Академический Документы
Профессиональный Документы
Культура Документы
Institute of Animal Genetics, Nutrition and Housing, àDepartment for Clinical Veterinary Medicine,
Section of Neurology, University of Berne, Bremgartenstrasse 109a, CH-3012 Berne, Switzerland,
§
Heinrich-Heine-University of D€ usseldorf, Institute for Neuropathology, Moorenstrasse 5, D-40225 D€usseldorf,
Germany, Veterinary Neurological Consulting Services, 1476 LakeviewRidge, Dadeville, AL 36853, USA,
àà
Department of Clinical Sciences, School of Veterinary Medicine, TuftsUniversity, 200 Westboro Road,
North Grafton, MA 01536, USA, §§University of Z€urich, Tierspital, Neurology Section,Winterthurerstrasse
268, CH-8057 Z€ urich, Switzerland, Private praxis, Pfeffingerstrasse 41, CH-4053 Basel, Switzerland,
ààà
Private praxis, Birkenstrasse 35, CH-4055 Basel, Switzerland
SUMMARY
Labrador Retriever myopathy (LRM) has become a relatively common muscular disease. The objective of our pro-
spective study was to determine by segregation analyses a plausible mode of inheritance within a Labrador Retriever
population. Therefore we performed neurological examinations, as well as electromyographic and histopathological
evaluations of 58 closely related dogs. Seven dogs with an average age of 27.8 months had clinical signs consistent with
LRM including exercise intolerance or fatigue. The diagnosis was based on neurological deficits and confirmed by
histopathological results of muscle biopsy. We found in all cases obvious differences in fiber calibre size associated with
texture disturbances. In addition, we found 41 clinically normal dogs with histological findings consistent with LRM.
Three genetic models, the major gene, the mixed inheritance as well as the environmental model, were evaluated by
segregation analyses. They were applied to an extended pedigree including 164 non-randomly ascertained related
Labradors. According to phenotype the clinically examined dogs were divided into two different data sets. One data set
distinguished between clinically normal and abnormal dogs, the second data set between histopathologically normal and
abnormal dogs. We concluded that the clinical form of LRM is transmitted by a major gene and controlled by an
autosomal recessive mode of inheritance. Furthermore, for expression of the subclinical form an additional gene or an
environmental factor is responsible. Our findings suggest that LRM is similar to limb-girdle muscular dystrophy in man
and therefore, may be used in the future as an animal model. Ó 2002 Elsevier Science Ltd. All rights reserved.
LABRADOR Retriever myopathy (LRM) is a re- et al 1996). Severely affected Labradors may have a
latively common muscular disease that mostly affects megaoesophagus that is confirmed radiographically
several dogs from the same litter (McKerrell and (Watson et al 1988). In most affected dogs electromy-
Braund 1987). The disorder was first published in 1976 ography (EMG) reveals positive sharp waves, fibrilla-
by Kramer and co-authors and is also known as ‘‘he- tion potentials and bizarre high frequencies, however,
reditary myopathy’’, ‘‘inherited myopathy’’ or ‘‘mus- motor nerve conduction velocities (MNCV) are within
cular dystrophy’’ worldwide (McKerrell and Braund normal limits (Moore et al 1987).
1986; Gortel et al 1996; Klopp and Smith 2000). The histopathological evaluation of muscle biopsies
The clinical signs are variable, but usually charac- confirms the diagnosis of LRM. The characteristic
terised by generalised intermittent weakness and findings are the reduction in number, necrosis and fiber
ataxia. Affected dogs between 6 weeks and 7 months of group atrophy of type 2 fibers. Other morphological
age have a short stilted gait, and sometimes demon- changes are fiber regeneration, proliferation of con-
strate ‘‘bunny hopping’’ and in rare cases ventroflexion nective tissue, internal nuclei and sometimes phagocy-
of the neck accompanied by a kyphotic posture. Pos- tosis. In some specimens muscle fibers are
tural reactions are normal or reduced and tendon re- hypertrophic. Peripheral nerves do not show histo-
flexes may be absent (Amann 1987). The musculature pathological changes (McKerrell and Braund 1986).
shows generalised atrophy but is not painful on pal- The pathogenesis of LRM is still unclear and limited
pation. In addition, serum levels of creatine kinase studies of hemodynamic and metabolic aspects of af-
(CK) activity may be moderately increased (Gortel fected muscles have been inconclusive (Amann et al
1988; Mehta et al 1989; Olby et al 2001).
To date, investigations of the genetic aspects of
*
Corresponding author: A. Jaggy, Birkenstrasse 35, CH-4055 Basel,
LRM have included only small groups of sporadically
Switzerland. affected dogs. It has been postulated that the disease is
E-mail: tim_bley@hotmail.com transmitted by a simple autosomal recessive mode of
0034-5288/02/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved.
232 T. Bley et al
inheritance (Kramer et al 1981; McKerrell and Braund with an average age of 39.9 months (range: 4–96
1987). Since Labrador Retriever myopathy affects both months).
sexes, the disease is well distinguishable from the X- Physical and neurological examinations were carried
linked muscular dystrophies of the Groenendaeler out on all Labradors. In addition, serum CK levels were
Shepherd, the Samoyed or the Golden Retriever (Van determined in 19 dogs. For further diagnostic proce-
Ham et al 1993; Valentine et al 1986). dures, probands were anesthetised with 3–7 mg/kg
We had the opportunity to perform a neurological bodyweight PropofolÒ given intravenously, and there-
examination, as well as electromyographic and histo- after incubated and maintained with halothane.
pathological evaluation on 58 closely related Labrador Thirty-two Labradors were tested by EMG as described
Retrievers. The purpose of our prospective study was by Moore et al (1987). Briefly, we used a monopolar
to characterise the disease within that population and exploring electrode and a ground electrode (Medelec
to establish a plausible mode of inheritance by segre- Accessories, Oxford Instruments, Hawthorne, USA).
gation analyses. The following muscles were examined for the thoracic
limbs: m. extensor carpi ulnaris, m. flexor digitalis su-
perficialis, m. biceps and triceps brachii; for the pelvic
MATERIALS AND METHODS limbs m. gastrocnemius, m. biceps femoris, m. flexor
digitalis lateralis and m. semitendinosus. Additionally,
Dogs and clinical evaluations
we investigated paravertebral and the muscles of mas-
The first case presented at the Clinic for Small An- tication (m. masseter and m. temporalis). MNCV of
imals in Berne, Switzerland, was a 4-month-old male either the left or the right peroneal nerve (n ¼ 6) was
Labrador Retriever (dog #843, Fig 1). Thereafter, his determined using Teflon-coated electrodes in combi-
whole nuclear family (8 littermates and parents), two nation with a ground electrode (Moore et al 1987). The
fullsib families (n ¼ 16) and 31 further related dogs recorded potentials were displayed on the Multiliner E
were evaluated. Thus 58 Labradors were included in Neuroscreen PLUS (Jaeger/Toennies) and stored on
our investigation, of which 33 were male and 25 female computer for further evaluations.
FIG 1: Extended pedigree including 164 dogs. / , female/male, not examinated; / , female/male; / , female/male, biopsy abnormal; / , female/male,
clinic and biopsy abnormal. Dog #843 was the first case of histologically confirmed myopathy. Every dog is marked by an identification number.
Genetic of Labrador Retriever myopathy 233
Muscle biopsies were collected from all 58 Labrador for allele frequency p and the allele transmission
Retrievers as described by Amann (1998). Further- probabilities, was calculated under two different as-
more, muscle samples from 12 healthy dogs of various sumptions: the latter mentioned were set according
breeds, ages and sexes were collected. to Mendelian segregation, namely 1, 1/2 and 0 re-
The selected muscle samples [m. semitendinosus (n spectively, and the values of the general transmis-
¼ 38), m. tibialis cranialis (n ¼ 15), m. quadriceps sion probability could vary between 0 and 1.
femoris (n ¼ 14), m. biceps femoris (n ¼ 2), m. triceps
2. The mixed inheritance model: The major locus was
brachii (n ¼ 1) and m. rectus abdominis (n ¼ 2)] were
frozen in isopentane cooled in liquid nitrogen modeled as in the major gene model. Furthermore,
()135 °C) and cut in transverse planes at 5 lm with the the underlying liability was assumed to be addition-
cryostat microtome maintained at )20 °C (Braund et al ally controlled by a normally distributed additive
1978). Sections were stained according to the following polygenic component with mean 0 and standard de-
histological and histochemical techniques: hematoxylin viation ra . This model was described by the same pa-
and eosin (HE), GomoriÕs modified trichrome/Engel, rameters as in the major gene model and by the
reduced nicotine-amide adenine dinucleotide–tetrazo- polygenic heritability, where h2 ¼ r2a =r2a þ ð1 r2a Þ.
lium reductase (NADH-TR), routine adenosine tri- 3. The environmental non-transmission model: Neither
phosphatase at pH 9.4 (ATPase), oilred/Sudan, a major gene effect nor polygenic effects were as-
succinate dehydrogenase (SDH) and periodic acid– sumed. It modeled the situation as if no genetic com-
Schiff. Muscle biopsy specimens revealed signs of se-
ponents had any influence on the disease.
vere muscle fiber degeneration, as evidenced by a
spectrum of muscle fiber diameters caused by fiber at- FisherÕs exact t-test was used to evaluate frequencies of
rophy and compensatory hypertrophy, fiber disruption morphological changes in muscle samples between
and degradation, metachromasia of damaged muscle males and females, as well as between clinically af-
fibers in the Gomori modified trichrome/Engel reaction fected and normal dogs (SAS 1996). The significance
and loss of type 2 fibers as demonstrated by the
level was set at P 6 0.05.
NADH-TR reaction. Nerve biopsies were performed
as described by Braund et al (1982). Briefly, nerve tis-
sue was stretched on wooden tongue depressors using RESULTS
pins, fixed in 2.5% glutaraldehyde and processed for
Dogs
paraffin fixation.
Myopathy was clinically diagnosed and historically
confirmed in 7 Labrador Retrievers (#572, 843, 844, 1025,
Pedigree and segregation analyses 1026, 1027, and 1028) that were presented for exercise
A total of 164 Labrador Retrievers with pedigree intolerance or fatigue (Fig 1). Their age ranged between
certificates were included in the study, of which 58 dogs 4 and 69 months (average: 27.8 months). The neurolog-
were clinically evaluated as previously described. The ical examination revealed poorly muscled dogs (n ¼ 6)
latter population was analysed with two different: data with decreased postural reactions (n ¼ 7) including ab-
sets as follows: data set 1 included clinically and his- normal wheelbarrowing (n ¼ 7) and hopping (n ¼ 7).
topathologically normal Labradors (n ¼ 10); in addi- All spinal reflexes were decreased (n ¼ 3) to absent
tion, 48 dogs with histopathological findings were (n ¼ 4). All results of blood (n ¼ 19) analysis including
classified as abnormal regardless of whether the clinical serum C K level determination were normal with the ex-
results were normal (n ¼ 41) or abnormal (n ¼ 7). ception of dog #843 (C K ¼ 730 U/l). Electrodiagnostic
Data set 2 consisted of clinically healthy dogs regard- examination was done in 32 dogs. Fibrillation potentials
less of the histological results; they were classified as and positive sharp waves in all proximal limb and par-
being normal (n ¼ 51). However, dogs with concomi- avertebral muscles were found in four dogs, all belonging
tant clinical signs of LRM and pathological findings to the clinically abnormal group. The results of MNCV
were classified as being abnormal (n ¼ 7). (n ¼ 11) were within normal limits (n ¼ 7).
Three different genetic models, that estimate allele The histopathological changes were not uniform in all
frequencies and transmission probabilities were tested affected dogs. They consisted of marked differences in
by segregation analyses. Calculations were performed fiber calibre size (P < 0:001) associated with texture
by the maximum likelihood procedures (Elston and disturbances (Fig 2). In most cases type 2 muscle fibers
Steward 1971; Morton and McLean 1974; Lalouel et al were dystrophic (P < 0:001) (Fig 3). The latter men-
1983) using the statistical software PAP package tioned condition (P < 0:021) as well as hypertrophy
(Hasstedt 1994) and maximised with the NSPOL pro- (P < 0:05) was observed in type 1 fibers (Figs 3, 4). Ne-
gram (Gill et al 1986). The differences between the crosis (P < 0:001) and atrophy (P < 0:05) of fiber type 1
likelihoods of the following three models were com- were also significantly frequent. The mean diameter of
pared with likelihood ratio tests. the smallest fibers was less than 5 lm and of the largest
fibers more than 150 lm. Replacement of muscle fibers
1. The major gene model: The major locus was mod- with fat and fibrous tissue was obvious (P < 0:001).
eled having two alleles in the Hardy–Weinberg equi- Mononuclear infiltrates, regeneration, hypertrophy of
librium, i.e. three different genotypes could be type 2 fibers, pathological enzyme activity, giant cells
distinguished. This model, which provided estimates and phagocytosis were uncommon findings (Fig 2).
234 T. Bley et al
TABLE 1: Genetic parameters of segregation analyses for myopathy in Labrador Retrievers expressed through different
genetic models
Genetic parameters Major gene with Mendelian Major gene Mixed inheritance
segregation
Data set Data set Data set
1 2 1 2 1 2
Frequency of the normal allele (A) 0.17 0.69 0.29 0.74 0.29 0.74
(0.07) (0.12) (0.15) (6.16) (0.15) (0.16)
Transmission probability of AA 1a 1a 0.99 1.00 0.99 1.00
allele A in genotype (0.51) (0.00) (0.15) (0.00)
Aa 0.5a 0.5a 0.11 0.44 0.11 0.44
(0.10) (0.16) (0.10) (0.16)
aa 0a 0a 0.05 0.00 0.05 0.00
(0.05) (0.00) (0.05) (0.00)
Polygenic heritability – – – – 0.02 0.02
(0.00) (0.00)
)2 In likelihood 68.18 32.28 50.24 32.13 50.24 32.13
Data are mean (standard error); bold: estimated parameters.
The environment model has a calculated )2 In likelihood of 52.15 for data set 1 and 37.67 for data set 2.
a
Present value; –: no value calculated.
tolerance and fatigue. All dogs exhibited signs of lower segregation analyses of data set 2, which gives a phe-
motor neuron dysfunction. It is well established that notypical distinction between clinically normal and
LRM is only confirmed by histological evaluation abnormal animals, confirmed the previous postulated
(McKerrell and Braund 1986). Therefore, we per- mode of inheritance. We therefore concluded that a
formed a standardised muscle biopsy on every single simple recessive Mendelian inheritance did not fit the
case. In addition, our controls included 10 related data, although the presence of a major gene with a very
Labrador Retrievers, as well as 12 dogs of different high frequency of deleterious allele (0.71) was most
breeds. Their histological results were consistently likely. Since its transmission probability was very high
negative. Furthermore, we examined closely related (0.89 instead of 0.5), we showed that this allele does,
nuclear families, as well as unrelated single Labradors. however, not segregate according to Mendelian ex-
To our surprise we found a large number of clinically pectation We do not believe that the segregation dis-
normal dogs (n ¼ 41) with histological findings consis- tortion is caused by a preferential selection of gametes
tent with myopathy. However, hypertrophy of type 2 that include the deleterious allele. However, the results
and type 1 fibers was only observed in clinically ab- of the segregation could also be influenced by false-
normal dogs. One explanation could be that fiber hy- negative histopathological results. Indeed the clinically
pertrophy may reflect an attempt to compensate healthy and histopathologically normal parents #47 and
malfunction through loss of morphological structures 824 (Fig 1) had one litter of six offspring (#741–746)
and indicate an advanced state of myopathy. We think with histopathologically confirmed LRM. We con-
that these observations may represent a subclinical cluded from all these results that the clinical form of
form of LRM. Because our probands are used as guide LRM is localised on a single major gene and trans-
dogs and therefore are not stressed by extensive mitted by an autosomal recessive mode of inheritance,
physical exercise, we believe that the latent form may whereas the subclinical form may be expressed by an
become obvious only under strong physical work. additional gene. We cannot exclude, however, that the
Abnormal EMG findings were only obvious in dogs latter may be influenced by some environmental
with clinical symptoms. All other dogs had no abnor- factors.
malities. Thus, this diagnostic procedure was not sen- In regard to these bivalent results, we postulate that
sitive enough to detect subclinical stages of the disease. several forms of a hereditary myopathy may be present
Similar results have been documented by Hoskins and in our dog population. It is interesting to note that the
Root (1983), who stated that histologically confirmed limb-girdle muscular dystrophy (LGMD) of man re-
LRM is not always accompanied by concurrent EMG presents a heterogeneous group of diseases that is
abnormalities such as fibrillation potentials and posi- characterised by mild to severe forms of progressive
tive sharp waves. muscle degeneration (Moreira et al 1997). In addition,
We were able to show the genetic relationship the clinical and histopathological findings are similar to
among the clinically evaluated Labradors by pedigree the described findings in our dogs (Fardeau et al 1996).
analyses and prove, thereafter, that the disease has Hence, LRM could be an acceptable animal model for
indeed a hereditary basis. In addition, the results of further research purposes.
data set 1, which phenotypically distinguished between Additional genetic investigations will have to study
histopathologically normal and abnormal animals, were an extensive group of phenotypically characterised
consistent with an autosomal recessive inheritance as Labrador Retrievers. They could lead to further ge-
already suggested by other authors (Kramer et al 1976; netic results and finally to the development of genetic
Gortel et al 1996). However, in our present families the markers. The collaboration of breeders for routine
frequency of the deleterious allele was extremely high, muscle screening and appropriate strategies are needed
with a range from 0.31 to 0.26. Moreover, the results of to selectively control the disease.
236 T. Bley et al