Exercise Induced Collapse: A Case Study for Dog Genetic Disease with Dr. Kari Ekenstedt, DVM, PhD

Dr. Kari Ekenstedt, DVM, PhD, is an Assistant Professor of Anatomy and Genetics, and the head of the Canine Genetics Lab at Purdue University. Dr. Ekenstedt received her DVM and PhD from the University of Minnesota. She works on the cutting edge of the latest genetics methods to better understand canine health and disease. Part of her research focus is on the development of new genetic tests to help breeders make responsible decisions for their breeding programs.

Her research has addressed important canine health topics including peripheral neuropathy in Leonbergers, myelopathy in Pugs, and neuronal degeneration in Great Pyrenees. She has also published on the appearance of “hidden” genetic mutations in purebred dogs that lead to unexpected coat color variations. Dr. Ekenstedt’s work has been supported by the AKC Canine Health Foundation and the National Institute of Health.​ One memorable moment from her research career was when Dr. Ekenstedt obtained DNA cheek swab samples from 126 dogs in a single day at a field trial; this was not only a record, but also a reliable way to develop carpal tunnel!

If you liked Dr. Ekenstedt's talk, you can find all of the Health Symposium videos here!

Transcript

Dr. Keri Ekenstedht [0:00] I’m really excited to be here. What my plan is is to walk you through the discovery process. I am a faculty member at Purdue University in the Veterinary College. I am a veterinarian and also a geneticist. I run a dog genetic lab here at Purdue, where we investigate new diseases, new conditions, and try to figure out making those happen genetically. My plan today is to tell you a little bit about how that looks, from the researcher's perspective, and then we’re going to use Exercise Induced Collapse (EIC) as an example or model for how that works. Apparently my slides have gone missing, so I’m just going to freewheel here a little bit. Before I was at Purdue, I was up at the University of Minnesota. That’s where I did my PhD and my DVM, and I was faculty there for a little while as well. I was there during the whole discovery process for EIC and kind of had that insider’s view so I’ll be able to tell a lot of interesting things from that perspective. The main process or goal of my talk today is to help you, as a breeder, realize that genetic testing isn’t just a hurdle that you have to get over, but instead to try to get you excited about the scientific process and how it works and talk about the really important role that breeders have when it comes to trying to figure out the genetics behind any of these new conditions that come up. 

[1:57] We’re back! We found my slides. I think we’ll actually get started here now, so thanks everyone for your patience. Hopefully this will be a really fun session. Like I said, I’m at Purdue University. I’m at the Veterinary College here. I was at Minnesota before that, so I was really involved in the development of EIC as a test and so can use that as a model to walk you through how solving genetic diseases works when we work with breeders and dog owners.

[2:32] The first thing that you have to do is ask yourself: Is this thing that you’re seeing in front of you (the disease, the trait, the condition) inherited or not? Both Dr. Oberbauer and Dr. Meurs before me talked about that a little bit. It can be tricky. The example I like to use is cataracts because there are several forms of hereditary cataracts out there in dogs, but also they can happen from injuries, toxins, and even disease processes like diabetes. We have to be careful, and we can’t just pin everything as genetic. Dr. Meurs touched on this a little bit already, but the things we need to consider are: What age does the disorder become evident? Is it consistent? If you’re seeing it in multiple dogs, is it happening in roughly the same time frame? Also, are there relatives, litter mates, cousins, half-sibs that are affected? That’s always a sign. And is it something that’s known to occur in the breed? (Which is also definitely a sign, with the disclaimer that of course new things can pop up that haven’t ever been seen before.) 

[3:47] Today we’re going to use a single-gene disease (you may hear these called mendelian diseases) as an example of how this works. These are really more straight-forward, less confusing, so that’s why we’re going to focus on that. We’ll use EIC as an example, not only because I had an insider view on the whole thing but also because there’s multiple breeds affected with EIC, and that test has been around for a while. So some of you have probably heard about it. Maybe some of you have even used that test. Any mutation (and you’ll hear the geneticists say variation or variant or just a change) in even one gene can (has the potential to) result in disease. That’s what we’re going to talk about today as an example, because it’s less complicated. There have been hundreds of single-gene diseases described and solved in dogs and somewhere in the ballpark of 200 of those are available as tests. Ditto for people. Maybe even in the thousands for people of single-gene diseases and conditions that have been described. This is a really important point, everybody. Mutations or changes in the genome happen every single generation. Dr. Oberbauer mentioned that as well. Every one of us sitting here and listening and talking all have genetic changes in our genome that happened that could potentially be bad, could even be lethal. This is (I’m going to throw you back into hopefully your high school biology class, if you remember) miosis. It’s how we get variety. It’s how we get differences between people and between generations. Every time an egg or a sperm is produced, there’s changes in the genome. Usually those are neutral. Usually they have zero impact whatsoever. Sometimes they’re good and they confer an advantage, and sometimes they’re bad and they end up resulting in disease. I’ve bolded that this is not anybody’s fault! It’s a natural process. I point that out because I know I have talked to plenty of breeders who have experienced—we’ll just say bullying. That’s kind of the word on the streets these days. But ostracization and just public crucifixion for something popping up in their kennel lines. I’m here to tell you that that’s not your fault. These changes happen every generation. You shouldn’t feel like it was your fault that it happened, and you should be nice to other people when it happens to them, right? Because this is a natural process. There’s a few things we can do to minimize their impact (and we’ll talk about that later) but I like to point that out. I one time had a breeder tell me on the phone, she said, “I think breeders of ___ dog breed would rather stand on a stage and tell everybody in the room that they had genital herpes before they would like to admit that they had X disease in their kennel line.” I think that encapsulates how sometimes there can be a lot of bad feelings and that we need to work against that because these things can happen, and it’s not your fault. As you just heard from Dr. Meurs and Dr. Oberbauer, there are lots of genetic diseases and conditions that are not single-gene. They’re not straightforward. They’re not easy. The hip dysplasia we’re going to hear about… maybe we can get Smith back online. Also cranial cruciate rupture, idiopathic epilepsy (as you heard from Dr. Meurs) with the cardiac conditions. You’ll hear these called polygenic or complex diseases. When you go into a project, you may or may not know how to define what it is: if it is a single gene thing or if it’s not. Sometimes it doesn’t become immediately apparent. Maybe sometimes later it does, and we’ll talk about that. 

[7:34] EIC initially was observed (and mostly it was observed) in Labrador Retrievers, and especially it was seen happening in hunt test and field trial lines. But there were other related breeds, like Curly-Coated Retrievers and Chesapeake Bay Retrievers where this was happening as well. We were seeing it clinically. The very first research step is to investigate the condition, the disease itself. Describe how it looks clinically. Describe how to treat it. Describe how to diagnose it. All of those things. Initially, that was done by an interesting clinician, Dr. Sue Taylor, who was in Saskatuan, Canada. She has since retired, but she did a lot of that pionary work.

[8:14] What are the clinical signs of EIC? What does it look like? Typically, these were young adult dogs (averaged between 5 months and 3 years or up to 4 years). The average was 14 months. That actually coincides right with heavy training beginning in a lot of these field trials and hunt test dogs. The other thing that was very clear about these dogs is they were the very fit, very muscular, very excitable dogs that had a lot of drive—the kind of Labrador that just does not want to stop retrieving. A lot of that excitable drive. What it looked like was muscle weakness or a lack of coordination or incoordination. They kind of looked clumsy, and then it would progress to this life-threatening collapse after just 5-10 minutes of really strenuous exercise, typically. It’s important to note a couple of things: these dogs were totally there mentally. Their mentation was normal. They could hear you talking to them. They could respond to that. They were what we would call bright, active, and alert. Very normal. I say unless the episode progresses, and we’ll talk about what that looks like a little later. These dogs are hot. By hot I mean temperature. You stick a thermometer in their rectum when this collapse is happening, and they’d have temperatures of like 107 degrees Fahrenheit. Initially, people thought it was heat stroke. But it’s not heat stroke! Because dogs that have heat stroke—yes, they collapse and, yes, they’re hot. But those dogs usually have really dull mentation. They’re kind of out of it mentally. They take a long time—hours to days—to recover from it, if they recover from it at all. These dogs were recovering very quickly. What we found out is that that’s normal hot. When you have a Labrador or a Chesapeake Bay Retriever or any dog doing strenuous exercise, they get that hot. It’s their normal rectal temperature at those sorts of strenuous exercises. So, they were hot but this was not heat stroke and it was normal hot. Now, I’m going to keep my fingers crossed and hope that the first video will work. 

[10:29] This is the video of the little Black Labrador in the garage. You can see this little black dog. He’s got a toy or a ball in his mouth. You can see already that he’s dragging those back legs, his pelvic limbs. But he’s BAR: he’s very bright, he’s very active, he’s very alert. He just interacted with his owner to go chase the ball again, which (as you’ll learn) should not be happening. He looks clumsy as he’s zooming around, or not zooming so much on his back legs. You can find these videos all over Youtube. The good thing is here he’s stopped, and this is what you should do is make him stop as the collapse episode progresses. Unfortunately, maybe for the sake of the video, we have him do a little more here, so you can see some bunny hopping. It’s definitely not normal, and then he realizes, “I’m having trouble ambulating or walking,” so he just lays down because his legs are not working for him. I think we can stop the video there and go back to the slides. 

[11:36] The first sign of an EIC collapse is that “rocking” gait or just kind of a clumsy-looking gait. What happens is those rear or pelvic limbs become weaker and weaker and eventually they cannot support their weight anymore, and they become uncoordinated or clumsy. These affected dogs, because they have such high drive—Labradors and other breeds that have a lot of drive to them—they don’t stop running. They’re still mentally active and normal, they just want to keep going. Then they’ll start dragging their back legs, like you saw in that video. The problem is that rear leg collapse can progress all the way up to a four-limb weakness. If you don’t intervene, it can progress to a complete inability to move: paralysis of the diaphragm and eventually death. Like I said, most of the dogs that get these collapses are totally conscious and alert, and they may still try to run and retrieve. As I mentioned, we need to stop that.

[12:35] What to do: stop exercise with these guys at the very first hint of any clumsiness or wobbliness. If you get them to stop, you make them lay down. You don’t let them move around. Help cool them off. Maybe offer some cool water. Of course, that’s how they sweat, through their panting. In 5-20 minutes later, they’ll be fine with no leftover weakness, no stiffness, no lameness. The whole thing is non-painful. They’re not experiencing pain during any of this, which is probably why they keep going. I put this little Clip Art in here—the beer stein—because early on when we were investigating this disease, we didn’t think it was fatal, and then we heard a story (and the stories started trickling in). The story that broke my heart the most was the story of a college kid and his frat buddies, and they may or may not have imbibed some alcohol, and the college kid’s dog had EIC, and he thought it was funny. So he just kept throwing the ball, and the dog kept trying to go after it, and that dog died. The point is to stop all exercise immediately. 

[13:49] What else contributes to these EIC collapses? Early on, everybody thought it was the temperature outside. “Oh, it’s too hot for the dog to be out there doing this. The heat is just contributing.” Actually, no. The ambient temperature is not a critical factor in these collapses. It turns out that if a dog is affected, they’re less likely to collapse while they’re swimming. A lot of these hunt tests and field trials will send the dogs through a body of water. Probably because the water is a little cooler, right? That’s helping keep their body cooler. However, if a dog has a collapse episode while they’re in the water… We have heard stories of dogs drowning because their legs aren’t working, and they can’t swim. In order to, again, prove that the ambient temperature is really not an impact, we have had severely affected dogs that collapsed during retrieving in frigid temperatures while they’re breaking ice to retrieve waterfowl. Clearly it’s not that ambient temperature that’s contributing.

[14:44] Excitement does seem to contribute. As I said, these are dogs with really intense, excitable personalities. Some of them would collapse without much exercise at all. They’d be able to go do a hunt trial or a field test just fine. Then, if you got the bumpers out at home—collapse episode. Or they’d be fine in all of their training, but the grandkids would come over and then—collapse episode. Because they’re just so excited. We’ve seen both negative and positive types of pressure on these dogs create collapse. Some of them are fine unless they’re under electric collar pressure. That sort of thing. Or they’re fine until they’re actually working with live birds. There seems to be some excitement-level trigger factors.

[15:25] How do we diagnose it? Before, in order to diagnose it, we ruled everything else out. That was laborious and time-intensive. Now there’s a simple genetic test to diagnose it. The rest of this talk is to talk about how we got there: How did we get this test? How did it become available so people could use it?

[15:45] The first thing is going to be initial contact with a researcher, like myself or Dr. Mickelson, who was really in charge of the EIC project at Minnesota. It might be a breeder who reaches out to a researcher. I’ve had breeders reach out to me. It might be a veterinarian. In this case, a clinician like Dr. Taylor up in Saskatchewan. Sometimes a pathologist. I’ve had a pathologist reach out to me and say, “You know, I did a post-mortem exam on two dogs from the same litter, and they had the same disease and it’s weird, and I know it’s genetic in people, and we should look at this.” It could be any of those factors. Regardless, even when a pathologist tells me that, I usually try to circle back to breeders, because if we don’t have motivated breeders to help us, things almost never are successful. We need a large number (as many as we can get) of samples, accurate phenotypes. By phenotype, I mean we need to know if the dog has (affected) or does not have (unaffected) for the condition that we’re looking to potentially investigate. And can we get them relatively rapidly? If it takes 5 years just to get 6 samples, that’s not a project that’s going to get off the ground very easily. Sometimes there aren’t that many dogs that are affected because you’re catching it really early. I had a project like that, where we caught it—I think there were 8 total affected dogs in the entire world. We caught it. We solved it. We offered a genetic test before any of that even had to get out there further in the breed. Sometimes it’s just a matter of how many samples we can get. This is really important: if there’s one or two take-home things to learn from this talk, this is one of them. Confidentiality is guaranteed. We are never going to give your name or your information or the registered names of the dogs or the call names of the dogs to anybody, other than the person that owns that dog. If we couldn’t maintain confidentiality, why would a breeder want to work with us? We absolutely never reveal that information. It doesn’t go in papers. It doesn’t go in talks. Nothing. It’s all completely confidential. I guarantee you: if you’re working with a good researcher, they will not discuss a sample with anyone but that dog’s owner. To the point where, as an example, I’ve had some breeders who were frustrated with me because we were trying to collect samples for a project, and there were a couple of breeders who were kind of running it, and they had reached out to their various puppy owners, told the puppy owners to send us samples. The samples were trickling in. We usually send a quick email and say, “Hey, your sample has arrived,” directly to the owner of the dog. And then the breeder was coming back to me: “Hey, do you have a sample from this one, this one, this one from her litter?” And I said, “I can’t tell you that. You don’t own the dogs anymore, so you’re going to have to ask the dog owner. If they’re willing to talk to you, they will have gotten an email from me if I have their sample.” I literally wouldn’t even talk to the breeder about those puppies, because that’s how seriously we take confidentiality. So, funding is always an issue, right? This is a conversation that will often come up early with breeds or breeds clubs, breeders. We’ve had breed club funding. I had a breeder who contributed some money and then her employer matched it. We’ve also certainly been funded by AKC Canine Health Foundation. Other foundations out there that will fund work in dogs: Morris Animal Foundation. There are ways to get funding. But of course none of the research is free.

[19:32] I’m not going to do this whole slide, but I realize that to most of you (including my own family) what I do for a job is kind of a mystery. I wanted to just sort of explain. Most of the researchers who work on these discovery-type projects are at universities. Most of them run laboratories that are funded, usually via grants. These are not-for-profit endeavors, with very rare exceptions. They might ask you to pay for the blood draw on your dog and pay for the shipping to send the sample to us. That’s why. We’re a not-for-profit, so that’s your contribution to help keep labs running. That’s why. Certainly, if you get a cheek swab sample kit from Embark (they’re one of the funders for this talk) or any other commercial laboratory, they’re a for-profit. They do great things for dogs, but they are still a business with that model. So of course they’re going to give you the swab, and they’re going to pay for shipping. You just may not expect that to be different in a research situation at a university, but we’re not-for-profit, so sometimes the money is just a little different. And then I just put a list of all the things that faculty members (or primary investigators) who run the lab, like I do, have to do besides work on your project for your breed and your dogs that you love. It’s not that they don’t love your dogs, too. I love all dogs. But they’re very, very busy people. They may have to take an approach where they work on the projects that they think are most likely to be successful. Sometimes I have experienced dog breeders who reach out to me and say, “I’ve been in touch with so-and-so and so-and-so, and none of them wanted to work on this.” That can be a really frustrating experience as well, but this might be part of the reason that you may run into that. Just be prepared. 

[21:28] When we started working on EIC, we collected samples from affected dogs and extracted DNA from all the samples, obviously. We collected DNA samples from relatives of affected dogs and definitely wanted medical information on all those dogs, so we can have an accurate diagnosis. And then we collected samples from unaffected dogs. They’re called controls. In this case, we wanted dogs that were old and of the same breed (of course) and had no history of collapse. Why does it matter that the dog is old? Well, it would be great if you wanted to send me a sample from your 1-year-old Labrador who’s never collapsed. I’ll probably still take that sample and put it in the freezer. But if that dog is only 1 and has never collapsed, it could still collapse. Right now, its phenotype (whether it’s affected or unaffected) is unknown. But if you have a 14-year-old Labrador who you use to do field trial and it never collapsed, that’s a really good control. So we’re usually looking for the ones that are going to be kind of the best. And that phenotype—that characterization—is vital. The next thing we do is build pedigrees. We’re going to do that so we can try to figure out: What’s the mode of inheritance? How does this disease look like as it’s being inherited? I’ll confess: this is not a Labrador pedigree you’re looking at. I chose a different one from a different breed in a different disease. You’ll notice that there’s no identifying information on that pedigree whatsoever. I’m not going to walk you through all of it because probably a lot of you have already drawn pedigrees or seen them. But in this case, the filled-in symbols here with the dark completely filled in are affected dogs. What we’re looking for, because autosomal recessive is the most common type of genetic disease in dogs (outside the polygenic or complex ones), we’re looking for things like inbreeding loops. If you get a dog somewhere back here who was a really popular sire and he ended up contributing down through this stud dog here and also this way through these females, you could get something kind of from both sides; that’s called an inbreeding loop. That’s how maybe a brand-new mutation did happen, and in this case, it did. If you had a popular sire and he was a carrier (so he just had some copy of it), you could get it coming down through both sides and then showing up in these litters and producing affected puppies. That’s what’s happening here. EIC definitely looked autosomal recessive. Autosomal just means it was not a sex-linked condition, so it was seen equally in males and females. In EIC, I can tell you that we did trace it back to a very popular sire—again, who shall remain nameless. But the point is that we now know that that sire is not where the mutation originated. We’ve actually found that mutation in other breeds as well. He just happened to be the one who sent it everywhere into Labradors. The take-home message there is that popular sires can be a problem if they become too popular. I think some of the message there is to make sure you’re maximizing the diversity of your gene pool and not using that sire too much. 

[24:43] I already said autosomal recessive is the most common mode of inheritance for genetic conditions in dogs. Single-gene, okay? Single-gene conditions. That’s as opposed to things like dominant conditions. You may remember this from a high school biology class maybe. Or sex-linked; there are a couple of sex-linked conditions that we know of in dogs that are more rare. In order for a dog to be affected with an autosomal recessive, single-gene condition, they have to inherit 2 copies of the mutation: 1 from each parent. In this case (and in EIC as well), dogs that have 1 copy of the mutation or 2 normal (non-mutation) copies are going to be clinically normal. They do not have collapse. Just so you know, this fancy word “heterozygote” just means carrier, just means 1 copy. The problem with autosomal recessive diseases is they’re very hard to eradicate because you don’t know who the carriers are until you accidentally (or unintentionally, because you don’t know) breed those two carriers together and get affected dogs, which is a very slow sort of horrible way to try to figure out what’s happening genetically, which is why we love to try to solve these things.

[25:55] At this point, the role of the breeder is to round up samples, to help us round up samples. Contact your puppy owners. Encourage submission for those people. Help us explore funding options. Sometimes that can be through your breed club. Sometimes breed clubs will include a letter of support for us when we write grants. I just have to say this, too, because I’ve experienced this with breeders: ostrich behavior! Sticking your head in the sand and trying to pretend that it’s not a thing in your line and hoping it’ll just go away or even ostrich behavior by the whole breed. It’s probably worse in the long run because all that’s going to do is let these alleles get dispersed further. Nobody wants to produce unhealthy puppies, right? We all want to produce healthy puppies! And then I say “or worse, deceitful behavior” because I did have (and it was with the EIC project) a person who submitted a sample for what happened to be a stud dog. Here’s the thing: in genetics, and in the modern day, if you send us a dog and tell us that he’s a stud dog for all these puppies, and if he never matches with any of them, we know that you’ve sent us the wrong sample. In this case, it was definitely done on purpose. You may have gotten away with it because we still didn’t have a sample from that dog, but we knew it. That kind of thing isn’t going to help us, for sure, and it’s probably not going to protect your reputation the way this person thought it would. I, of course, think transparency is great. But everybody has to make their own choices. Either way, like I said, we would never be releasing specific information about any owners or any dogs. And then patience. The projects don’t always work out immediately. I had one project that I worked on and we solved it in 4 months. It was amazing. I’ve had other projects I’ve worked on for years—years—and we still haven’t solved them. Sometimes they don’t work out immediately, and sometimes they don’t work out ever. Sometimes it turns out they’re not single-gene diseases, and it turns out it’s actually polygenic, and that just makes it a lot harder. We need more samples, more affected dogs, more control dogs, that sort of thing.

[28:06] What genetic tools are the researchers using? They’re going to go to one of two or maybe both of these modern tools. First tool is a marker. By marker, I just mean it’s a pin in a map. Think about it like this: you, out there in the world, need to find me, one human being, out of the 330 million or so in the US. A marker is going to get you into the right state and even maybe into the right city. Now, there’s still 250,00 people in the Lafayette area (where Purdue is). It’s closer to 300,000 when the students are here. I mean, 300,000—you’d rather sift through that many than 330 million, right? Markers are just pins on a map that get you to the right place. The marker that most of us are using now are SNPs (single nucleotide polymorphisms). Lots of words. It just means a one-letter difference, so an A to a C or an A to a T or a C to a G, in that DNA sequence. It can be between the two chromosomes in one dog. It could be between a chromosome in that dog and a chromosome in another dog. We all have these. We all have millions of SNPs. The vast majority of them are harmless. They happen in the in-between DNA that’s not coding for specific proteins, like Dr. Oberbauer talked about. The commercial array now that’s available for research—the best one, in my opinion, is 710,000. For $100-$200 per dog (I can get it for $100 because I’m special), I can test 710,000 SNPs on that dog all at once, simultaneously, and get both copies. Because, of course, with SNPs you’re going to have one on each chromosome, one from mom and one from dad. That’s a lot of data. That’s 1.5 million pieces of data just for one dog, if you count both copies. There’s some complicated statistics and data management that have to be done. All that’s done (when you think about our metaphor) is told you that Keri is somewhere in the West Lafayette city. Those 300,000 people. You still haven’t found me. You’re just a lot closer. That’s what markers do. Whole genome sequencing is exactly what it sounds like. It literally just sequences the entire genome of a specific individual. So, from beginning to end of every chromosome. Now, you can pick different coverage rates. Some people don’t want to spend as much money, and they can only pay for 2-4x coverage, where every base in the genome is sequenced 2-4 times. Remember, the average mammalian genome has 2.5 billion pairs of letters. That’s about 5 billion letters total (because they line up in that double-helix structure) in each mammalian genome. I’m a little bit of a data snob. I like to have really deep coverage because some areas end up covered really well and other areas don’t have good coverage. So if you’re paying for 2x, you’re going to end up with areas that aren’t covered at all. I like to hedge my bets and get everything covered, so I like to go for 35-40x coverage, so pretty deep coverage of that genome. That costs about $1500, which sounds like a lot but actually the most expensive part is paying somebody to process that data. One raw data set, at this coverage rate for one animal, is 50 gigabytes of data. Gigabytes! All that has to be processed, so they do that on supercomputers here on campus. It takes a lot of time, and it takes money because you’ve got to pay someone to do it. We take that information, and it would be aligned with the existing genome maps. There are a couple of different dog genome maps that are out there. Those are the main tools that we would be looking at in a genetic project. Now the question is: Which one do you do first? Obviously cost and time are involved. It’s going to depend a little bit on that, and it’s also going to depend on how many samples you have. There have been recent papers where investigators had one affected dog (maybe the dam) and they whole genome sequence both of them because there’s only two of them. Let’s just go ahead and whole genome sequence everything because the nice thing is, with whole genome sequencing, you already have all the mutations for each of those individuals. Then it just becomes a data-sorting kind of thing. 

[32:37] For EIC, because it was done a while ago (it makes me feel old, we did the marker route. We used markers. That got us to a location on a chromosome. This is the paper here. (There’s my name, right there!) Each one of these blue dots is a marker that we tested. The higher it goes up this way, the more statistically significant it was, basically, in the difference between affected dogs and unaffected dogs. It happened to be Canine Chromosome 9. That’s great. It got us into that nice region. But we’re still stuck with somewhere in West Lafayette. Somewhere in there, is Dr. Ekenstedht. There’s still a couple of hundred thousand people to sort through. What can we do to get closer than that, to narrow it down even more? We do something called haplotype block analysis. Those words don’t matter so much as just the concept. Each two bars here is a dog. LR1 (Labrador Retriever 1) has two bars. Labrador Retriever 2 has two bars. There’s one Chesapeake Bay Retriever here. What we did was look to see which regions did they share completely? Which regions were they all exactly the same? This whole region may have been West Lafayette but now we’ve got it narrowed down to, “Keri’s in the vet school. She’s in one of the vet school buildings.” Now there’s still a couple of vet school buildings, still maybe a thousand or 1,500 people here, but now you’ve really narrowed it down. That’s what we did. And then we looked at the position to see which genes were in there. That’s what Panel B on the bottom is showing: it’s showing you all the genes that are in those regions. There’s still 15 genes in there. I can tell you, because I’ve been doing this long enough, that everywhere you look in the dog genome, there is a gene that does something in the nervous system. It just always happens! In fact, there’s a couple in here that have to do with neurological function. What you would do is you would hope that maybe only one of them in there would have neurological function. And you’d start looking at that. We’re going to skip over the fumbling around we did with the different genes. We’re going to skip over that and go straight to DNM1, which is here, and talk about dynamins.

[34:56] This is going to be a little biochemistry. I’ll keep it at a level where nobody needs to panic. I think it’s just so cool when we get through this how the mutation and the biochemistry of it totally explains the phenotype, the collapse that we just saw. There’s three dynamin genes: 1, 2, and 3. We’re just not creative people. “Let’s just number them!” And they code proteins that help in vesicle formation. Okay, what are vesicles? When you come down to the end of a nerve (you guys know your nerves conduct signals, either signals of pain or signals to move), they have little bulges at the end of the nerve (the actual cell) and there’s a little space at the end of it called the synaptic cleft. Little vesicles, like bubbles, form there. The way that nerve talks to the next nerve or the way that nerve talks to a muscle cell (to tell that muscle cell to contract) is through neurotransmitters. You guys know neurotransmitters. You’ve heard all the drug commercials on TV for serotonin and dopamine and epinephrine. Maybe some of you have to have an EpiPen. There’s plenty of neurotransmitters out there that you’ve heard of. Those neurotransmitters are little tiny molecules inside these little bubbles, and they have to get sent across space to the next nerve or to the muscle. Then those little bubbles get recycled. Dynamin is this guy (bottom right, that little gray color) that helps pinch off the bleb. It makes the new vesicle by pinching it off, or like a collar that tightens down and twists and releases it. If we don’t have neurotransmitters, if we can’t pinch off our blebs and send the blebs out and then back in where they need to be (between the nerve or between the nerve and the muscle), then communication stops. Anything coming from the brain trying to go to a muscle to, say, the pelvic limb (“you’re running right now; you need to be contracting”)—it just stops because there’s no communication happening. 

[37:10] This is just a different picture of dynamins so don’t freak out about it. Just look at these little red guys. The red guys are dynamin. This little bleb is the end of the nerve, the green bar at the bottom. As we move from left to right, the little bleb is forming. There’s neurotransmitters inside it. That red collar (the proteins of dynamin) come together, pinch and twist, and release that bleb (containing neurotransmitters) to conduct communication (nerve signals), either between nerves or from the nerve to the muscle. Dynamin is really important, obviously. 

[37:45] What I think is really cool is that we looked at that, and we were like, “Well, nobody’s ever heard of dynamin before but let’s see if we can look for a mutation in that gene,” so we employed something called Sanger sequencing. This is different than the whole genome sequencing I mentioned before. Sanger sequencing just sequences small little bits of the genome (maybe 1,000 to 1,5000 letters at a time, sometimes even less) and then we just sort of slowly walk our way across the gene. We eventually sequence the entire gene but in tiny little pieces. It’s less data-heavy. What we’re looking for is somewhere that the affected dogs were different in their letters compared to the unaffected dogs. We’re looking for mutations or variants. If we have done whole genome sequencing, we would have had the variants already. This is a different step we had to take because we opted for the marker approach in the beginning. Ultimately, we identified one letter change in the dynamin 1 gene. It’s a G to T, so guanine to tyrosine if you remember your AGCTs from genetics in high school. It was in an exon. That just matters because the exons are the parts that get turned into the protein (in the last figure, that red protein that became the collar that pinches off). This letter change changed the amino acid and that altered the protein. It meant the protein wasn’t quite working the way it was supposed to.

[39:14] The next step after that is to say, “We better check a bunch more dogs.” If they’re available! Sometimes they’re not. Like I said, I’m working on a paper right now where we only have 8 affected dogs in the entire world. We tested as many as we could. But in this case, there were a lot of Labradors and so we tested a bunch more Labradors to see if this mutation really was segregating with affected or unaffected status the way we thought it would. Here T is the mutation and G is reference (or normal). We had 104 affected dogs, and 101 of them were homozygous for the mutation. Autosomal recessive—you need to have 2 copies to express the disease, and these were all dogs that collapsed. Now, there were 3 that collapsed and looked totally normal, so they probably collapsed for a different reason. Maybe one of them did have heat stroke. Maybe one of them had a cardiac thing, like Dr. Meurs talked about. It’s not perfect, but statistically, this held up. It was highly significant. And then we had unaffected dogs. Remember, unaffected—you’re not collapsing and this is autosomal recessive, then you can be homozygous (2 Gs) or heterozygous (GT)—65 dogs and 55 dogs—and you’re unaffected so these are dogs that never collapsed. Then we had 12 dogs that never collapse but did have homozygous TT. We’re going to talk about them a little later. But anyway, statistically, this told us, “Yeah, you got the right mutation.” So that’s pretty exciting. The other thing a researcher is going to do is look and see if that gene has ever been described in any other model. By model, I mean people or mice or cats or horses or pigs. Is there a mutation in this gene that created the same disease in some other species? There wasn’t in DNM1 at the time, which made us less sure—until we found out that there was a mutation in DNM1 in fruit flies. Fruit flies have been the workhorses of the genetics world. I never thought I’d do anything with fruit flies. But the fruit flies really provided the last interesting piece to the puzzle, because they demonstrated temperature sensitivity. Now, if we could go ahead and roll video 2, that would be wonderful. 

[41:37] In this video, you’ve got 2 jars of fruit flies. One of them (and you’ll figure out eventually) are fruit flies that all have a mutation in their DNM1 gene, just like hte Labradors do. If you look over on the top left, it’s 9:05 in the morning, and you can see the temperature (28.2; that’s pretty warm). That’s definitely over room temperature. Body temperature is like 36. So it’s getting warmer. We’re going to watch most of these fruit flies congregate up at the top of the jar. But as it gets warmer and warmer and warmer, they are (and you’ve probably figured it out) in the red-capped jar. They are literally, pun totally intended, dropping like flies. These flies (with their dynamin mutation) can’t fly anymore. They’ve lost the ability to fly because it’s really warm. So it told us the mutated DNM1 was working fine until it got too hot. And then the flies all were unable to fly or, if you’re a Labrador, keep running with your back legs. This is not work that was done in our lab. This was from somebody else. But the zoom-in here is just showing you all the flies. They're not dead. They just can’t fly. They’re at the bottom because they just can’t get their lift to do their flying. Then we’re going to see, 10 minutes later, they’ve brought the temperature back down to less hot. They’re recovering. Ten minutes later, 20 minutes later, recovering, recovering. I think eventually here we’ll see all the flies recover, even that last loner fly in the red jar does eventually get back up on his wings. I think we can stop this video and go back to the slides.

[43:33] So what was really cool about this is that we were able to finally put it all together. The Dynamin 1 protein works fine, even when it’s mutated. But as soon as it’s mutated or altered, it doesn’t work as well in those high temperatures. By that, I mean body temperature. Dynamin 2 and 3 (don’t forget 2 and 3 are out there) can handle the low-frequency neurological stimulation, so baseline stuff. Dynamin 1 is the workhorse when you’re doing any sort of sustained neuronal stimulation, like strenuous exercise. (Or for these fruit flies, flying and being at the top of the jar.) So the Dynamin 2 and 3 is why this collapse is not fatal immediately. They’re also why, if you keep the dog going, Dynamin 1’s not working and it’s gotten too hot—2 and 3 eventually just can’t keep up, and that’s why it can progress to becoming fatal. That means this is a mutation that’s loss-of-function, temperature-dependent, and totally reversible because when the dog cools off (the body temperature), those muscles work fine again and Dynamin 1 is working fine again.

[44:48] We can totally connect this mutation to the disease. The dysfunctional Dynamin 1 stops working when the muscles get warmer. We know muscles get warmer when they contract. All of us have experienced this. We exercise, we get warm. This is where I miss getting to see you guys in person, because I usually get a little interactive. Which muscles in the body are the biggest? This is when people have a little Charlie Brown parent action, can’t hear anything, and then someone will finally yell out, “Hey, my butt muscles are pretty big!” Yeah, your gluteal muscles are actually one of the biggest-size muscles in your body, as are everything in your upper pelvic limb, so your thigh, the quadriceps, and the hamstrings. Those are the biggest muscles you have in your body. They get the hottest first. That’s exactly why we see pelvic limbs—it’s the same in dogs. Those are the biggest muscles they have. They get hottest first, and that’s where Dynamin 1 stops working first. That’s why we see the collapse that looks exactly the way we saw it in the video, which I just think is so cool. So, carriers! If a carrier has 1 mutated copy of Dynamin 1 but the other one is not mutated, there’s enough functional Dynamin 1 that they actually can perform just fine. We know lots of carrier dogs that hunt test and field trial, and they’re totally fine. They’re truly unaffected. 

[46:10] The next step in research is to establish the frequency, so we’re going to try to figure out how common is this mutated copy amongst all the copies out there in the breed. We found about a 3% rate of affected Labradors (Labradors that had 2 copies of the mutation). And we found about a 37% rate of carriers. I want you to remember that number; that’s why it’s bolded and underlined: 37%. This was a combination across all kinds of Labradors—field trial, hunt test, conformation (which are the show Labradors), pet Labradors, service lines of Labradors. It’s probably a little biased because of course we were studying the disease, so people were sending us affected dogs. Updating that would require a truly unbiased sample cohort, totally randomized, which as far as I’ve seen in the literature, nobody’s tried to update. I think the numbers have probably gone down. I would hope that the affected number has dropped to almost 0 because there’s no reason anybody should be producing an affected dog anymore. Of course, if you submit a whole litter, those dogs are going to be pretty closely related and so if you’re including whole litters in your numbers, that’s going to skew your numbers as well. Those are the numbers that we had. I would expect them to be a little lower these days. 

[47:27] So what’s next? Next, if the researchers are really, really sure (and we were) that we had the positive mutation identified, then they’re probably going to offer the test so that breeders can start using it. As a researcher, I want healthy puppies as much as you do. We’re not going to sit and not tell you and not offer this test as soon as we can. Some researchers will choose to just license it directly to a commercial company. Other researchers are set up to be able to offer that test through their laboratory. Usually, any profits (at least I can speak for my lab and Minnesota) that are coming in for the testing were just funneled right back into the lab, so keeping the technicians employed and advancing additional research. Again, it was set up in a not-for-profit sort of way. Any samples that had been previously submitted got their test results for free. If you helped contribute to the research in the first place, then once we’re ready to offer the test, we put together the report and everybody whose sample we already tested, we get the results out and you get that result for free. I think some people should look at that and think, “Hey, I want to participate right away, upfront, and yeah, I’m going to have to pay shipping or whatever, but at least I’m going to get results (if there’s results to have) someday.” Usually, the researchers will designate a cutoff point and say, “Anything submitted after this is not going to be free anymore, and you’ll have to pay for the testing.” If it’s the right timing, the research might be presented at a scientific conference. We have a dog and cat genetics conference—international—that happens every couple of years. Several times, I’ve spoken at National Specialities for a specific breed to let the breeders know. And then publish a paper—that’s the next step. That’s the scientific review process, which is truly the actual bar that any science should have to cross. After something’s been published and it’s available in the public domain, then the test is likely to be picked up by a bunch of different commercial testing companies, sometimes added to panels and so forth. That just increases your options. It’s certainly much more efficient if you can get 5 tests done at once, right? Rather than having to do them one at a time. 

[49:44] The good news is prognosis for EIC affected dogs is pretty good. They can live good lives as family pets, but they usually can’t stay (if they’re having collapse episodes during field trial), then they probably can’t keep doing field trials. There’s things like that that have to be limited: intense exercise, excitement, that sort of thing. As they get old, they tend to collapse less. This picture is just from a friend of mine, Susan, and her dog, Scooter, who’s my god-dog. I was his first veterinarian when he was a little baby puppy. You can see he’s quite gray. He’s an old man. I think he’s 16. He never had EIC, never collapsed, but if he had when he was younger, he has slowed down. So collapsing becomes less of an issue as dogs slow down their energy levels.

[50:28] Other research steps that the researchers might do is they might want to look at a bigger population for the mutation, and they might want to check other breeds. We published this paper as a follow-up, and what it did was it finally actually separated out the different lines in Labradors, so it looked at pet Labradors, field trial Labradors, show Labradors. I will say one of the kind of surprising things that we found is there were a lot of show Labradors that had this mutation, homozygously, and they were not collapsing. (We’ll talk about that in a second.) 

[51:02] And then we looked at other breeds. We did find this mutation in other breeds. This is the list that was published in that paper. Here, this speaks to what Dr. Oberbauer was saying as well. It doesn’t mean that these breeds collapse. This is where genetics can get a little more complicated. I put this OES in here, an Old English Sheepdog, because we found some OES that were homozygous and they looked like they should be affected, but have any of you met an OES that has the drive and crazy personality as a Labrador? I haven’t. Every OES I’ve ever met has been sort of the dog equivalent of a stoner. They’re so chill. I don’t mean that in an offensive way. I mean that as they’re just really chill dogs. They just didn’t collapse. I don’t think we’ve ever seen any Pembroke Corgis that actually collapsed either. 

[51:55] There’s something interesting going on here, and that’s called variable penetrance. That’s the geneticist word for it. There were some dogs that had the disease genetically (this gets to what Dr. Meurs was talking about in terms of risk). They were at risk. They never collapsed. This might be those show Labradors. Again, the show Labradors are (as a bunch) kind of dialed down a little bit from the more hyperactive, highly-driven field trial Labradors. There’s some selection differences in terms of personality. We see this incomplete penetrance where some of the dogs did not show the disease. Maybe it’s because they’re just not participating in strenuous activity. Okay. We’re going to be honest: trotting around a show ring is not as intense as flying as fast as you can after the duck from the slingshot. Guide dogs—they have stressful lives. It’s a different kind of stress. They don’t seem to collapse as much. The difference in these clinical signs seems to depend on lifestyle as well as temperament. This is where some of the OES and some of these other breeds come in, too, where nobody studied it. Nobody has validated whether this mutation in this breed has any actual meaning, clinically. We can say for sure: it does in Labradors. It does in Chesapeake Bay Retrievers. It does in Curly Coated Retrievers. We’ve shown that. But in other breeds, it’s a big question mark. To my knowledge, nobody has studied it to any significant level. 

[53:20] There’s another paper that came out in 2018 that found the EIC mutation in the Cotons, the Parson Russells, and the Rhodesians. Again, to my knowledge, these breeds aren’t doing a lot of collapsing but we also have to recognize that there might be some risk there. Right? It would be good information to know. I think if you look at some of the testing companies that are out there who are offering this test, there probably are other breeds on this list. They just haven’t been published. My question that I say is: “Is this test applicable?” Without validation, without some sort of scientific validation, it’s hard to answer. We need to be tracking that. They’re areas that are certainly ripe for additional research. I should also throw out here at this point that EIC has definitely been seen in mixed-breed dogs, not just as carriers, but we’ve actually seen homozygous mixed-breed dogs where maybe they had some Labrador from one side and a different Labrador from another side, and they were still mixy-mixed. They still ended up with 2 copies of this mutation. So that can happen, even if you don’t have a purebred dog. 

[54:33] The other thing that’s really cool—and this is why I said we can say that that one Labrador stud dog who did contribute to spreading the mutation around Labradors is not where the mutation originated—we can say it originated before the breeds were even formed, because I don’t think there’s a lot of crossover happening between Labradors and Cotons. Right? This allele (this mutation) goes pretty far back. We call that “ancient,” even though it’s a couple hundred years. Ancient. As I said, we’ve seen it in mixed-breed dogs, both as carriers (1 copy) and as affected (2 copies), so I had to throw this in here. Doodle breeders, I know there’s quite a few Doodles in this group. Be careful. If all you ever do is cross a Labrador to a Poodle (to my knowledge), EIC has not been found in Poodles. The variant has not been seen in Poodles. If all you ever did was cross a Labrador with a Poodle, you’d be fine because the worst you’d ever do is make a carrier. But if you’re then taking your half-and-half (your Labrador and Poodle crosses) and crossing them together, now you can end up with Labrador coming from both sides in a Labradoodle type scenario. So you do have to be careful. Even if you’re doing breeding like that, it’s still something you need to be aware of.

[55:49] So breeding recommendations: best treatment now is prevention. We should not be making affected puppies. We’re going to talk about which breedings are okay. I want you to remember that carrier frequency. We’re going to talk about Labradors as our model here. The carrier frequency was 37%, so 37% of the dogs had at least 1 copy. 

[56:11] Now, Dr. Oberbauer talked about this a little, and this is the other place where usually I would get the audience interacting with me. We’re going to just have to do this without that. In this slide, N (or blue) is the reference (or normal allele, so the non-mutated version). Red is going to be the E (or mutated) version. If you remember your Punnett squares from way back in high school, you put the 2 copies your sire has, and you put the 2 copies that your dam has, and you start making puppies in these 4 squares here. We have done that with all the potential crosses. I think all of us can agree that it’s totally fine to breed a clear dog to a clear dog. We approve. And I think we can all agree that nobody should be making affected puppies, so we should not be crossing affected with affected. We shouldn’t be crossing affected with carrier. Purple here is 1 of the normal copy and 1 of the mutated copy. We don’t want to make these affected puppies here. We should not be breeding carrier to carrier, because 25% of the time, those puppies are going to be affected. That’s how we found this disease in the first place. We’re going to cross this one out, this one, and this one. Now this one: you’ve got an affected dog. It’s probably not an ideal breeding. You would never want to breed this dog to anything other than clear. Every single puppy in your litter is going to be a carrier. That’s not ideal. But there are some genetically affected dogs out there that don’t collapse. Breeding is so much more than just one test. You’ve got to think about all the things, right? The hip dysplasia test, the elbow dysplasia test, the personality, behaviors, all the things that you select for. If this dog is amazing—really amazing—and if you bred it to a clear, I do not have an ethical problem with you creating all carrier puppies. You would want to make sure that none of them were going to get bred to another carrier. We don’t want this happening. So, some people have problems with that, but I say that an EIC affected dog still has a decent life as long as you take care of it the right way and manage it the right way. And if everything else was really good about the dog, we’re not going to throw out the baby with the bath water here. But we’re going to be really selective about who that dog is bred to. And then this one: this is breeding a carrier with a clear. Well, half of those puppies are going to be clear, and half of them are going to be carriers. I say this is totally fine. In fact, I encourage you to do this. Here’s why: remember the carrier rate in the breed. Almost 40% of all the dogs out there already have 1 copy. They look like this. They’re purple. The knee-jerk reaction (and I hear this from veterinarians all the time) is “You should just spay and neuter all your carriers. Just get them out of the gene pool.” Well, okay—if you want new disasters! If you take 40% of your gene pool out and get rid of it, you are going to have disasters. We actually encourage it—especially if this carrier is a really great specimen, a really great example of everything else you want to breed for and it just happens to have 1 copy of this mutation. Fine! Breed it to a clear dog. Make sure you’re breeding to a clear dog, so you’re not producing any affected puppies. Then test the litter. Slowly, over generations, maybe this little guy here (another carrier) is your next show prospect. But maybe the clear one is. That’s the one you keep, and that’s the one you keep breeding. Slowly, over time, that mutation can be eliminated from the breed. But we don’t want to do anything catastrophic. So please, please, please if there’s nothing else you remember from this talk, for any autosomal recessive condition (especially if the carrier rate is really high in the breed), we want to make sure we’re not creating new problems by rapidly eliminating too much of the gene pool that goes along with if you were just eliminating your carriers. Hopefully that makes sense. This is the point where I really wish I could see your faces and see if you’re tracking with me or not. 

[1:00:22] I promise I’m almost done! So what happens next? All that stuff got published. You know, research always produces new questions. That’s what’s really great about it. We started testing these Border Collies, and they did not have the DNM1 mutation, so it was not the same as EIC. Now, as I describe it, it’s going to sound like EIC. They would do some strenuous exercise, and it would trigger this collapse, and it would look neurological and they were normal the rest of the time. They recovered after 5 minutes or up to half an hour, and they were normal again. Weird, right? Sounds like EIC. But the difference is they were quite disoriented. They had dull mentation. They were not responsive to owners’ verbal calling. They would sway and stagger. It’s not just restricted to Border Collies. It has been seen in some of the other herding breeds. Voila! New disease to study. So we’re going to circle right back to where we started at the beginning of this. We’re going to try to describe it clinically. Sue Taylor did—I don’t think she’s published it yet; I haven’t checked—the same kind of clinical investigations that she did on the Labradors when EIC was a new thing. And then of course there’s genetic investigations ongoing still at the University of Minnesota. If you want more information on that, just google “University of Minnesota Border Collie collapse.” It’ll come right up. I can tell you it doesn’t look like it’s going to be a single-gene disease though. It looks like it’s going to be more complicated, unfortunately. It just makes it harder to solve.

[1:01:47] Let’s wrap up! EIC as a case study. These single-gene diseases are what I call the “low-hanging fruit.” If we can get our hands on them, we have the genetic tools now where we can usually solve them pretty easily, and then help breeds avoid producing affected puppies. The complex and polygenic diseases are definitely harder. If we could have an easy test, genetically, for hip dysplasia, we would, right? That just hasn’t happened yet. I’m sure Dr. Smith will mention that. I do think that new single-gene diseases are going to keep popping up. That’s just because of the way dog breed structure is. If we just let our dogs all free-for-all, mate, and we weren’t kind of keeping our contained gene pools of this breed or that breed, it would be less of an issue. But now we have all the tools where it’s not such a big deal, and we can really work with breeders and hopefully catch them quite early. Like I said, I have this other paper I’m working on right now. We caught it so quick there were 8 affected dogs, and that’s it. And now there don’t ever have to be any more. That was pretty fun. The scientific process itself is really fascinating, but it can’t be done without teamwork and without the really involved help of great breeders. 

[1:02:58] With that, I have no idea where I’m at with time. I don’t know what time I started. I don’t know what time I was supposed to end. Everything got all kind of goofed up, so I don’t know if I have time to take questions now or not. Probably not. We could probably just save questions for me. I’ll be back on the panel. These are all my critters. No, there is not a dog in any of these pictures. Yes, these are two different cats. I just tell people that’s because if I had a dog, I would be biased. This way, I can love all the dog breeds the same. 

Dr. Judi Stella [1:03:33] Thank you, Dr. Ekenstedt. I’m sorry that we’re not going to have time for Q&A. Trying to keep us within our time frame. But we will have you back for the panel discussion, so we can ask some questions then. Just so everyone knows, we are going to skip the coffee breaks for the rest of the afternoon and just keep going with our speakers. Thank you so much, Dr. Ekenstedt. We look forward to having you on the panel. That talk was amazing. It was really, really useful, good information for everybody.