,000 Genome sequences, the Europeans were trying to put a database together of about 10,000 Genomes etc. But this number of genomes, let’s say 200,000 genomes that you can find today in the databases, are really not enough. The databases are very poor. So challenge number two, in genetic medicine is to enrich the databases for millions of genomes. Now one important thing here in databases is that if you have a database of variants, they’re worth nothing, if you don’t have a phenotypic assessment on them, a phenotype link to the variant. So the effort is to populate the database with variants, but also to populate the databases with phenotypes.
Then there’s the politics of databases which is important, most of the databases are on the other side of the Atlantic, and some of the databases are on this side of the Atlantic, and other databases are in Asia. And each one, because they get the money from a national pocket, and not from the international pocket, then the databases have a flag, have a color. And I argue in science that the databases should not be national, nationalistic. But the WHO which is in Geneva and I’ve shook hands with some of them, they should take this under their umbrella, and create genotypic-phenotypic databases that are international, accepted by others, by everybody, funded by the national programs, and that will solve the problems of accessibility and the problem of privacy. Now let me say one word about privacy, of course we legislate for privacy and we legislate against discrimination. However, some people argue that privacy does not exist. And if privacy does not exist in behavior, perhaps it would be difficult to have privacy in genomic variants. but the best argument for people contributing their genomes to the world’s database is that your client, your patient, comes for a genetic diagnosis, when we find a variant, that causes a disorder, how do we find the variant. We did it; we compare this variant to the databases. So we have a diagnosis because others contributed their genomes to the databases. So this altruistic solidarian argument is the best argument I think for people who contribute their variants because if they don’t contribute their variants, we cannot make a diagnosis on an individual basis.
George Chrousos: So Professor Antonarakis, your talk is very excellent in my opinion because you’re trying to be extremely vigorous with science in this particular case applied to genetics. But there is something where I would like to ask a few questions and also permit the audience to ask any questions that they want because we don’t talk about clients, we talk about patients. We are doctors and we work in university hospitals and genetics made great progress so in the 60’s we in the 70’s we preferred PCRs, and in the 80’s and 90’s we preferred sequencing. And now we have genome and we have and some geneticists say the is dead or chromosomal abnormalities.
But we still measure glucose and we need to incorporate genetics in our work, in our daily work. Briefly, how do you consider that geneticist and clinicians should get together to organize techniques to use for what purpose interpretation of technique. I understand all of the which take all of the databases, we know where it’s going and we know that we will find an insertion producing mutation and chromomycosis and creating a stock current is highly pathogenic if a specific phenotype that we are working with.
So how should we organize, collaborate together in a clinic with a patient, with a client, a human being, Geneticist and Pediatric Endocrinologist. How? and in a country like Switzerland how many laboratories of genetics do you think you need to take care of the Swiss population in terms of monogenic decisions?
Antonarakis: Well I will start with the last question. Which is quite easy actually. Before the introduction of the Genome analysis and diagnosis, the estimation was that we need one genetic center for about 1 million people. So Switzerland has 6 million people, so 6 genetic laboratories in different places to serve the population. Now that the Genomic analysis is a reality, and the search for variants is more laborious and the genetic counseling becomes more complicated. Then we still need 6 genetic centers, one per million, but with more people per genetic center. So the answer is, the question really is not how many centers we need but how many specialists we need, because we need six centers still, but we may need fifty people per center instead of three people per center.
Chrousos: What do you mean a specialist, you mean geneticist doing clinics?
Antonarakis: Now, when we talk about specialists, I’m talking about two different branches, one medical branch and one laboratory branch. And we consider both of them equally important. So the medical branch is MD’s and the laboratory branch is PhD’s and since you’ve mentioned the country that I work in, my adopted country also, after the US, is that we have two specialties one for MD’s and one for PhD’s and there’s always the politics etc. and I don’t want to go into this, but you need both. And now that cancer became a genetics problem, because as you know cancer is a genomic problem of somatic cells and has to do with driver variants and rearrangements in the genome, now we need more and more and more to deal with the cancer categorization, the treatment choice, the monitor of the treatment, and the change of the treatment with relapses. So the six centers that I’ve mentioned to you, you need to add one center per million people for cancer genetics. So we’re talking about twelve now with several people in each one of them.
Now let’s go to your first question. What do we do with genetic investigation? If you like to look at this holistically, the best thing to do is to sequence the genome. Sequence once, and interpret many times. And the answer to the question is also financial. If the sequence of the genome goes below $400 a genome, which may happen in two to three years, then that will be the detection of choice. And I argue and I see it that many countries will introduce neonatal genomes, so everybody that will be born will have a genome done, and databased and this will be one component of the medical record indisputable and not questionable, and then we base our medicine on additional tests beyond this. In practice though, because genome sequence is still expensive, insurances doesn’t pay in many countries for genome sequence, insurances may for exome sequencing, all the protein coding genes as we have said. Then this is the diagnostic method of choice, but if you have a very strong clinical suspicion of one gene involved, you can only do that one gene, or if you suspect clinically a chromosomal disorder, then you can do a chromosomal analysis, perhaps not with the regular karyotypes but with micro hybridization or with some kind of cheap sequencing method in which you see the copy number of chromosomes plus the junction fragments of translocations. The diagnostic thing is in flux actually and it largely depends not on technology that exists, but on the cost.
Of course this is a vicious cycle, a new technology will bring the cost down, and therefore we use that technology. The problem in medical practices is that my generation and your generation will soon disappear from medical practice, and the new generations they don’t know or they’re not used to this clinical thinking that our generation had. The new generation will do genotype first, blindly and without any thinking, and then they will start to think about it after this.
Argente: I can see the reasons actually, I can see that in Spain I can see that in the States and I can see that everywhere I go, they start thinking about medicine as genetics. So that means that if you have a specific disease or a specific phenotype you need to find out an explanation due to a mutation in one specific gene. As a monogenic disease. Of course the base can have a syndrome, the syndrome can also be explained the genetic techniques or not according to our knowledge. But today, most of the medical residents, I’m talking about my personal experience in the field, they’re not thinking in terms of physical exam, you know what happened with your family, what kind of diseases you have, but in terms of genetics immediately you know, everything should be genetics. And the geneticists say, medicine isn’t genetics, even if you have a fracture in your thigh, can’t be explained by genetics really. But the geneticist like you right now, are really more prudent and say please don’t do medicine like that at least so far, because right now we have the possibility of being wrong, we have the possibility of applying something in an appropriate manner, or maybe what we consider is appropriate, maybe tomorrow we will know that it is not. And of course we follow all of the reports from the states that we’ve acquired so far, the amount of variants, mutations, pathologic mutations that are not, and this is quite important in order to follow medicine. Don’t you think so?
Antonarakis: I couldn’t agree with you more; we do our best with whatever we have. So with the prior knowledge that we have and the knowledge that we get, as I’ve said before, one good hypothesis is that genetics is extremely important for the etiology of disorders, on the other hand if we use only genetics today, we explain 0.3% of what is out there. So because of the lack of knowledge, we cannot apply genetics only today. There was a study two years ago by an international collaboration in which they asked the question; from all the patients that will go to a hospital today, how many of them will have a genetic disorder? a problem in the genome. And this was done from many monozygotic twins in which people share the same genome in the zygote actually, when the monozygotic twins grow older, they accumulate their own variants and they diverge with age, and it was estimated that 49% of the suffering in a hospital today, with todays knowledge, is due to the genetic variation. About 50%. And for that 50%, we can identify a very small fraction of the real variants that causes them. Now, you said something very important too, that we can say something about the high impact variants. Just remember this term. High impact variant that cause the Mendelian disorders. Or the digenic disorders or the trigenic disorders, or the oligogenic disorders. However, the medium impact variants and the low impact variants, we do not know much. And this is a mega challenge in medicine, not only in my field to identify the low impact variants that contribute to disorders. I’ll give you a low impact variant that you all know, there’s a gene for apolipoprotein E, that makes a protein, the apolipoprotein Apo E, and the Apo E comes in three flavors, three colors. A blue, a red, and something in between, a red/blue, according to two amino acid substitutions in positions 112 and 154. The red/red is cysteine in those two amino acids, and the blue/blue is another amino acid. The blue we call Apo E4. And a specific fraction of the population is homozygous for this E4. They have two copies of E4. Now if you happen to have a homozygous for E4, you have a 15 times more of a probability to develop Alzheimer’s disease than the normal population, if you are Caucasian, European. So that is what we call a low impact variant. Because it increases your possibility of a disorder, but is not deterministic like the achondroplasia variant for example. Everyone that has the achondroplasia gene, there’s a mutation in one amino acid, everybody has short stature and the bone abnormality. So the determinism goes a hundred percent to fifteen times more than the population. So the high impact variants we know, we conquer all of them, the low impact variants is a tremendous challenge, not only this but the low impact variants, they don’t act in in isolation actually even the Mendelian variants don’t act in isolation. There is no really Mendelian disorder. In all our books now we call them neo Mendelian because there are modifying factors of high impact too that change the phenotype. So for the common complex phenotypes there are many low impact variants that contribute to disorders. And perhaps there is one medium impact variant that triggers the appearance of a phenotype.
Argente: Good, so let me ask you a last question before passing the microphone to all of the members here in this living room, you’re Greek, come from a Greek origin, I come from Spain and we can understand each other because we are speaking in English. My question is; do you think that most of the clinicians today speak the same language that geneticists speak?
Antonarakis: Of course not.
Argente: So we need to do a good deal of teaching people how genetics works etc. because we are spending a lot of money. In Europe for example, the number of private laboratories, doing any kind of techniques with no bioinformatics vigorous control is amazing. And we are spending tons and tons of euros doing something that most of the people really don’t know what they’re doing. And this is not moral. This is unethical in my opinion.
Antonarakis: I agree with you partially. However, you need experts. Socrates said: don’t look at what many people think, only look at what people with knowledge think. And the geneticist is something else that happened in our generation and I’m very proud of it, it is that the geneticist now has an organ of investigations. Something. When I started, the geneticist had the phenotype. He was a helper. A helper of others. Phenotypic assessment. Short, tall, dimorphic, big ears, etc. Now the geneticist, or the genetic medicine, or medicine, has this organ that’s the genome. As the cardiologist has the heart, and the neurologist has the nervous system, and the psychiatrist has I don’t know what, the geneticist has a different organ that is big but finite, don’t forget the finite. And the computer makes the finite shorter and shorter and shorter as time goes on and artificial intelligence and machine learning makes this shorter and shorter and more approachable. So I never expect that everybody will be on the same wavelength. And why be? we’d become monomorphic again. And I think the diversity of thought is healthy, and if there is diversity of thought, there is a need for MDs to take care of the others.
Ana Latronico: Thank you very much for this very interesting and unique conversation, and I have a question about the relevance of imprinted genes, what is your view about them? because in pediatric endocrinology, we have a lot of imprinted disorders, some of them are implicated in puberty and growth. Another point is related to animal functional studies, how they have impacted the new era of human genetics and if they are so important, as we previously thought. In the genetics of human pubertal development, mice studies did not help. More recently we have another imprinted gene, associated with precocious puberty called DLK1 on chromosome 14. Interestingly, chromosomes 14 and 15 are very similar in some aspects, especially because both contain a cluster of imprinted genes. Then I totally supported your idea about animals and we should be more careful about when we require functional studies using animals.
Antonarakis: Thank you very much to discuss the idea of imprinting. As you probably know, imprinting means that a gene is expressed only if it comes from one of the two parents, paternally expressed if it comes only from the parent when it comes from the father and when it comes from the mother it’s not expressed. It’s a peculiar type of inheritance, there are several imprinted genes that we know, and with a large project like the GTEx, the genome tissue expression project in which transcription of thousands and thousands of cells from different tissues are studied, then we will know by the alleles that are expressed, how many total genes we have with imprinting, we don’t have that many. And most people think that we will not find many more of protein coding genes that are imprinted. They are very interesting, it’s about a hundred, if you look at a mouse 120, people expect to find no more than 200, this is an exception, but perhaps for development it is very important. The other point I would like to make is that we have genes and networks that change in different stages of our lives, in puberty for example, the regulation of genes changes, because some genes are lined up, transcription factors, and then the regulation changes. The best example perhaps in puberty you have better examples.
But the best example that we use in my community is hemophilia B. There’s a hemophilia in which there is a mutation in a promoter, one type of hemophilia B not all of them, and when the child is a child, they’re hemophiliacs. They need the transfusions, they need factor 9, they need care, they bleed etc. and when they reach puberty hemophilia B becomes very mild. Because there is a different transcription factor that lights up and then binds to this promoter area better than the one in childhood, and then it cures the hemophilia B. So there are several examples like this that the gene regulation changes throughout life. And perhaps in my stage of life, the third stage, the regulation changes yet so one needs to study us. That brings me to the mouse experimentation. We now say that each human could be a project. A research project. Human, one individual, how we change through the years, through the environmental stresses, through the networks that change within us which are also the environment, the microbiota in our gut, everything that touches us, then we need to study in each human because each human has a unique genome and a unique environment.
Stefano Cianfarani: A very short and general question about the impact of epigenetics on genome function and the impact of epigenetics in medicine.
Antonarakis; Undoubtedly epigenetics is important to give you the short answer, you’re talking to someone who is biased or is the variation of the genome. There are many experiments that concluded, and the evidence is mounting, that a good fraction of epigenetic variation is based on the genetic variation. So in other words, your epigenetic marks are probably different from mine, and you can say that the environment makes them. However, a good fraction of them is because of your genetic variation that predisposes or allows the epigenetic mark to be on you and not on on me. But this is a biased opinion.
Christos Yapijakis: A very short but philosophical question. Do you think there is evidence in the human genome supporting the theory of intelligent design in evolution?
Anotonarakis: I will give a short answer. Probably not. There are two forces that work in the evolutionary process; chance and necessity. Mutations and selection. There’s nothing else that matters. Variants occur in nature, wakes up in the morning and has mutations, notes them down in a book in some genome and then necessity comes in selection and says it’s not useful for me. If it’s not useful, you have it, you die with it. If it’s useful, I like to have it myself too. So if there’s intelligence on this, you answer.
Argente: Thank you very much Prof. Anotonarakis, it has been a great pleasure to be here with you, I never thought that one day I could be at your house with a geneticist like that with an excellent great phenotype.