Congenital Myasthenic Syndrome in Labradors (COLQ): The 2014 Discovery

SamOur Lab pup lies down after a few minutes of fetch. His parents were EIC clear, so what else could it be?

Elena MarshSeveral things, and one of them is rare enough that few owners have heard of it. A 2014 Clemson and Texas A&M paper described two Labrador sisters whose weakness followed exactly that pattern.

SamThe breeder’s paperwork lists a dozen DNA results. Wouldn’t a serious one already be on there?

Elena MarshNot necessarily. CMS is not on the Labrador’s CHIC list even as an optional test, and OFA’s own Labrador statistics hold only 269 results for it, against more than 325,000 hip evaluations.

SamIf a test exists, should I just order it and stop worrying?

Elena MarshA result answers one narrow question about one letter of DNA. The 2014 team screened 288 unrelated Labradors and found none carrying it, which tells you how rare it is and how much a single result can and cannot cover.

Bottom line: Congenital myasthenic syndrome (CMS) in the Labrador Retriever is a rare, autosomal recessive failure of nerve-to-muscle signalling caused by one change in the COLQ gene, c.1010T>C (p.I337T). Researchers at Clemson University, Texas A&M and UC San Diego identified it in 2014 in two littermates whose weakness got worse with exercise and worsened further under a standard myasthenia drug challenge. Only dogs with two copies are affected; carriers stay healthy. A DNA result is risk information, not a diagnosis. A weak puppy needs a veterinarian’s neurological workup, and a clear result does not exclude other forms of CMS, acquired myasthenia gravis or exercise-induced collapse.

The Labrador spent 31 years as the most registered breed in the United States and has now sat at No. 2 for the third year in a row behind the French Bulldog, according to the American Kennel Club’s 2025 ranking. A breed that common already has a long list of DNA tests attached to it. COLQ-related CMS is one of the tests that most Labrador owners never see discussed, partly because only one affected family has ever been published. This article works through that family, the gene, the numbers that have accumulated since, and the practical choices a buyer, owner or breeder in the US or UK can make with them.

Two sisters, nine pups and a drug challenge that backfired: the 2014 discovery

The 2014 PLoS One report by Rinz and colleagues began at the veterinary teaching hospital of Texas A&M University. Two female Labrador puppies, six weeks old, arrived with a three- to four-week history of exercise-induced tetraparesis: weakness in all four legs that appeared when they moved and eased when they stopped. They came from a litter of nine. Six littermates showed no signs, and one had died at birth. Both parents were clinically normal and had already tested clear for exercise-induced collapse (EIC) and centronuclear myopathy (CNM), the two neuromuscular conditions a Labrador breeder would normally screen for.

On neurological examination the pups walked with a short, choppy stride that shortened further the longer they moved, and their spinal reflexes were reduced. Electrodiagnostic testing showed what neurologists call a decrement: when a nerve was stimulated repeatedly, the electrical response recorded from the muscle (the compound muscle action potential, or CMAP) fell with each pulse, at every stimulation frequency tested. Nerve conduction velocity was normal for their age and needle EMG was unremarkable. Muscle and nerve biopsies showed no specific structural change, which argued against a congenital myopathy or neuropathy. Neither autoantibody against the acetylcholine receptor (AChR) was detected in serum or bound to muscle, which ruled out autoimmune myasthenia gravis.

The clue the authors singled out came from a drug challenge. Medicines that block acetylcholinesterase are the standard way to unmask myasthenia gravis, because they leave more acetylcholine at the junction and usually produce a short burst of strength. In these pups the challenge made the weakness worse. The authors read that worsening as a sign of acetylcholine overload desensitising the receptors, which fits a synaptic form called end-plate acetylcholinesterase deficiency, in which the enzyme itself is missing from the junction. One puppy was euthanised at seven weeks as its signs progressed; the other was euthanised after electrodiagnostics and tissue sampling.

SamSo the hint was that the usual myasthenia test made them weaker, not stronger?

Elena MarshExactly. Rinz and colleagues flagged that paradox as the reason to look at the enzyme’s anchor. They then genotyped about 173,000 SNPs across nine family members and found one candidate region that fit.

The genetics followed. The team genotyped 173,662 SNPs in seven littermates and both parents and checked 18 candidate genes known to cause CMS in people or animals. Only the region around COLQ, on canine chromosome 23, was identical by descent in the affected sisters. Sequencing found a single substitution in exon 14, c.1010T>C, replacing a conserved isoleucine with threonine (I337T) in the C-terminal domain of the protein. Both affected puppies carried two copies. Sixteen relatives carried one. None of 288 unrelated Labrador Retrievers or 112 dogs of other breeds carried it at all. It was the first COLQ variant described in a dog, and the paper also notes that two human CMS patients had been reported homozygous for the same change. The authors suggested that linebreeding in the family’s background had allowed a rare variant to meet itself.

An enzyme without its tether: why a missing anchor weakens muscle

A nerve tells a muscle to contract by releasing acetylcholine across a narrow gap, the neuromuscular junction. To end each signal, acetylcholinesterase (AChE) has to break that acetylcholine down quickly. In skeletal muscle much of this enzyme is held in place by a collagen-like tail, the product of COLQ. In 1998, a Mayo Clinic group showed in human endplate acetylcholinesterase deficiency that this tail anchors the asymmetric enzyme into the basal lamina of the endplate, and that recessive COLQ mutations make the normal asymmetric form disappear from muscle. The human condition is a severe congenital myasthenic syndrome.

The 2020 classification by Mignan, Targett and Lowrie in the Journal of Veterinary Internal Medicine sorts the canine and feline syndromes by where the fault sits: before the gap (presynaptic), in the gap (synaptic) or on the muscle side (postsynaptic). COLQ falls in the synaptic group, labelled acetylcholinesterase deficiency. That placement explains why the Texas A&M puppies reacted badly to a drug that inhibits an enzyme they were already short of. It also explains why COLQ disease sits apart from receptor-deficient syndromes, which can look identical at the bedside.

The inheritance is autosomal recessive. A dog needs a copy from each parent to be affected. A dog with one copy has a working second copy and, in every family studied, has been clinically normal. Gene databases add one quirk worth knowing if you search online: OMIA entry 001928-9615 lists both canine COLQ variants, with the Labrador change recorded as variant 900 and its first report dated 2014. The breed list there contains just two names, Labrador Retriever and Golden Retriever, and the variants are different in each.

Cleared for EIC, untested for this: the second collapse gene in Labs

Ask a Labrador buyer what makes a young dog fold after hard exercise and the answer will usually be EIC. That instinct is reasonable. EIC is common in the breed, and the OFA/CHIC programme for the Labrador Retriever requires a DNA-based EIC result from an approved laboratory, alongside hips, elbows, an ACVO eye exam and the D locus (dilute) test. CNM, a cardiac exam and prcd-PRA are listed as optional. Congenital myasthenic syndrome does not appear in either column.

The Labrador Retriever Club’s 2016 health statement to the AKC describes the same set: OFA hips, elbows, annual eye exams and EIC testing, with CNM, prcd-PRA and dilute recommended. Across the Atlantic, the Royal Kennel Club’s Labrador page lists Good Practice DNA tests for CNM, EIC, MCD, prcd-PRA, SD2 and STGD, with HNPK added under Best Practice. CMS is absent there too.

The practical consequence is narrow but real. EIC and COLQ-related CMS are caused by unrelated genes. A puppy whose parents are both EIC clear, as both parents in the 2014 family were, has said nothing about COLQ. Wisdom Panel’s condition page describes the Labrador form as producing collapse after 5 to 30 minutes of exercise, with recovery after a few minutes’ rest and a return of signs if the dog keeps running. Many owners would read that description as EIC. The age differs somewhat: the company places first signs at 12 to 16 weeks, while the published pups were already affected at six weeks, and a 2026 Dobermann study by Shelton and colleagues, summarising earlier CMS variants in six breeds, put onset at roughly 6 to 12 weeks across the board.

SamSo a breeder who did every CHIC test could still sell me an affected pup?

Elena MarshIn principle, yes, if both parents happened to carry COLQ c.1010T>C. The 2014 parents were EIC and CNM clear and still produced two affected daughters, because neither of those tests looks at this gene.

That does not make the absence a scandal. A breed club adds a test when evidence shows a condition is a meaningful problem in the breed, and one family with two affected puppies is a thin basis for that. It simply means the CHIC number on a pedigree should not be read as a statement about this disease in either direction.

Three numbers that measure different things: 0.217%, 1 in 5,000 and 269

The largest population figure comes from the 2023 PLoS Genetics analysis by Donner and colleagues, built on genotypes from about 1.05 million dogs tested through Wisdom Panel. Among 16,856 Labrador Retrievers, the allele frequency of the Labrador COLQ variant was 0.217%. That is the share of all gene copies in those dogs, not the share of dogs that are carriers, and the two should not be swapped. Across the whole dataset, 187 dogs carried one copy and none carried two. The Golden Retriever variant was not found in any Labrador, and the Labrador variant was not found in any Golden. In purebred dogs, neither variant showed a non-zero frequency in any breed other than the one in which it was discovered. Mixed-breed dogs were a different story: 114 of the single-copy carriers of the Labrador variant were mixed breeds, out of roughly 810,000 in the dataset.

Wisdom Panel’s own condition page gives a second figure: the variant was “found in 1 in 5,000 dogs” in the company’s testing. That is a company statement about its whole customer base, not a Labrador-specific figure, and the page does not define it precisely. It should be read on its own terms rather than set beside the published allele frequency as if the two measured the same thing.

The third number is the one most relevant to breeding programmes. OFA’s Labrador testing statistics show 269 dogs with a congenital myasthenic syndrome result on file: 267 normal and 2 carriers. For comparison, the same table lists 325,416 hip evaluations. The carriers in that small group are a reminder that the variant exists in registered lines, but 269 voluntary submissions cannot be turned into a breed-wide carrier rate.

Two other groups deserve a sentence. AnimaLabs lists the Labradoodle alongside the Labrador for this test, and the DogWellNet harmonisation database rates both the Australian Labradoodle and the Labrador Retriever as breeds with some evidence for the condition. Anyone breeding Labradoodles from Labrador lines is drawing from the same gene pool and can test for the same variant.

Antibody negative, decrement positive: separating it from acquired myasthenia gravis

The disease most often confused with CMS is acquired myasthenia gravis (MG), an autoimmune disorder in which antibodies attack the acetylcholine receptor. In a 2016 mini-review in Neuromuscular Disorders, G. Diane Shelton of UC San Diego summarised the distinction: MG is acquired and immune-mediated, whereas CMS comes from inherited structural or functional faults at the junction. A Labrador puppy with a COLQ variant does not have an immune disease, and the two should not be described as the same thing.

The test that separates them is the AChR antibody titre. Shelton’s Comparative Neuromuscular Laboratory at UC San Diego, founded in 1990 and run as an international reference centre for companion-animal neuromuscular disease, lists the test on its 2026 submission form (revised 01/01/2026) at $161.00. It needs 1 to 2 ml of serum shipped with a cold pack by overnight or second-day courier, and samples must be analysed within 14 days of collection. The form states that a positive titre is diagnostic of acquired myasthenia gravis in dogs and cats. In a young Labrador with exercise-related weakness, a negative titre therefore pushes the search towards an inherited cause. The same form prices a complete muscle profile, for a first muscle biopsy, at $182.00, and asks clinicians to call before collecting biopsy specimens.

SamIf the antibody test is negative, does the DNA test then settle it?

Elena MarshIt confirms or excludes one variant. Shelton’s 2016 review lists CHRNE, COLQ and CHAT as canine CMS genes, and a 2017 mixed-breed case turned out to be a receptor-deficient form that no breed test would have caught.

Other forms of CMS belong on the list a neurologist works through. The first canine CMS was recognised in Jack Russell Terriers in 1974, and a 2015 study traced it to a CHRNE frameshift that cuts receptor protein by about 90%. Old Danish Pointing Dogs carry a CHAT variant, and in 2026 Dobermanns were linked to an AGRN variant. In a 2017 Frontiers in Veterinary Science report, a six-month-old mixed-breed dog with progressive weakness, a decrement on repetitive stimulation and a normal AChR antibody result turned out to lack receptor protein on intercostal muscle biopsy; it was a postsynaptic syndrome, not COLQ. Exercise-induced collapse, centronuclear myopathy and other Labrador neuromuscular disorders also sit on the differential list.

One safety point overrides the diagnostic sequence. Breathing difficulty, choking or bringing food back up shortly after eating need an urgent veterinary visit, not a wait for test results. In the Golden Retriever COLQ family, a puppy with stridor and a history of regurgitation developed laryngeal paralysis at 11 months and died of aspiration pneumonia two days after airway surgery, and in affected cats megaesophagus and aspiration are recurring causes of death. These signs were not described in the two Labrador sisters, but they are part of how myasthenic disorders turn dangerous.

Kits that name I337T, and what each one costs

Because the variant is a single substitution, almost any genotyping laboratory can read it. The useful check is that the listing names the Labrador variant, c.1010T>C or p.I337T, rather than the Golden Retriever change c.880G>A.

• UC Davis Veterinary Genetics Laboratory offered “Congenital Myasthenic Syndrome (CMS) in Labrador Retrievers” at $55 as a single test, with a $5 discount on three or more dogs, or $25 as an additional health test. Turnaround was at least 15 business days, from a cheek and gum brush sample. These figures come from an archived copy of the page dated July 2025; the live page could not be read, so prices may have changed.
• Genomia (Czech Republic) lists “CMS in Labrador Retrievers” at USD 56.00 per test, shown without VAT, with a usual turnaround of 12 business days.
• AnimaLabs (Croatia) lists the Labrador Retriever type at EUR 54.90 including VAT, from a cheek swab or EDTA whole blood.
• Wisdom Panel names the Labrador variant on its health condition page and cites the 2014 paper; the price of the panel that includes it was not captured for this article.
• Embark carries an item called “Congenital Myasthenic Syndrome, CMS (COLQ, Labrador Retriever Variant)” in its health list, but its individual page could not be retrieved, and Embark is not among the providers DogWellNet lists for this test.
• Paw Print Genetics had a listing for the Labrador type, but the page now redirects to a notice about its merger with Orivet; no current price was available.

DogWellNet’s provider table for COLQ-related CMS names EVG Diagnostics, Laboklin, Paw Print Genetics, VetGen, Weatherbys Scientific, Wisdom Panel and AnimaLabs. UK owners will find Weatherbys and Laboklin among those names, although no sterling price for a UK laboratory was captured here. For a single dog in the US, the archived UC Davis figure suggests that adding COLQ to an existing order costs less than the antibody blood test a neurologist might run, which is one reason testing breeding stock is cheaper than working up a sick puppy.

Planning a litter: what carrier status changes and what it does not

For breeders, the recessive pattern makes the arithmetic simple. Two clear parents cannot produce an affected puppy for this variant. A carrier bred to a clear dog produces no affected puppies, although some offspring will be carriers. Only a pairing of two carriers can produce affected puppies, which is how the 2014 litter came about. Testing the dam and sire before mating therefore prevents affected litters without forcing a carrier out of a programme.

SamShould every Labrador breeder be adding this test now?

Elena MarshThat is a judgement for breed clubs, not a requirement anyone has published. Donner’s 2023 data found 187 single-copy carriers among about a million dogs and no affected ones, so it is rare, but cheap to check in lines that use linebreeding.

The 2014 authors linked their family’s problem to linebreeding, and the Golden Retriever COLQ family reported in 2020 showed the same picture: affected pups shared a sire who was also their maternal grandfather. Close pairings bring rare recessive variants together, so the case for testing is strongest when the same ancestors appear on both sides of a planned pedigree.

What testing cannot do is certify a puppy as free of weakness. A clear result covers one position in one gene. It says nothing about other CMS genes, acquired MG later in life, EIC or the long list of orthopaedic and muscular causes of a reluctant, tiring young dog.

A one-year window in California, and a policy clause that excludes the test itself

US consumer protection for puppy buyers varies by state, and most windows for hereditary conditions are short. California is an exception for pets bought from a dealer. Under California Health and Safety Code section 122160, if within one year after the buyer takes possession a California-licensed veterinarian states in writing that the dog has a congenital or hereditary condition that adversely affects its health, the dog is considered unfit for sale. The buyer can then choose a refund, an exchange, or keep the dog and claim reasonable veterinary fees for diagnosis and treatment up to 150 percent of the purchase price, plus sales tax. Sales by breeders fall under a separate California statute that was not reviewed for this article, so buyers from private breeders should check that law or their contract.

Pet insurance matters because the signs of this disease can appear within weeks of a puppy going home. Embrace’s FAQ states that breed-specific, genetic and congenital conditions are covered by its accident and illness policy as long as the condition is not pre-existing to the policy and its waiting periods. The same FAQ lists “DNA testing or cloning” among its exclusions, so the genetic test itself is an out-of-pocket cost even where the disease might be covered. Embrace’s genetic-conditions page adds that some insurers charge extra for a rider covering genetic conditions or restrict or exclude them entirely.

The gaps a buyer should know about before reading too much into any result

Much of what owners would like to know about this disease in Labradors has simply not been measured. No further affected Labrador has been reported in the literature since 2014, so the clinical picture rests on two puppies. Both were euthanised young, so there is no natural history for the Labrador form. OMIA’s summary of the canine COLQ syndrome, drawing on the 2020 classification review, notes that signs often progress despite treatment and that most reported affected puppies died or were euthanised. No comparative treatment study exists for dogs.

There is also no published carrier rate for UK or European Labradors; the million-dog dataset is drawn largely from US customers. Whether OFA or any breed club plans to add CMS to its recommendations has not been announced. And the same COLQ gene causes a separate, well-documented disease in Devon Rex and Sphynx cats through a different change, c.1190G>A. Figures from those cats, and from the Golden Retriever family, should not be borrowed to describe Labradors.

Read against those gaps, the most useful message is modest. COLQ-related CMS is a real, rare cause of exercise-linked weakness in Labrador puppies that standard breed panels do not cover. An inexpensive DNA test can identify carriers before mating, and a weak puppy deserves a full veterinary neurological workup rather than an assumption that it must be EIC.

Frequently asked questions

Q. Is congenital myasthenic syndrome the same as myasthenia gravis?
No. Myasthenia gravis in dogs is usually an acquired autoimmune disease in which antibodies target the acetylcholine receptor. COLQ-related CMS is inherited and present from birth. The AChR antibody blood test helps tell them apart: a positive titre indicates acquired MG, whereas the 2014 Labrador puppies were antibody-negative.

Q. My Labrador’s parents are EIC clear. Does that cover CMS?
No. EIC and COLQ-related CMS are caused by different genes. Both parents of the affected 2014 litter had clear EIC results. Only a test that reads COLQ c.1010T>C tells you about this variant.

Q. Can a carrier Labrador develop weakness from this variant?
Not from this variant alone. The condition is autosomal recessive, and the 16 one-copy relatives in the 2014 family were not reported as affected. A carrier can still develop weakness for unrelated reasons, so symptoms should always be examined by a veterinarian.

Q. My puppy tires quickly and lies down. What should I do first?
Book a veterinary examination and describe exactly when the weakness starts and how long rest takes to bring recovery; a short video helps. Diagnosis needs a veterinarian, often with referral to a neurologist for electrodiagnostics and antibody testing. Choking, coughing after meals or laboured breathing need same-day care.

Q. Is the Golden Retriever test the same as the Labrador test?
No. The Golden Retriever form, reported in 2020, is caused by a different COLQ change, c.880G>A. In the million-dog dataset neither variant was found in the other breed, so order the listing that names the Labrador variant.

Q. Can affected dogs be cured?
There is no cure reported, and no controlled treatment study exists for dogs. Most published affected puppies died or were euthanised young. Any management plan for a dog with confirmed CMS has to come from the treating veterinarian.

References

• Rinz CJ, Levine J, Minor KM, Humphries HD, Lara R, Starr-Moss AN, Guo LT, Williams DC, Shelton GD, Clark LA (2014). A COLQ missense mutation in Labrador Retrievers having congenital myasthenic syndrome. PLoS One 9(8):e106425.
• Tsai KL, Vernau KM, Winger K, Zwueste DM, Sturges BK, Knipe M, Williams DC, Anderson KJ, Evans JM, Guo LT, Clark LA, Shelton GD (2020). Congenital myasthenic syndrome in Golden Retrievers is associated with a novel COLQ mutation. J Vet Intern Med 34(1):258-265.
• Donner J, Freyer J, Davison S, Anderson H, Blades M, Honkanen L, Inman L, Brookhart-Knox CA, Louviere A, Forman OP, Chodroff Foran R (2023). Genetic prevalence and clinical relevance of canine Mendelian disease variants in over one million dogs. PLoS Genet 19(2):e1010651.
• Mignan T, Targett M, Lowrie M (2020). Classification of myasthenia gravis and congenital myasthenic syndromes in dogs and cats. J Vet Intern Med 34(5):1707-1717.
• Shelton GD (2016). Myasthenia gravis and congenital myasthenic syndromes in dogs and cats: A history and mini-review. Neuromuscul Disord 26(6):331-334.
• Ohno K, Brengman J, Tsujino A, Engel AG (1998). Human endplate acetylcholinesterase deficiency caused by mutations in the collagen-like tail subunit (ColQ) of the asymmetric enzyme. Proc Natl Acad Sci U S A 95(16):9654-9659.
• Rinz CJ, Lennon VA, James F, Thoreson JB, Tsai KL, Starr-Moss AN, Humphries HD, Guo LT, Palmer AC, Clark LA, Shelton GD (2015). A CHRNE frameshift mutation causes congenital myasthenic syndrome in young Jack Russell Terriers. Neuromuscul Disord 25(12):921-927.
• Proschowsky HF, Flagstad A, Cirera S, Joergensen CB, Fredholm M (2007). Identification of a mutation in the CHAT gene of Old Danish Pointing Dogs affected with congenital myasthenic syndrome. J Hered 98(5):539-543.
• Shelton GD, Coates JR, Steiss JE, Guo LT, Platt SR, Minor KM, Friedenberg SG, Cullen JN, Bullock G, Hansen EA, Katz ML, Johnson GS (2026). Congenital Myasthenic Syndrome in Doberman Pinscher Dogs Is Associated with a Homozygous Missense Variant in AGRN. Biomolecules 16(8):1099.
• Blakey TJ, Michaels JR, Guo LT, Hodshon AJ, Shelton GD (2017). Congenital Myasthenic Syndrome in a Mixed Breed Dog. Front Vet Sci 4:173.
• OMIA 001928-9615: Myasthenic syndrome, congenital, COLQ-related in Canis lupus familiaris (dog). Online Mendelian Inheritance in Animals, University of Sydney.
• Wisdom Panel. Congenital Myasthenic Syndrome (Discovered in the Labrador Retriever). Health condition page, accessed 2026-10-17.
• Comparative Neuromuscular Laboratory, University of California San Diego. Service and submission form, revised 01/01/2026.
• DogWellNet / International Partnership for Dogs. Harmonization of Genetic Testing for Dogs: Congenital Myasthenic Syndrome – COLQ related, accessed 2026-10-17.
• Embrace Pet Insurance. Coverage FAQ, accessed 2026-10-17.

About the author

Elena Marsh

Editor & writer (not a veterinarian)

A writer with a molecular-biology background and a lifelong dog and cat owner. Not a veterinarian — she translates peer-reviewed genetics research and primary data into plain language, always as information rather than diagnosis.