Free magazine for dog enthusiasts everywhere K9 Perspective on-line magazine. Dog information resource. Go to page one of this issue Go to page 19 of K9 Perspective issue 5 Go to page 21 of K9 Perspective issue 5 mans best friend

Sleeping dogs don't lie:

Discovery of the canine
narcolepsy gene

By Mark Springer

Doberman pups
A litter of dobermans born from a back-cross of a narcoleptic doberman and a normal dog. Prancer, one of the pups pictured here, grew up to become a cover dog for the journal 'Cell'.
FOR the first time, life science researchers have isolated a disease gene in dogs that provides clues to the origins of a disorder also found in humans. A 10-year hunt for the gene responsible for canine narcolepsy culminated in the discovery of a gene named Hcrtr2 (hypocretin receptor 2). The canine narcolepsy gene codes for a hypocretin cell receptor in the hypothalamus region of the brain. When mutated, the Hcrtr2 gene causes narcolepsy in three different breeds of dogs.

Apparently, mutations in the gene disrupt a signalling system in the brain that promotes wakefulness in Doberman pinschers, dachshunds, and Labrador retrievers. The hypocretins, a class of molecules previously not known to be associated with sleep regulation, may also play a significant role in human narcolepsy and other sleep disorders. Led by Dr. Emmanuel Mignot, a team of researchers from the Stanford University Center for Narcolepsy in Palo Alto, California, used the ABI PRISM(r) 377 DNA Sequencer from PE Biosystems (now Applied Biosystems) to help them find the DNA sequence and chromosomal location of the Hcrtr2 gene [Cell. 98: 365-376, August 6, 1999].

With the help of the DNA sequencer Mignot's group developed hundreds of new molecular markers for the dog genome. These markers allowed them to isolate and identify the gene by positional cloning. In addition, the DNA sequencer automated mutation analysis techniques, which accelerated the researcher's discovery of the causative mutations in the canine narcolepsy gene.

Tired pup probably narcoleptic
This doberman puppy shows signs of excessive daytime sleepiness, one of the signs of canine narcolepsy.
Using the most informative markers to gradually isolate the disease gene, a team of 10 Stanford researchers eventually pinned the Hcrtr2 gene to chromosome 12 in canines, identifying three different mutations in the same gene. Dachshunds,

Labrador retrievers and Doberman pinschers each have a different mutation, but all have damage to the same gene. In each case, restriction fragment length polymorphism (RFLP) analysis revealed a flaw in the gene.

"For the dachshund, we have a number of sporadic cases and one familial case," notes Mignot, co-director of the Stanford Center for Narcolepsy (with Dr. Seiji Nishino) and the principal researcher in the decade-long project. "In Dobermans, we have only one familial case. For the Labradors, we found mostly familial cases and one that was reported as sporadic."

Mignot noted that there were a few cases of canine narcolepsy in which attempts to breed carriers did not result in genetic transmission of the sleeping disorder. In those cases, the situation may be more complicated than a monogenic autosomal recessive inheritance of narcolepsy.

New role for hypocretins in sleep
Hypocretin ligands - also called orexins - are peptides that bind with hypocretin cell surface receptors. They are found strictly in the hypothalamus region of the brain. Although there are only about 1200 hypocretin-producing cells - located on both sides of the brain - arrays of axons extending from the hypothalamus distribute the neuropeptides widely throughout the brain. "Hypocretins are positioned to affect a lot of different systems in the body - including known sleep centres - as neuromodulators of daily activities," says Mignot.

Excitement causes sleep attack

Cataleptic attack:
1. The dog spots a treat.
2. He dines enthusiastically.
3. Overcome with the emotion of his good fortune, he is down for the count (a few seconds of muscle paralysis).

Before Mignot's recent discovery, researchers thought that hypocretins were solely involved in regulating aspects of feeding behaviour. Now, after the isolation of the canine narcolepsy gene, more evidence exists for the role of hypocretins in regulating sleep behaviour than for their role in feeding behaviour, according to Mignot.

"I believe that the hypocretin system will eventually be found to be one of the key neurotransmitter systems for sleep regulation and will become as significant as such well known neurotransmitters as acetylcholine and norepinephrine," Mignot predicts.

Breeding dogs with narcolepsy
Interestingly, dogs with narcolepsy show many of the same symptoms that are found in narcoleptic humans. Although no one knows whether dogs dream or not, they do show brain activity during REM sleep similar to that of humans. By attaching electrodes to the heads of sleeping dogs, sleep specialists have identified in the dogs increased brainwave activity characteristic of REM sleep. These kinds of tests enable sleep specialists to establish which dogs show early onset of REM sleep, a phenotypic trait of narcolepsy. In addition, narcoleptic dogs exhibit another tell-tale sign of narcolepsy: cataplexy. Just as ataplectic attacks in narcoleptic humans are brought on by extreme emotions, the same is true for narcoleptic dogs. In narcoleptic dogs, the sight of a snack or the emotion of playing together ignites the abrupt, temporary paralysis of all postural muscles, suddenly dropping a group of excited dogs into a silent heap on the floor. Attacks last only a few seconds, but can occur hundreds of times a day.

"Dogs experiencing cataplectic attacks are not asleep," Mignot notes. "They are awake, but momentarily paralysed. Their eyes are open and they can move them, but their postural muscles are completely relaxed."

In the late 1970s, A.S. Foutz et al. first discovered that dogs inherit narcolepsy in an autosomal recessive manner with full penetrance (Sleep 1: 413-421, 1979), meaning that if they receive a bad copy of the gene from each parent they will express the disease phenotype. This study associated canine narcolepsy with a disruption of an unknown gene, which was named canarc-1. However, no one knew where within the 38 dog autosomes (non-sex chromosomes) canarc-1 resided.

Decade-long hunt for the narcolepsy gene
In 1989, shortly after joining the Stanford Center for Narcolepsy, Mignot launched his quest to establish the genetic underpinnings of narcolepsy. Mignot came to the Stanford Center for Narcolepsy 13 years ago after studying pharmacology at Necker Enfant Malades medical school, in France. There, he had focused primarily on the effects of drugs on the brain. Shortly after he arrived at Stanford, he became interested in sleep research and today is one of the leading authorities on the molecular basis of sleep disorders. Believing the human genetics to be too complicated, Mignot decided to undertake linkage analysis studies with dogs because the disorder had already been established in canines as a single-gene autosomal recessive trait.

Sleep comes at any time of day
Bo, a narcoleptic dachshund, asleep in the grass.
For linkage analysis studies, researchers study families, or isolated populations, and attempt to trace phenotypic traits through generations of offspring. When successful, linkage analysis associates specific genes or characteristic genotypes with select observable traits such as symptoms of a disease. The deliberate breeding of closely related animals makes canine breeds akin to geographically isolated human populations, ideal for linkage studies.

Mignot's exhaustive linkage studies would not have been possible without the foresight of Dr William Dement, director of the Sleep Disorders Research Center at Stanford. It has been 20 years since Dement, a pioneer in sleep research, who studied under Nathaniel Kleitman, the discoverer of REM sleep, first became aware that narcolepsy can affect dogs. After giving a lecture on human narcolepsy, Dement was approached by a woman who mentioned that she had a dog that showed the same symptoms of narcolepsy that he had just described in his talk. A short time later, researchers at the Stanford Center for Narcolepsy started seeing dogs with narcolepsy.

Initially, they reviewed only sporadic cases, such as occur in poodles and dachshunds. But, after acquiring a number of narcoleptic dogs, they tried to breed them; a difficult task considering that excitation often brings about cataplectic attacks. Still, the sleep specialists persevered and after a few years established a colony of hundreds of narcoleptic dogs, including Prancer, a narcoleptic Doberman who recently appeared on the cover of the journal Cell {Cell. 98: 365-376, August 6, 1999].

"For every dog that is born in the colony, we first look at the phenotype to see if it is narcoleptic and then look at the DNA," notes Mignot. To establish linkage, Mignot performed back-crosses with hundreds of dogs, breeding recessive homozygote narcoleptic dogs with a heterozygote. Eventually, these breeding practices linked a particular locus marker with the signature traits of narcolepsy. Linkage analysis, a labour-intensive effort under the best conditions, was further complicated by the fact that in the early 1990s, the time of Mignot's studies, no dog genome markers were available to the Stanford researchers. Molecular markers are essential for isolating genes. They serve as signposts throughout the billions of base pairs of DNA, helping researchers to pinpoint the exact chromosomal location of a gene.

For humans, sets of markers had been developed for each chromosome, but for dogs, no such set had been developed at the time of Mignot's linkage studies.

"Initially, we found linkage pretty much by chance," notes Mignot. Using a DNA fingerprinting technique and a candidate gene approach, Mignot hoped to find a polymorphic marker that co-segregated with the sleeping disorder in several generations of dog offspring. Fortunately, he found one, an immunoglobin cross-reacting DNA segment that completely co-segregated with the narcolepsy phenotype.

"It was pure serendipity that when we found this marker it appened to be on the same chromosome as the canarc-1 locus for narcolepsy," notes Mignot.

Of course, when his team of researchers first found the marker, they did not know where it was located. It could have been on any one of the dog's 38 autosomes.

At first, the Stanford researchers thought that immunoglobin genes within the cross-reacting DNA segment might be associated with narcolepsy. After all, previous studies in humans had shown an association of two human leukocyte antigen (HLA) alleles with the sleeping disorder. As HLA alleles code for specific antigens that help the body distinguish between self and non-self, and immunoglobin genes code for various parts of antibodies, agents which attack foreign antigens, it seemed plausible that immunoglobins might have something to do with narcolepsy in dogs.

"We cloned the immunoglobin genes, but we did not find that canine narcolepsy was immunoglobin related. We concluded that the marker was not functional in narcolepsy, but it was just a marker, so we were momentarily stuck," Mignot concedes. Usually, at this point, researchers can use sets of previously developed markers and through positional cloning narrow down the location of a linked marker, but genome libraries of markers for dogs did not exist.

Taking a dog genome walk
With a marker of several thousand base pairs in hand, but nowhere to place it in the dog genome, Mignot's team faced a daunting task: taking a chromosome walk through billions of base-pairs of DNA to find out on which dog chromosome the marker resided. "In the middle of the project, I realized that there was no way that we would get through this project unless we had an automated sequencer," says Mignot. Enter the ABI PRISM(r) 377 DNA Sequencer. The slabgel DNA sequencer automates electrophoretic separation of bases or chemical letters in DNA samples. It can decipher and spell out in real-time the sequence of hundreds of DNA bases in a matter of hours, dramatically simplifying large-scale sequencing projects like Mignot's.

"I used the sequencer for almost everything. We used it every day, and ran it around the clock," Mignot notes.

Turning to BACS
To walk through the dog genome, Mignot first tried bacterial viruses, or phage, as cloning vectors. But phage can only accommodate clones with 15,000 base-pair segments per insert. Considering it takes about a week to perform all the associated molecular biology to walk the length of just one of these 15 kb segments, Mignot calculated it would take around 40 years to walk across the dog genome.

He needed to find a vector that would accept larger-sized DNA inserts. The answer appeared to be bacterial artificial chromosomes (BACs). With BACs, the researchers were able to walk 100,000 bases with each insert or step, but still there were so few recombination events that this approach was taking too long.

Fishing for synteny
At this point in the project, Mignot made a major breakthrough. He decided to take advantage of what was already known about the human genome and search for regions of conserved synteny, known stretches of DNA in humans that are the same in dogs and other mammals. He found a human HLA gene named Myo6.

This human gene had a similar DNA sequence to that of the immunogloblin dog marker, and had already been mapped to human chromosome 6. After developing a probe based on the Myo6 gene sequence, Mignot performed fluorescence in situ hybridisation (FISH) and successfully labelled the homologous gene on a dog chromosome spread. The fluorescently labelled Myo6 probe lit up a region on dog chromosome 12.

Once the researchers knew that chromosome 12 held the marker linked to canarc-1, their work had just begun. Now they needed to identify dozens of additional markers before they could establish the precise location and the DNA sequence of the narcolepsy gene. Again, for this important task, they leaned heavily on the sequencer. Mignot's research team isolated microsatellite markers using mini-libraries of dog DNA. They built these libraries with selected genomic BAC clones from canine chromosome 12.

Following overnight incubation, they extracted plasmid DNAs from all positive bacteria colonies and sequenced them on the 377 DNA Sequencer. By using the automated DNA sequencer for direct sequencing of the ends of BAC clones, the researchers developed the majority of markers necessary to narrow the search for the canine narcolepsy gene.

"Using the sequencer, we constructed the mini-libraries, sequenced all the clones and developed enough microsatellite markers to flank the region around canarc-1. "It was a Herculean task," notes Mignot. By the end of the project, Mignot's team had developed hundreds of microsatellite markers scattered throughout canine chromosome 12. A third of those markers turned out to be informative, bracketing the gene of interest, and helped to identify the exact chromosomal position of canarc-1, which turned out to be the Hcrtr2 gene. "The two best things I ever did for this project were to build this BAC library with Dr. Pieter de Jong, and to devote the sequencer to this project," Mignot remarks.

Mutation analysis fingers broken cell receptors Following positional cloning of Hcrtr2, RFLP analysis suggested that a possible genomic alteration in the vicinity of, or within, the canine Hcrtr2 gene was the cause of canine narcolepsy. Genomic sequencing, again performed with the sequencer, was used to identify the mutation in Dobermans as a 226-base-pair short interspersed nucleotide element (SINE) inserted in the third intron of the gene, upstream of the fourth encoded exon. The SINE insertion results in aberrant splicing of the Hcrtr2 gene transcript, leading to production of a truncated non-functional hypocretin receptor. The non-functional hypocretin 2 cell receptor causes narcolepsy in Dobermans. Surprisingly, the researchers did not observe the SINE insertion in canarc-1 positive Labradors or dachshunds, suggesting that other mutations in the Hcrtr2 transcripts might be responsible for the abnormal mRNA molecules they found in these two breeds.

A similar analysis of Hcrtr2 gene mutations for all narcoleptic dogs studied revealed that, in all cases, a truncated hypocretin receptor 2 protein product resulted in either a non-functional hypocretin receptor or one that cannot fit properly in the cell membrane. - Abridged.

Reprinted with the kind permission of BioBeat(r) Online Magazine (, Applied Biosystems.

Go to page 19 of K9 Perspective issue 5
Issue 5Page 20
Go to page 21 of K9 Perspective issue 5

Copyright 2001-2010 Paperclip Publishing
All rights reserved