Lecture: Basic Genetics for the Ophthalmologist

Every ophthalmologist today is faced with genetic information, whether it be in our journals or the nature of the eye diseases we see. During this webinar we will review basic genetic information that can assist ophthalmologists in “going back to school” to refresh their knowledge of genes, mutations, and chromosomes to apply to a better understanding of their patients and test results. We will also review basic inheritance patterns to develop basic concepts for counselling.

Lecturer: Dr. Alex V. Levin, MD, MHSc, FRCSC, Wills Eye Hospital, Philadelphia, USA


DR LEVIN: Hi! I’m Alex Levin, chief of pediatric ophthalmology and ocular genetics here at Wills Eye Hospital in Philadelphia. I’m delighted that you’ve joined us. We’re gonna talk today about basic genetics. Some of you have sent questions in advance. We’re hoping today will be the first in a series of talks that will cover increasingly more clinical information on ocular genetics, but really, we want to start at a place where we can talk about the basic foundation that’s gonna let people be able to use genetics in a way that’s most comfortable for them. Before we get started, and this is part one of two of the basic series, what I’d like to do is have you answer some questions, some questions that will allow us to assess your current knowledge and see how we do at the end. So without further ado, let’s do the question part. You’ll see the questions as they come up. You’ll get 30 seconds for each. We’ll talk afterwards, and then I’ll be completely open to hearing your questions. So the question is: Which of the following is a disease causing mutation? A nonsense mutation, a missense mutation, a frameshift change, or I don’t know? So let see what you say. You should be pushing on your cortisone *screen which one you think is correct. Give it 30 seconds. So 48%, I don’t know. Let’s go on to the next one. A polymorphism… Causes disease, is a normal variant, occurs only in affected family members, and always involves one codon. While you’re doing this, it sounds kind of basic, doesn’t it? But what I’m going to show you is how important all this is clinically. So 51% say it’s a normal variant. We’ve got 8% causing disease, 19% only in affected family members. And 20%, only one codon. If the patient has more than one organ system involved, that means more than one gene must be involved, miRNA is the cause, it could be due to a single gene, or you should order whole exome sequencing? Here we go. So we say it could be due to a single gene, at 60%, and one more question. A negative genetic test means the wrong gene was tested, the wrong disease was suspected, the wrong test selected, or any of the above. And your answers are: Any of the above, for 68%, very good, and we see some scattered responses for wrong gene, wrong disease, and wrong test. So why do we have to know this stuff? Why do we have to spend time talking about basic genetics? And the reason is that genetics has proliferated throughout medicine, and in particular ophthalmology, for reasons I covered in my last talk. Every journal you read in ophthalmology now has basic genetics in it. Every patient you see, genetics somehow play a role in their disease. And when we talk about playing a role in disease, we really need to be thinking about: What do these tests mean? How do we order a test? How do we know what test to order? But most importantly, how do we interpret the results? Because now any of you out there in the world can easily order a genetic test. But when those results come back, you need to know how to interpret those results. Because if you don’t interpret them correctly, it could lead to parents and families making decisions about whether to have children, when they get pregnant, what to do with the pregnancy, people at risk may not know they’re at risk, or think they’re at risk when they’re not. There are so many spillouts related to a basic idea of genetics that it becomes increasingly important to know this information. So it will be very basic. I’ll take you through stuff you may have learned in medical school. We’ll go through the updated version, just so we know as a baseline today what the basic genetics concepts are, and I’m gonna show you how they lead to understanding what we do clinically. So genes are made up of nucleotides. Adenine, guanine, thymidine, and cytosine. And in RNA we have uracil instead. And this AGTC forms the code. And we read the code, these letters, as you can see, in triplets. And you can see these triplets — there’s a yellow one and a white one. I’m assuming you can see my pointer. And as we go through this, we call this the codons, or the triplets, and each of these is translating — is a code for an amino acid. Now, you don’t need to know this. No one knows this wheel by heart. But it’s very, I think, instructive to look at it. It tells us if we go in a code from a CGC, we get an R. If we look at our amino acid grid, R is for arginine. You notice there are many ways to get to arginine. That’s the brilliant diversity that God put into our DNA, that allows us to have some variation with a limited number of letters in the sequence. Likewise over here, if we go ACC, we get a T. Come down here, that T is a threonine. A different amino acid. But look what happens here. If we go TAA… You get that asterisk. What that asterisk means is it’s a code to stop reading the DNA. It’s a stop codon. It ends the reading that we’re making a protein. Now, genes have a structure. We can recognize a gene. Just like you can recognize a dog and say that’s a dog, and that’s a cat. You can recognize a gene, and their structures are very similar. They have areas called exons. These exons are the part of the gene that will become a protein. Genes can have one exon, or 50, 60, 70 exons that vary in sizes. And in between these exons we have stretches of DNA that are called introns. In the old days we used to refer to introns as junk DNA. But now we know they are important. If you have a mutation in an intron, that can be disease causing as well. The beginning of the intron is called the donor site, the end is called the acceptor site, and that’s because this material is gonna get spliced out to connect exon one to exon two in RNA and subsequent protein. Every gene also has usually an upstream promoter area. This promoter is going to create either RNA or protein and is going to affect the expression of the rest of the gene. Let’s say turning on the gene, turning up the gene. In a very important region. So you could have no change in the gene, but have a change in your promoter, and you get the disease, because you can’t turn the gene on. Now, what do genes do? Three things. Some make building blocks, things like fibrillin and collagen that make the structure of our body. Some make enzymes. And then we have the most important genes, the regulators or transcription factors, they’re called. And these are the ones that turn on and off other genes. If you think about an orchestra, an orchestra has many different instruments in it. You’ve got the strings, the woodwinds, the trumpets, and all of that. Every orchestra needs a conductor, who sit at the front of the orchestra and tells who to play when. You play a little quieter, you a little louder. Faster, slower. And that’s what the transcription factors do. If we consider all the instruments in the orchestra as building blocks or enzymes, the transcription factor is the orchestra conductor, saying how others should play, so they’re expressed at certain times of development, not others. Some upregulate, some downregulate the other instruments. But you can imagine if you had an orchestra conductor who had narcolepsy and was falling asleep all the time, or attention deficit disorder with hyperactivity and was out of control, either way, you’re gonna have bad music coming out of that orchestra. The transcription factors that are too heavy handed or too light handed for a mutation can cause disease. Now, DNA is transcribed into messenger RNA. Messenger RNA is translated into proteins. You can see we’ve connected the exons into this RNA. Now, obviously everything I’m gonna tell you today is much more complicated than what I’m gonna tell you. But suffice it to say that these very complex processes that allow this process to occur also involve many genes and the processes involved in those genes of transcription and translation are gonna affect many genes as well, all of which are trying to be transcribed and translated. Now, one gene can only make so many different proteins. In fact, it can make many, many different proteins, and it does that through a variety of ways. Sometimes the protein that’s secreted is then modified, posttranslational protein modification. We see that in insulin, where the first part of insulin is chopped off. But we could also make isoenzymes, so we could splice together the genes in different ways, to maybe use exon one and three and splice out exon two in some parts of the body. This is another brilliant way to create some diversity in the genome. So say in the eye exon two is not used for a particular protein. You might see in collagen — a mutation in exon two won’t affect that organ. Won’t affect the eye in that case. Now, mRNA we shouldn’t discount. RNA is very important. Some RNA is never made into protein. Things like microRNA, used in our body to regulate the complex processes we’re talking about. And we can have mutations in genes that affect RNA and actually never affect a protein directly, but would cause changes in proteins, because the RNA are regulating the making of protein later on. The human genome is quite complex. There’s over three billion nucleotides. But in there is only about 20,000 genes that code for proteins we know, and many of those we just don’t know what the function is. Less than 2% of the genome actually codes for proteins, and that means when you order a test like whole exome sequencing, which is a test for all of those genes that code protein, you’re really getting less than 2% to 3% of your genome. The rest of the genes are doing other things you’re not testing for. So a negative whole exome sequencing test simply means you may not have tested a gene that’s out there causing that disease. We’re over 99% identical. If you look around the room we’re sitting in, everyone looks different. We’re all different. So you ask… How can that be? That we’re 99% the same? If you take 0.01% of 3 billion, 1% being 30 million… 0.01% is still a four million nucleotide difference. 3 to 4 million nucleotide difference between us. So there’s lots of variation in the human genome. There are no unique genes. There’s no Alex Levin gene. You all have the same genes that I do. The difference is the sequence of the genes. And there are variations in the sequences that make us different from each other, without causing disease. In fact, our genome is 90% the same as mice, who diverged from us 50 million years ago in evolution. We’re 75% the same as the fly. In fact, the control gene for the eye is over 70% the same as the fly, even though the fly eye is so different from us! So these are really important concepts, when we consider how genes have stayed the same. The variation is variation in us which causes differences that are not disease causing. Those are polymorphisms. So we all have polymorphisms in our genome. And what we define as disease may have some gray zone in the middle. Red hair is an example. Red hair we don’t consider a disease. It’s very rare in the population. We know the gene that when changed can cause red hair, but we accept that. Certainly we don’t accept congenital heart disease or retinitis pigmentosa, or other genetic disorders, because in those conditions, the gene change is causing a disease. That’s when we call the gene change a mutation. So whether or not a gene change is a polymorphism or a mutation is simply defined by the fact of whether it is disease-causing. And when you order a genetic test and you get a result, the result tells you that there’s a sequence variation, but it’s not telling you if that sequence variation is disease causing or just a simple human variation, a polymorphism. This is very important. Laboratories, when they give you the report, will sometimes try to distinguish this. They may say it’s a polymorphism. They may say it’s a mutation. They may call it a variant of unknown significance, meaning they can’t tell the difference. But even those reports are subject to error, and must be checked thoroughly, to get the evidence to let us know: Is this a polymorphism? Therefore benign? Or a disease causing mutation? Very critical point. We all have, as I said, polymorphisms. Probably over ten million of them in each of us, occupying 0.3% of our genome. We also have deletions, duplications, major areas of our genome that are upset in various ways that don’t cause any disease. Therefore when we do a test that looks for a deletion, how do you know the deletion is actually causing the disease? It might actually be normal. So how do we tell? We look at mutations, we ask some questions, and we get reports back. We do this in every patient. Does the change in the gene predict a biologic error? For example, we know that when you mess with cysteine — cysteine is a crosslinking amino acid — that’s usually bad. That’s usually gonna change the structure of a gene, and therefore the linking of a gene. Therefore that’s usually bad. We also ask the question: Is it present in ethnic controls? So if this change looks like an important change, we have to know: How does this compare to other people from the same country or the same ethnic background? And in fact, sometimes we find that it’s a normal variant for that ethnic background. And doesn’t cause disease at all. That’s why people from China look like people from China and people from India look like people from India, and people in certain races look like — it runs in that group and it’s normal. The next question we ask is: Does it segregate with the disease? Does the change segregate with the disease? So that means we have to test family members. So if the change is truly a mutation, that mutation should only be present in the affected individuals, and not present in the unaffected individuals. So therefore a test on the one person may not be sufficient. You need to test other family members. We also look at evolutionary conservation. In other words, does this change occur in a part of the gene that’s stayed the same since flies or mice? And if that’s the case, we know this is a very important part of the gene, and gives us more evidence that indeed this is an important change that could be disease causing. And sometimes we can demonstrate this. We’ll use models in the laboratory or animals, and we can even use protein prediction tools — in silico tools — to help predict. They’re less reliable, but we sometimes use them to predict what the biologic impact will be. So this is the drill we need to do in order to interpret the result. We have databases out there and literature we can use. We always do that when we see every patient. And I hope you get the feeling that for every hour I spend with a patient, we’re spending hours in the background, trying to sort all this out in terms of their test results. Has this change been previously reported as disease causing? There are similar variations? In other words, maybe this patient has a change in this codon that’s never been reported before, but other patients with changes in this codon were either normal, meaning polymorphism, or diseased, meaning it’s a mutation. So there are many things we can do to instruct ourselves from the literature and databases to interpret tests as well. So I guess the bottom line is: Don’t order a test unless you know what to do with the result, and that’s very, very important. Even if the test is free, and there is free testing available more and more today in various places. You need to know how to interpret the result so you can best care for your patient. What does a negative test mean? Well, you send this test out, you get back a result that says that the test is normal. Well, you might have the right disease. But you may have tested the wrong — or maybe the gene is not known. So maybe is retinitis pigmentosa, but we don’t know all the genes that cause retinitis pigmentosa. It still is retinitis pigmentosa, but you tested either the wrong gene, maybe your panel didn’t have enough genes on it, or maybe the gene has yet to be discovered for that disease. You might have the wrong disease. I thought this was Stickler syndrome, so I tested genes for Stickler syndrome, and the test came back negative, because it isn’t Stickler syndrome. It’s another disease that looks like it. Maybe it’s Wagner syndrome, from another gene that doesn’t cause Stickler syndrome. Maybe the analysis was not complete. This is very, very important. If you don’t test the whole gene, and you only test known mutations in that gene, you could be missing novel, new mutations in the gene. Were the introns tested, or just the exons? Was the promoter tested? These are all questions that we have to ask, to know whether the analysis was complete. Again, did you test the promoter? And then there’s other things that can fool us. Non-paternity. It’s been estimated that in North America, when blood is sent to the lab, non-paternity may be as high as 10% of the time, where daddy is not daddy. And that is usually known to the mother, but sometimes not. And these are cases where our genetic inheritance patterns in families are also tricky and could be unusual. There could be unknown incest, unknown rape, unknown relationships in a family, which make our interpretation of test results difficult. And lastly, there could be error. Sometimes it’s just error. The lab could make an error. There could be a swap in the specimens. So we could have the wrong test, we could have the wrong disease, we could have the wrong gene. All of these things are possible reasons for a negative. Let’s look at some changes in genes and see what they mean to us. We’ll take this strip of DNA up here, divide it into its codons, and what happens if we change the TAC to a TCC? We’re gonna change the A to a C. Well, this is called a missense change. Why is that? Let’s go to our wheel. Just to review, we’ll take a TAC and go to a TTC. So here is our TAC. It gives us a Y, which is tyrosine. We’re gonna change that to a TCC, that gives us S, which is serine. All we’ve done is we’ve swapped one amino acid for another. Is that a polymorphism? Or is that a disease? You don’t know. That could be entirely benign. Or it could be the cause of the patient’s disease. And the only way you can sort that out is by doing all the things I told you. Does this create a biologic error of importance? Is it gonna change a binding site or the shape of the protein? Have other people with this change been disease caused? Is this only seen in patients in the family who have the disease, versus those who do not? Those are the kinds of questions we’re gonna have to ask, to sort that out. Let’s do another one. Let’s take a TAC and turn it to a TAG. We call that a nonsense change, because what it did was it took our TAC, making tyrosine, and converted it to a TAG stop codon. So what’s gonna happen is we’re gonna truncate this protein. We’re gonna stop translating. We’re gonna have a piece of the protein missing. Is that a polymorphism, or does that cause disease? Certainly looks bad, doesn’t it? But it could be perfectly benign. And probably some of you out there have some of these polymorphisms in your body. Maybe it’s the end two proteins that aren’t important. Maybe our body can adjust without having all of this protein made. Sometimes this is just benign. How are we gonna figure out if it’s a polymorphism or a mutation? We go through all the steps I just taught you. This is a frameshift. We’re gonna take this G out, and by taking this G out, instead of reading GGC, the body is gonna read three at a time. So it’s gonna frameshift all the way down, and that’s gonna be a garbled protein. It’s gonna have amino acids on there that won’t make sense, and eventually it will stop. Is that a polymorphism or a mutation? You don’t know. It could be perfectly benign. We have to ask all the other questions. Genes live on chromosomes. Chromosomes live in our cells. If you were to open up a cell, you would get all of your chromosomes that spill out. You can notice that we can stain these chromosomes to give these banded patterns, and that would help us line them up in pairs, one from our mother, one from our father. Remember that despite all the ways families are made today, you always have to have a male giving one and a female giving the other. We have all of our chromosomes, 1 through 22, they get smaller as they get higher in number, and we have our sex chromosomes, the X and Y. So we have 22 autosomes — that’s where the phrases, autosomal dominant, autosomal recessive, that I’ll talk about in my next talk — come from. And we have XY being male, XX being female. The sex chromosomes. Bands are not genes. Bands are regions. They contain genes. You’ve got 22 chromosomes. Genes are quite small. Figure it out. There’s a lot of genes per chromosome. The short arm of the chromosome is called the P arm, for petite. The long arm is Q, the next letter after P. There’s a constriction in the middle, the centromere, and we call the ends of the chromosomes telomeres. Often there’s a concentration of genes just under there. And you’ve probably heard a lot in today’s world about telomeres and telomere shortening as being one of the causes of aging. I’m hoping I have long telomeres right now. Now, when a chromosome is aberrated, and that can be a duplication — here’s the nomenclature for duplication. dup5q22. What does that mean? If we go back, we can name our gene by their banding. This would be section 1, subsection 2. Here we have section 2, subsection 4, subsubsection 1. So we can name the regions on chromosomes. And just part of that could be missing or duplicated. So here’s a duplication of just the long arm of chromosome 5, region 2, subregion 2. Duplication 5q22. And there’s all sorts of ways that the chromosome could be altered. If you could alter a chromosome enough so that you could see that change under microscope, by definition there has to be more than one gene involved. And therefore you would expect that there would be more than one organ system involved, and very, very frequently, the brain is one of those. The brain being the most malformed or most affected organ in the body. Guess what’s number two? The eye. And therefore it would not be uncommon for a patient to be developmentally delayed, maybe have an eye problem, and so on and so forth. And in general, we consider that if a patient has three different unrelated malformations, then we order a karyotype, karyogram. We look at the chromosomes, looking to see if more than one gene is involved, thus explaining why more than one organ system is involved. There’s different definitions of these. Should they be all major? One major, two minor? And so on and so forth. But when you’re looking at the eye plus other parts of the body, or even different malformations within the eye, think about whether the chromosomes have been altered by deletion or duplication or other shape changes. How do we look for those pieces? There’s different tests we can use. There’s a test called fluorescent in situ hybridization, or FISH, where we use dye to see what’s missing in there. In this patient, one copy of the chromosome has two reds, but the other does not. It’s deleted from the other copy. The problem with this test is you have to know what you’re looking for. If you’re thinking about, for example, Wager syndrome, which we’ll talk about, you have to have the probe specifically for that, to do this. Nowadays, more than anything, we do what’s called chromosomal microarray. An amazing test. Where on a slide, essentially, we can test for many, many, many different — what we call copy-number variations. Deletions or duplications. In the genome. This test is fascinating. You know, we can do a one million mutation screen, in hours, for less than $500. And less than ten years ago, it would have been more than $500,000 and taken months. Today is the anniversary in 1774, oxygen was discovered. Now, that’s only 250 years ago. We have come a long way, a long, long way, and really our technology today is fantastic. But there’s another twist on this. Not only do you have multiple organ systems, by multiple different genes. One gene can also cause multiple organ system involvement. How would that happen? Let’s take fibrillin. Fibrillin is a protein used in multiple part of our body. We use it in the wall of the aorta, we use it in the zonules of our eye, and other places as well, as a basic building block. So if you get a mutation in your fibrillin gene, you may have Marfan’s syndrome. And you can have aortic root dilation and the other skeletal manifestations of your disorder. Bardet-Biedl syndrome. Many genes can cause it, but each of those many genes, in and of itself, is used in many parts of the body, because it’s a basic cilia protein. And cilia are important in migration and adipose tissue, thus the obesity in that syndrome, in our eye we have cilia left over, so to speak, in photoreceptors, and cilia are important for cell direction, the making of hands and feet, so therefore we might get polydactyly. All the features of Bardet-Biedl syndrome. Ciliopathies in general are multiorgan diseases, each of which can be caused by an individual gene. So when we see multisystem disease, we’re asking: Is this one gene being used in multiple organs or is it more than one gene that’s been deleted or duplicated by a chromosomal aberration? Depending on whether we recognize the disease, we may choose, if we recognize — oh, I know this is Marfan. Let’s just test fibrillin. Or you may say… This is some syndrome with lots of things in it. I don’t know where to start. Let’s look to see if there’s a chromosome first, and we would do a karyotype and a microarray. The most famous contiguous gene deletion system, the first to be described, was WAGR syndrome. Wilms tumor, aniridia, genital abnormality, and retardation. Think how important it would be to detect this. Maybe that genital abnormality is just a mild hypospadias you can’t see at birth. Maybe at one month old, it’s hard to tell if there’s gonna be developmental delay. And certainly they won’t have their Wilms tumor yet. So when we see patients with aniridia, we start with an array. That’s why we choose that test. We could choose just the PAX6 gene to see if there’s a mutation in that, but ultimately we want to know if this is this disease, because the patient could be at risk of Wilms tumor or not. So you can take away that fear from the family by a simple test, or give them the hope of a better cure and treatment, by early detection, knowing that the patient is at risk. So what tests do I order? You know, that’s a hard question. And it really depends on what you’ve got sitting in front of you. Sometimes we order single gene sequencing, because we look at the patient, and we say… I know this is Stargardt disease. It has all the features. How did I get there? Well, I looked at the OCT, the fundus autofluorescence, the ERG, maybe intravenous fluorescein angiography. All of these things to get me to recognize that this is almost certainly a mutation in the ABCA4 gene. Having said that, Dr. Jay Ibanez, a graduate of our ocular genetics program, now in the Philippines, did a study, and showed that over 35% of the patients that come to see me with a diagnosis of Stargardt don’t have Stargardt disease! And based on our phenotype, we might not even choose to test for Stargardt disease. So single gene sequencing is really helpful if you suspect: This is what I have in front of me. Sometimes we can’t be that exacting from diagnostic testing alone. And might order a panel. Many forms of retinitis pigmentosa from many different genes look the same. I can’t tell you: This is. I might be able to tell you from the family it’s dominant or recessive or X-linked recessive. That may narrow down the number of genes I have to test, but ultimately sometimes we have to test many genes. We talked about microarray on the bottom, when we would order that. But ultimately we have available to us whole exome sequencing. And I see many of you going to whole exome sequencing. But keep in mind the more genes you test, the more noise comes back. Noise meaning basic human variation. And the more work you’re gonna end up doing trying to sort out: Is this disease causing? Or is this a polymorphism? A benign polymorphism? So we really try and reserve whole exome sequencing when our more goal-directed testing, single gene, panel gene, or in patients where we just don’t know where to start, we have a syndrome in front of us, and we have no idea, we’ve never seen anything like it, it’s never been reported, we searched the literature, we can’t find anything, then we might expand whole exome. And then remember that whole exome is only 2%, 3% of the genome. We have available a research tool today, whole genome, to sequence the entire genome. But as you imagine, the noise level is getting higher and higher, because we all have differences, and trying to sort out which are disease causing, which are benign polymorphisms, not to mention the fact that we start to get more information we don’t know. I’m testing for a disease, retinitis pigmentosa, and find out that the patient is at risk for Alzheimer’s. Do you want to know that? How do we disclose? What surprises are in store? There’s ethical issues, financial issues — who is gonna pay? There’s availability issues. But ultimately it comes down to phenotype. And as advanced as we are with molecular genetics, in the end the most important person is the clinician. A good clinician, and we are training ocular geneticists around the world, because that’s what we specialize in. We’re rare disease doctors. We’re not smarter than anybody, but we see these rare things over and over again. We know what diagnostic test to order to allow us to get specific, to narrow down and say: This is what you have. This is what I think. And I can do the one gene test, instead of the many gene test. Instead of the whole exome or whole genome. Phenotyping is really the critical first step to determining what your testing sequence is going to be. So I’m gonna end there. We’ve got four minutes left. Let’s see what we’ve learned. And then we’re gonna open up for questions. I also have a bank of questions that came in beforehand, that some of you kindly submitted, that we can fall back on. There is a way on your screens to enter questions. And I would love to hear your questions. What we talked about today is to give you a basic outline of an understanding of the human genome, enough for you to hopefully think about whether you’re prepared to do genetic testing, how to use that genetic testing, how to interpret that genetic testing, and the impact that correct versus incorrect genetic testing will have. So let’s ask you the question again. We’re gonna ask you… Now, which of the following is a disease causing mutation? Nonsense, missense, frameshift, or you don’t know? Remember, just a little hint about all we talked about and what it takes to distinguish mutation from a polymorphism. Which of the following is a disease causing mutation. The studio audience says… We’re getting better. I don’t know. That’s the correct answer. Because those changes could be completely benign. Same with frameshift. Looks bad, but could be completely benign. With any of those changes, missense, nonsense, frameshift, until you do all the work to determine: Is it biologically significant? Does it run in that ethnic group? Are family members affected who have the disease, but those who do not don’t have the change? Segregation? What happens in the literature? Until you answer all those questions, the answer is I don’t know, and a missense or frameshift could be benign. What about a polymorphism? Is polymorphism disease causing? Is that what we call a normal variant? Does it occur only in the affected family members? The ones who have the disease? Does it just involve one codon? What do you think? We talked about polymorphism today. Let’s see what you learned. There we go! Well done! 84% got it correct. Normal variant. Polymorphisms are not related to the affected family members. In fact, it doesn’t matter whether they’re in the affected or unaffected, because these individuals have a benign change. So if the mom and child look alike because they’re mom and child, they’re gonna have the same polymorphisms that determine nose shape and ear shape and eye color. They look alike. But the disease that the child has, that mom doesn’t have, they could have the same polymorphisms, but the mutation is only gonna be in the child. Not in the mom. And as we discussed, polymorphisms can involve one codon, more than one codon, they can be deletions, duplications of whole stretches of DNA. If the patient has more than one organ system involved, then more than one gene must be involved? Or it’s a problem of microRNA? Or it could be just a single gene? Or the first thing to do is order whole exome sequencing? So the patient has multisystem disease. More than one organ. You tell me. Is it more than one gene that must be involved? Could it be a single gene? Is RNA the problem? Or let’s just forget about all of that, and order whole exome sequencing? What do you think? Yay team! It certainly could be a single gene. Just like Bardet-Biedl could be a single gene, Marfan could be a single gene, and we don’t want to order whole exome sequencing to start. We might want to look at a microarray first, or a single gene that fits the phenotype, like Marfan. Although multiple genes could be involved, they don’t have to be involved. A negative test could mean wrong gene was tested, it could be the wrong disease was suspected, it could be the wrong test selected for that disease, or any of the above. What does a negative test mean? Which of those options are correct? Well done. Well done. Any of the above could be possible. So I’m gonna open it up to questions right now. I’m gonna close this for a second, so I can see the question box that you may have sent in. Hold on. There it is. Great. So the first question is, from an ocular pathologist, what advice for genetic testing at histopathology slides? There are two answers to this. If the question is can you do DNA testing from histopathology slides? The answer is yes. It’s a complicated process. But you can do testing. But where do those slides come from? If the patient is deceased, you can get tissue. But most of the histopathology in ophthalmology come from living patients. And you can go back to the patient. But what your question highlights so importantly is how the pathologist must work together with the clinician, hopefully an ocular geneticist, to say: Here’s what I found on this slide, in this tissue. Rethink the diagnosis. The perfect specimen is corneal transplant. Sometimes we make the diagnosis of corneal dystrophy based on what the ocular pathologist found. The information the pathologist gives us, fed back to the clinician, can lead us to approach the patient, to do the appropriate blood test for the appropriate corneal dystrophy gene. The next person asks: What is the difference between a polymorphism and a phenotypic gene manifestation? That’s a great question. So polymorphism just refers to a gene variation — a sequence variation — that doesn’t cause disease. So the phenotype of a person actually includes normal things and abnormal things. So if you look at me, I’m not a very attractive man. My nose is very big. I have great big ears. See those? But we don’t call those disease in our society. You may think I’m ugly. My wife loves me, I think. But when we look at those features, we say… You know what? My nose works like any other nose. My ears work like any other ear. Those are my polymorphisms that give me a bigger nose, bigger ears, and maybe it runs in my ethnic group. I think the nose does. That’s part of my phenotype. But when we talk about the phenotype of a disease, if I had retinitis pigmentosa, we would be talking about those features in my eye. Bone spicule pigmentation, waxy disc, thin blood vessels, whatever. And those are the disease-specific phenotypic findings. There are no phenotypic gene manifestations. It’s phenotypic clinical manifestations. We’re trying to find the gene cause of those phenotypic clinical manifestations, and that gene cause will be a mutation, rather than a polymorphism. Next question is: What is the concept of genetic statistics, epigenetics, and how does it apply to glaucoma? Glaucomas are a disease like any other. Epigenetics is a phenomenon that occurs with every genetic disorder. I think the most lesion in ophthalmology is Leber hereditary optic neuropathy. We know that if patients drink or smoke, they’re more likely to express that gene and get optic neuropathy. That’s epigenetic. That’s environmental or local factors that affect the expression of genes in our body. Statistics is also very important in genetics. It’s important in several ways. For example, let’s say I have two children. And they both have retinitis pigmentosa. What’s the inheritance pattern? Well, you don’t know. Could it be that I have retinitis pigmentosa? And both my kids… Does that prove that it’s dominant? It may not. You can flip the head of the coin heads twice in a row. But if I had a hundred children, two had the disease, and 98 did not, you know for sure it’s not dominant, because that coin that you flipped heads or tails — when you start flipping it heads 100 times in a row, you’re gonna ask me to see that coin. Is that a trick coin? You can’t do that. So odds ratio, what we call Bayesian calculations, observed incidences, also statistics in populations, how frequent is this gene variant, is this part of this ethnic group — all these become important. They also become important in glaucoma because we don’t have a lot of specific one gene glaucoma diseases, and there are a lot of polymorphisms that influence whether you’re gonna get glaucoma. And that is all influenced by epigenetics. Very complicated, beyond the discussion, but really the future. This is a question from India. The challenge we face is getting the genetic testing to be affordable. It’s for a future talk, but know that people like ourselves here, at Wills, and partners at the University of Iowa, are working together to get affordable genetic testing to people worldwide. We will make this happen. Philanthropy will be a key. There’s not a hole in the bottom of the world where dollar bills are falling out — or rupees, in your example — so if we make philanthropic money for RP available, we will make this happen. I think right now today there are some labs that can do things very cheaply, like the Carver lab in Iowa. There are places where you can get research testing done. The clinical applicability is subject to some ethical questions, but I think the day is coming where everyone will have affordable testing. And you can always email me offline, and I can help you sort out how to get tested. Next question is: As a general ophthalmologist, what would be the first five abnormalities or pathologies to send to genetic counseling? I’m gonna change that question on you. Every genetic eye disease you see needs genetic counseling. And as we’ll talk about in future talks, you should never, ever, ever order a genetic test without appropriate genetic counseling. Delivered by you, or even better, by a genetic counselor. I could not do my job without my wonderful genetic counselor, and that counseling is so important, so patients’ expectations are adjusted appropriately. They know we may not find it, even though we spend money. They know that non-paternity is an issue. Pre- and posttest counseling is essential. And I think for any genetic eye disease that you see, especially when the family has questions about reproductions, at-risk family members, especially in those disorders like WAGR syndrome, where you know there’s a life threatening associated manifestation, or Marfan syndrome, where there’s an aortic aneurysm, you want to get those patients to see a geneticist, or an ocular geneticist, who can assist the family, counsel, and test. The next question is: Could a genetic disease be caused by an intron mutation? Absolutely. Now, that requires special testing to find those. And sometimes the gene sequencing comes back, and that’s because the intron wasn’t tested. So the answer is yes, and when we get a quote unquote negative test, but the phenotype… Gosh, it really looks like that disease. I really thought it was gonna be that gene. Everything fits for that gene. We might go back and delve into that gene more deeply, and look at the introns. Next question: Is there a mainstream genetic test that you recommend? That’s a difficult question. All genetic tests are becoming mainstream. If you go online, and go to clinical testing registries online, you can find what tests are recommended at what labs around the world. There’s no lab that does every test for every disease. But you can find almost everything — a clinical lab that will do clinical testing for your patient’s disease. By the mere fact that it’s available clinically, it is quote-unquote “mainstream”, whether it’s used or not, or whether people know how to use it is a whole different story. So you can find online those genetic tests. Is there a specific book for general ophthalmologists? That’s an unfair question. I’m gonna have to very, very carefully make a suggestion. We’ve written a book. Which is called the Wills Eye Handbook of Ocular Genetics. Which is kind of a basic book. Obviously I have a conflict of interest, although I think my royalties are two pennies a book or something. It has question and answer, differential diagnosis. It’s a thin paperback which goes through the major ocular genetic diseases. You may find it useful. Traboulsi has a tome, Ocular Genetics, a thicker hardbound version, which is more extensive and thorough, a wonderful resource. I also recommend the Online Mendelian Inheritance in Man. If you go online, put in OMIM, that’s a great resource to find any gene and the diseases it causes, and what its manifestations are. So there are specific websites — there’s a plethora. But for the general ophthalmologist, I am hopeful that our textbook, Wills Eye Handbook, is helpful. Another question is: What percentage of whole exome sequencing yields a diagnosis? When I see cases of rod/cone dystrophy with a negative panel result, I don’t normally go to a WES given the cost and the fact that it’s possible the reason is an unknown mutation. Do you recommend pursuing it? Yes, it’s useful when used by the right people, for the right clinical situations. The percent that’s positive depends on why you’re using it, and what you’re testing for. It does have a very important role in our diagnosis. But often I see patients who had whole exome… And they come to me, and I say… Well, you know, I knew all along it was that one gene. Because I can recognize the syndrome, because that’s what I do for a living. You have to remember that when you get results for whole exome sequencing, those results are given to you on the basis of a series of strategies. One is there’s 50-some genes that Labs of North America must report, because they’re life threatening results. You have to report Marfan, because the patient could have an aortic aneurysm that’s treatable. Then we look at the patient’s phenotype, the information that you send to the lab. And they say… Okay. What are the genes that might have something to do with this phenotype? But that’s a laboratory, trying to pick genes. They don’t really know. They don’t really see the patient. So they may be inaccurate, and we go back, and we say… Did you look at this gene? We didn’t look at that gene. We didn’t think it was related. Lo and behold, there’s the mutation. So whole exome sequencing requires a lot of collaboration between the clinician and the laboratory. To do this in a skillful, useful way. Thank you for the compliment. The question is: Can we send samples across to me? Yes, samples can be sent around the world. Almost every laboratory that does genetic testing accepts worldwide specimens. And I’m happy to communicate with you offline about that. Some countries, it’s harder to get specimens out of that country, through various governmental rules, but that’s something we can explore offline. We’re starting to get a little bit afield. The next question — we have three left — can retinitis pigmentosa be treated? The answer is it’s coming. It’s gonna happen very soon. We can talk in single digit years. We’re working on this now. Many centers. Clinical trials are currently underway for some specific genes such as the RPGR open reading frame mutations in X-linked recessive retinitis pigmentosa, there have been trials for rhodopsin and other specific genes. The answer is we can treat some patients now. Some patients will be too advanced. They’ll need stem cells. That can come as well. The best thing we can do is give them hope. Treatment will come. And get them tested. Get them to the front of the line for knowing what their gene is. Because it takes months. Maybe the testing in four to six weeks, but you have to test family members to figure out is it a polymorphism? Before you know this is what you have. Then you’re at the front of the line for a treatment trial, when treatment becomes available. Role of genetic testing in management of retinoblastoma. That’s a whole different lecture for another day. And we’re gonna end with: You mention PEX in your lecture and it’s a frequent finding. I don’t know if I mentioned PEX in my lecture. Those are the peroxisomal genes. They can cause a variety of phenotypes. Heimler syndrome, Zellweger. We use testing in specific situations where it might be useful. But you ask another important question. Should we think about genetic testing for a person who has eventually extraocular associated abnormalities like aneurysm? That’s a perfect example of a life threatening condition. My father died of an aneurysm. He had five aneurysms. What do you think I did? I sent him to the geneticist and they did genetic testing for aneurysm genes that we know. And it came back negative. Why did I do that? Because aneurysm is life threatening, and we wanted to know which family members may have silent aneurysms and not know. But at the same time, his test was negative, because we don’t know all the genes that cause aneurysm. So there’s only so much we can do. But certainly where is genetic testing most powerful? Areas where there’s treatment, areas where there’s life threatening, preventable things, like aneurysm, where we can do something before it gets to kill you. Families that really have a need to know and are contemplating their reproductive options. And diseases where we can have screening that makes all the difference, like glaucoma. So I’m gonna end there. Thank you very much. We’re a minute over. It’s been a great pleasure. Keep using Cybersight. Cybersight is a great tool for all of you. In so many different areas of ophthalmology, from surgical teaching, live case teaching, to lectures on all areas of ophthalmology. And you can link with mentors on Cybersight, and ask one on one patient questions. And have longitudinal relationships with these mentors from around the world. We’re all here to help you. Good luck. Have fun. Enjoy genetics. Thank you very much.

Last Updated: July 5, 2023

1 thought on “Lecture: Basic Genetics for the Ophthalmologist”

Leave a Comment