Brain (1996), 119, 295-308
Juvenile limb-girdle muscular dystrophy
Clinical, histopathological and genetic data from a small
community living in the Reunion Island
M. Fardeau,1 D. Hillaire,4 C. Mignard,3 N. Feingold,2 J. Feingold,2 D. Mignard,3 B. de Ubeda,3
H. Collin,1 F. M. S. Tome,1 I. Richard4 and J. Beckmann4
]INSERM U153, the 2INSERM U155, Universite Paris VII,
Paris, the ^Centre Hospitaller General, He de la RSunion
and the AGenethon Laboratory, Evry, France
Correspondence to: M. Fardeau, INSERM U153, 17 Rue
du Fer a Moulin, 75005 Paris, France
A series of patients affected by a muscular dystrophy, similar
to the original description of a juvenile scapulo-humeral
form by Erb in 1884 and fitting with the criteria used to
define limb-girdle muscular dystrophies, was discovered in a
small community living in the southern part of Reunion
Island in the Indian Ocean. A detailed clinical analysis was
conducted over 5 years on a cohort of 20 patients. This
community presented a high degree of consanguinity as it
was segregated from the majority of the island population
for more than a century. In previous molecular genetic
studies, the disease locus has been mapped to chromosome
15p. Mutations were recently identified in a gene located in
this region encoding for muscle-specific calcium activated
neutral protease (CANP3). Clinical, pathological, genetic
and complete identification of the mutations are presented
here, establishing, for the first time, precise clinico-genetic
correlations in this form of autosomal recessive, juvenile,
limb-girdle muscular dystrophy (LGMD).
Keywords: limb-girdle muscular dystrophy; calpain; epidemiology; muscle biopsy; molecular genetics
Abbreviations: CANP3 = muscle-specific calcium activated neutral protease; LGMD = limb-girdle muscular dystrophy;
SCARMD = severe childhood autosomal recessive muscular dystrophies
The concept of LGMD was, amongst the hereditary muscle
disorders, one of the most challenged and one of the most
difficult to establish as an entity. The early description by
Erb (1884) of a juvenile, scapulo-humeral type of progressive
muscular atrophy raised the first controversy; at about the
same time Landouzy and Dejerine (1885) described a 'pure
myopathy' with facio-scapulo-humeral involvement, and
emphasized in a following paper (Landouzy and Dejerine,
1886) that the facial muscle involvement in some cases
could be demonstrated only at the histological level. Erb's
proposition (1891) to include with his own cases the previous
observations of Leyden (1875) and Mobius (1879) was
generally accepted. Batten (1909) acknowledged a 'juvenile'
form of myopathy conforming to Erb's description, but
the nosological situation of this muscular dystrophy varied
considerably in the successive classifications. Bell was the
first clearly to differentiate this dystrophy (Bell, 1943), which
she considered as inherited as an autosomal recessive disease,
© Oxford University Press 1996
from the X-linked pseudo-hypertrophic recessive Duchenne's
muscular dystrophy and the autosomal dominant facio-
distinction was not accepted by others (e.g. Tyler and
Wintrobe, 1950; Levison, 1951; Stevenson, 1953, 1955). In
their classic paper, Walton and Nattrass (1954) adopted a
similar 'ternary' presentation of the major muscular dystro-
phies, and coined the term limb-girdle muscular dystrophy
to include cases of both sexes, beginning usually within the
first three decades, with a predominant involvement of
scapular, pelvic girdle and trunk muscles, without involve-
ment of facial muscles, with uncommon pseudohypertrophy,
moderately severe progression and usually an autosomal
The clinico-pathological homogeneity of this proposed
entity was, however, again rapidly disputed. With progress
in physiological and histopathological evaluation of muscle
disorders, it soon appeared that a number of patients
dystrophy. However, this
by guest on May 26, 2012
M. Fardeau et aL
considered to be suffering from LGMD were, in fact, affected
by other conditions such as spinal muscular atrophies,
congenital myopathies or metabolic disorders. Although some
familial cases with limb-girdle distribution of muscle
weakness and autosomal recessive inheritance in some inbred
communities continued to be identified as LGMD (Pfandler,
1950; Kloepfer and Talley, 1957; Jackson and Carey, 1961;
Moser et aL, 1966; Jackson and Strehler, 1968; Shokeir and
Kobrinsky, 1976; Shokeir and Rozdilsky, 1985), the relative
rarity of this dystrophy in most countries and the diagnostic
difficulties led some authors to prefer the term of 'limb-
girdle syndromes' (Munsat, 1977; Bradley, 1979) and others
to question even the reality of LGMD (Brooke, 1977).
The discovery of a group of affected families in the small
Reunion Island in the Indian Ocean (Fardeau et aL, 1989)
was therefore of great interest. The clinical presentation and
progression of the severity of these cases were very close to
the original description by Erb (1884) and fitted with the
clinical, histological and genetic criteria most commonly
adopted for LGMD. Immunocytochemical study of muscle
biopsies showed a precise distinction between LGMD
and other (more recently identified) conditions such as
Becker's dystrophy characterized by dystrophin abnormality
(Hoffman et aL, 1989; Bushby and Gardner-Medwin, 1993)
and the severe childhood autosomal recessive muscular
dystrophies (SCARMD) characterized by adhalin deficiencies
(Matsumura et aL, 1992; Fardeau et aL, 1993; Romero
et aL, 1994).
The families were concentrated in a small mountainous
area of the southern part of Reunion Island. Due to socio-
economic problems and geographical conditions, they had
remained isolated from the majority of the island population
for more than a century. This suggested an inbred community,
as was subsequently confirmed by the genealogical study of
the different branches of the pedigrees. Thus, a molecular
genetic study could be performed in a priori favourable
homogeneous conditions. This led to gene mapping of the
disease locus to the long arm of chromosome 15 (Beckmann
et aL, 1991). After genotyping additional markers on this
chromosome, the locus was restricted to a 7 cM interval
(Fougerousse et aL, 1994) and eventually 1 cM (Allamand
et aL, 1995). In this region, a gene encoding the large subunit
of the muscle-specific calcium-activated neutral protease 3
(calpain 3) was located (Chiannilchulchai et aL, 1995), and
mutations were found which cosegregate with the disease in
families of various origins (Richard et aL, 1995).
The present clinical, histological, immunocytochemical,
ultrastructural and genetic data of the Reunion Island families
are reported here.
Hospital in 1987. This led to the examination of a cohort of
20 patients belonging to eight nuclear families, totalling 87
individuals. Eight additional patients, four belonging to the
latter families and four to different nuclear families, were
found between 1988 and 1994. They were confirmed by
clinical, biological and genetic findings but were not included
in the longitudinal clinical study of the first 20 patients.
Each of the 20 patients was personally examined by one
of us (M.F.) on three different occasions (1987, 1990, 1993).
Manual testing of muscle strength (according to international
procedures; Medical Research Council, 1943) was performed
in most patients in 1988 and 1992 by the same trained
physiotherapist. The functional stage of the dystrophy was
graded from I to VII according to a scale proposed by
Gardner-Medwin and Walton (1974). Photographs of some
patients were taken with their consent. CT scans of lower
limb muscles were performed on 12 patients. Cardiac (15
patients) and respiratory (19 patients) explorations were
carried out initially and after 3 years. A psychometric
evaluation was performed on 14 patients according to the
Wechsler scale. Serum creatine kinase levels were measured
in all patients. Electromyographic examinations and motor
conduction velocities measurements were performed on 14
Histological and immunocytochemical studies
Five patients at different stages of the disease agreed to be
biopsied. A muscle sample was taken from the median head
of the deltoid muscle according to a 'mini-open' procedure.
The muscle fragments were deep-frozen in isopentane cooled
in liquid nitrogen in Reunion Island and shipped to Paris to
be processed following classical histoenzymological and
ultrastructural methods (Dubowitz and Brooke, 1973).
Immunocytochemical studies were carried out on these
biopsies with anti-dystrophin antibodies (DYS2 and DYS3,
Novocastra, Newcastle upon Tyne, UK), anti-dystrophin
related protein antibody (a gift from T. Khurana, Boston,
Mass., USA), and anti-50 kDa dystrophin-associated
glycoprotein antibodies (a gift from K. P. Campbell, Iowa
City, la., USA).
Pedigrees of the families were established. Thanks to the
cooperation of the 'Cercle G6nealogique de Bourbon', links
between the different nuclear families were identified and
traced back 12 generations. Intrasibship resemblance for
age of onset was approached by calculating the intra-class
correlation coefficient (Haldane, 1941).
Material and methods
An exhaustive field study was conducted in Reunion Island,
starting with the first families studied in the Saint-Pierre
Blood from healthy and affected members of the families
was collected by one of us (D.H.). Three patients belonging
to the group of eight additional patients were included
by guest on May 26, 2012
Juvenile limb-girdle muscular dystrophy 297
Fig. 1 Histogram of the ages of clinical onset.
in this study. Transformed lymphoblastoid cell lines were
established and the DNA was extracted. In an initial study
(Beckmann et al., 1991), a panel of 85 restriction fragment
length polymorphism probes was utilized for systematic
hybridization. These markers were selected on the basis of
their relative location on the genome (at about every 20 cM).
Subsequently, additional probes, mainly microsatellites, were
used in order to restrict the genetic interval around the
presumed LGMD2A locus (Fougerousse et al., 1994;
Allamand et al., 1995). Reconstruction of carrier haplotypes
was inferred upon phase determination by minimizing the
number of crossing-over events. These haplotypes consist of
close to 30 microsatellites and other markers spanning a 10
cM region around the disease locus. Screening for calpain
mutations was performed as described by Richard et al.
(1995). Polymerase chain reaction amplified products made
on DNA from peripheral blood from defined patients were
examined by heteroduplex analyses or by systematic
sequencing. In some instances, mutation detection was
performed on reverse transcription-polymerase chain reaction
amplified products, taking advantage of the illegitimate
transcription in immortalized lymphocytes (Chelly et al.,
The 20 patients (12 males and eight females, aged 15-64
years) were examined at different stages of the disease. The
pattern of muscular involvement was fairly similar in all
patients. The clinical data will be presented together.
The average onset of clinical manifestations was 13.5
years, with a large range of 3-30 years (Fig. 1). Mean age
of onset was 16.25±7.02 in males and 10.50±7.86 in females.
Fifty per cent of the patients presented with their first
manifestations at 8-12 years. In all cases, the first complaints
were difficulties in running, climbing stairs or mountains and
frequent falls. Only two patients (Cases 2 and 7) reported,
at the same time, difficulty with carrying heavy loads. Usually,
the time lapse between functional deficiencies of the pelvic
and scapular muscles was 2.5-5 years (Table 1).
When examined in the early stages of the disease (stages
II/III), patients had waddling gait and slight hyperlordosis.
Muscle weakness predominated in gluteus maximus and
adductors, and to a lesser degree, in gluteus medius, psoas
and biceps cruri. At this stage, even if not considered by the
patients as functionally limiting, the weakness was also
evident in the scapular girdle muscles, predominating in
latississimus dorsi, rhomboid, serratus magnus, pectoralis
major and, to a lesser degree, in trapezius, deltoid, biceps
brachialis and brachio-radialis (Fig. 2). Rectus abdominis and
abdominal muscles were markedly affected. There was clear
evidence of atrophy in the same territories, especially in the
glutei, posterior muscles of the thigh, pectoralis, rhomboid
and biceps muscles and there was slight contracture of the
At more advanced stages of the disease, characterized by
an inability to climb stairs (stage IV) or to rise up from a
chair (stage V), weakness and atrophy predominated in the
same groups of muscles, extending, in the upper limbs, to
the triceps brachialis and to a lesser degree to the radialis
and cubital muscles. The supra- and infraspinatus muscles
were spared. In the lower limbs, the weakness extended to
the quadriceps and, to a minor degree, to the tibialis anterior
and triceps surae. Spinalis dorsi were weak, but less so than
the abdominal muscles (Fig. 3).
When the patient was no longer able to walk without aid
(stage VI) the distribution of the muscle weakness remained
identical, extending in the upper limbs to the long extensor
and abductor of the thumb and to the palmaris muscle, and
in the lower limbs to the peroneus and tibialis posterior
muscles. The patient had marked hyperlordosis, contractures
of the triceps surae and also developed early contractures of
the elbow, hip and knee flexors. When the patient was
wheelchair bound (stage VII), severe contractures developed
around the hips, knees and elbows. Kyphoscoliosis occurred
in five patients. Only the small muscles of the hand and neck
muscles were, at this stage, of normal strength.
At all stages of the disease, muscle weakness predominated
in limb-girdle and trunk muscles and was almost symmetrical,
with very selective involvement of certain groups of muscles.
Marked weakness of neck muscles was rarely noticed and
these muscles often remained at normal strength in the late
stages. Facial muscles only showed exceptionally a minor
clinical involvement: one patient (Case 3) presenting with
the earliest onset and the most severe form, had a slight,
symmetrical weakness of the facial muscles. There was never
any clinical involvement of ocular and velopharyngeal
The disease process was almost purely atrophic and CT
scan examination clearly revealed this selective involvement
(Fig. 4). Only once was a transient calf hypertrophy noticed
(Case 15). Contractures were limited to calf muscles in the
early stages of the disease (three patients had an Achilles
lengthening in their early teens), but developed rapidly in
the late stages when the patients had difficulties in getting
up from a chair. The idio-muscular contractility was abolished
by guest on May 26, 2012
M. Fardeau et al.
clinical data on
Age in 1991
Age of onset of
Age of onset of
Age at loss
Achilleus length at 16 years
Achilleus length at 11 years
Achilleus length at 13 years
Died in 1991
•Functional stages graded according to Gardner-Medwin and Walton (1974): I = unable to run freely or stands up without using hands;
II = dandling gait, climb stairs without using the bannister; III = climb stairs only with the bannister; IV = unable to climb stairs; V =
unable to rise from a chair without aid; VI = walks only with aid, frequent falls; VII = unable to walk and to stand erect.
heart was found to be normal in all cases. ECG and
echocardiographic examination did not reveal any significant
abnormality except in one case (14), in which there was a
slight diminution of left ventricle contractility. In the later
stages of the disease, cardiac abnormalities were considered
to correlate with the respiratory insufficiency.
Psychometric evaluation showed below normal levels
(mean IQ = 77, with a range of 61-94), and verbal and
performance IQs of the same order. These values should be
interpreted in relation to the educational and linguistic levels
of the population, most of them Creole speaking. None of
the patients were considered as intellectually abnormal by
their families or neighbours.
Electromyographic examination showed abnormalities
compatible with a myopathic process in all examined
territories. No myotonia or spontaneous discharges were
found. Motor conduction velocities were normal.
Serum creatine kinase levels were usually abnormal, up to
20 times the normal values in the early stages, then decreasing
to reach nearly normal levels in wheelchair bound patients.
The disease was considered by all patients and their physicians
as slowly and steadily progressive. It should be noticed that
the rate of deterioration, roughly indicated by the time
lapse between the first symptoms (stage I) and the age of
commencement of the permanent use of a wheelchair (stage
VII), varied greatly (Fig. 5), but the earlier the onset, the
Fig. 2 Muscular involvement pattern. (A) Case 14. Predominance
of the atrophy on the glutei and posterior thigh muscles. (B) Case
7. Predominance of the atrophy and weakness on serratus magnus,
latississimus dorsi and trapezius muscles, with respect of supra,
infraspinatus and deltoid muscles.
in the early stages. Tendon reflexes were weak or absent in
most cases. In two brothers (Cases 10 and 11) gynaecomastia
A deficiency in diaphragm function was found in the
majority (12 out of 20) of the patients, resulting in a reduction
of lung vital capacity to 30-50% at stages V-VII. Only in
one patient (Case 18) did severe respiratory insufficiency
lead to death, at age 19 years. On clinical examination, the
by guest on May 26, 2012
Juvenile limb-girdle muscular dystrophy 299
Deltoid Triceps br.
Degree of weakness
Fig. 3 Schematic representation of the sequential involvement of 20 muscles in the Reunion Island type
with onset after 20 years remained ambulatory even, in one
of them, at an advanced age. Only one patient died during
the time of these observations: her first symptoms appeared
at around 8 years of age, she was in a wheelchair at 12 years
and died at 19 years.
It is noticeable that there was marked interfamilial
variability—the two most severe cases, and the three most
benign belong to two different nuclear families, whereas the
intrafamilial variability was low, both for the age of onset
and the stage of motor deterioration (Fig. 5).
Biopsies taken from the medial head of the left deltoid
muscle of five patients, two at stage III, two at stage V and
one at stage VII, were examined.
Light microscopic study of transverse sections of these
biopsies (Table 2) showed variation in muscle fibre diameter
without group atrophy, increased number of internal nuclei
and the presence of some necrotic and regenerating fibres.
On myofibrillar ATPase staining, the type I:type II ratio
was nearly normal in mildly affected patients and shifted
towards a type I predominance in those more affected, with
type I uniformity in the most affected one (Fig. 6). Mean
diameter of type I fibres was normal in the less affected
patients and decreased in the others. The diameter of type II
fibres, also normal in the less affected patient tended to be
slightly decreased in the more affected ones. A bimodal
distribution of type I (Case 11) and type II fibres (Cases 13,
11 and 17) with a peak of small fibres was found in three
biopsies (Table 2).
On the oxidative enzyme stains, an abnormal lobulation
of the type I fibres was seen in every biopsy, more marked
in the most severely affected patients (Fig. 6). Endomysial
Fig. 4 CT scan of thigh muscles. (A) Case 7. Selective
involvement of biceps cruri. (B) Case 12. Generalized
involvement of the thigh muscles, predominating on posterior
faster the progression. In two patients with very early
clinical manifestation (before age 6 years), loss of ambulation
occurred before age 10 years. In most patients with a clinical
onset between 8 and 12 years (10 patients) the loss of
ambulation occurred at around 20 years. The five patients
by guest on May 26, 2012
M. Fardeau et al.
Fig. 5 Evolution curves of the 20 LGMD patients. Functional stages graded from I to VTI according to
the Gardner-Medwin and Walton scale (1974). Small figures indicate the case number. Note that Cases
2 and 3 were sisters, and Cases 6, 7 and 8 belong to the same nuclear family.
Table 2 Main histopathological data from five muscle biopsies
Muscle fibre changes*
I muscle fibres
•Functional stages graded as in Table 1; deltoid strength graded according to manual testing score (MRC, 1943); intensity of the
muscle fibre changes graded from 0 (absence) to + + + (very marked/frequent); 'mean diameter of muscle fibres indicated by 'xx' when
the histogram showed a bimodal distribution.
collagen was markedly increased in most of the biopsies. In
two biopsies, motor end-plate morphology was examined in
teased muscle fibres using Koelle's technique, and was found
to be normal.
Ultrastmctural examination of the five biopsies did not
show any specific change and confirmed the great frequency
of myofibrillar disarray. In one biopsy, a small tubular
aggregate was found. A few motor end-plates were examined
in one biopsy, and their morphology was considered to
be normal. Immunocytochemical studies showed a normal
sarcolemmal labelling with anti-dystrophin, anti-utrophin
(dystrophin-related protein) and anti-adhalin (50 kDa
dystrophin-associated glycoprotein) antibodies (Fig. 7).
In summary, all these findings were compatible with a
primary muscular dystrophic process, without any specific
structural change or any detectable sarcolemmal protein
Genetic and epidemiological data
The 20 patients belong to eight nuclear families, five of them
being multiplex (Fig. 8). In most of them, the parents are
known to be close or distant cousins and also some of these
families are closely related, as the same patronymics (amongst
the most common of the first inhabitants of French origin on
the island) appear amongst the ascendants. Thanks to the
cooperation of the families, and to the records of the 'Cercle
Genealogique de Bourbon,' one common ancestor of 17
patients, who landed on the island on November 19, 1674,
has been identified (De Ubeda, 1992).
These families belong to a small community which
remained, for political, socio-economical and geographical
reasons, excluded from the rest of the population in the
middle of the nineteenth century, forming an isolate in the
high mountains of the south of the island (hence they were
called 'les Petits Blancs des Hauts'). Up to now, most of
Fig. 6 Muscle deltoid biopsies. Cryostat serial sections (X192): A-C, Case 7; D-F, Case 13; G-I, Case 12; J-L, Case 17. A, D, G and
F, haematein-eosin; B, E, H and K, NADH-tetrazolium reductase; C, F, I and L, ATPase after acid (pH 4.65) preincubation.
by guest on May 26, 2012
Juvenile limb-girdle muscular dystrophy 301
by guest on May 26, 2012
M. Fardeau et al.
Fig. 7 Indirect immunofluorescence with monoclonal antibodies
against dystrophin (NCL DYS2, Novocastra) and adhalin (IVD3,)
(generous gift of Dr Kevin P. Campbell) on serial transverse
cryosections of Patient 17 (Family 21) showing normal labelling
of sarcolemma of muscle fibres with both antibodies. It is worth
noting that in this patient the muscle-specific calpain is presumed
to be totally missing, as the mutation leads to an aberrant splicing
(Richard et al., 1995).
these families have lived in a few small poor villages around
Saint-Pierre, the main town of this part of the island (Fig. 9).
These conditions were propitious for inbreeding through
intermarriages and thus to manifestations of recessively
inherited traits. Complete genealogical records were difficult
to establish in some families. Even then, it should be noticed
that close (first cousin marriage) consanguinity was observed
only in two families. For most of them, however, con-
sanguinity appeared when traced back over several
generations. Moreover, intrafamilial comparison of age of
onset in the sibships with two or more affected siblings gave
an intra-class coefficient of 0.98 (/> < 0.001), which was
suggestive of the presence of two or more deleterious genes
in this population.
Thanks to the cooperation of many physicians, neurologists
and pediatricians of the island, an exhaustive field study was
conducted. This gave a minimum prevalence of 28 per
578 000 (population of Reunion Island), i.e. 4.8Xlfr5. It
should be added that associated disorders were scanty in
these families (one case of cystic fibrosis in family 15 and
two cases of gynaecomastia in family 17).
Molecular genetics studies and genotype/
The Reunion families were first examined for the co-
segregation of the disease trait with a series of restriction
fragment length polymorphism markers selected on the basis
of their relative location on the genome (spread approximately
every 20 cM). A linkage on chromosome 15 was obtained
with marker D15S25, giving a total lod score of 5.52 at
0 = 0.0, assuming a penetrance of 1.0. Use of additional
microsatellite probes and additional chromosome 15 linked
families revealed that the disease locus was flanked by two
markers separated by 1 cM, an interval which was found to
span a 4 Mb region (Allamand et al., 1995). Genotyping of
the R6union pedigrees with 28 markers extending over 10
cM around the LGMD2A locus led to the unexpected
disclosure of haplotype heterogeneity in this population
(Allamand et al., 1995). This was confirmed upon the
validation of the muscle-specific calpain gene (CANP3) in
the aetiology of the disease (Richard et al., 1995). In all
patients of this cohort except one (Case 18), mutations in
the CANP3 gene were detected. Four different mutations
were identified (Richard et al., 1995). Two of these were
effectively null mutants resulting from one splice site and
one frameshift mutation. The remaining two were missense
mutations (Table 3).
Exhaustive analyses of the calpain gene revealed a fifth
mutation, present in family 15, which involves a C-»T
transition at nucleotide 1872 within exon 16. The consequence
of this mutation was the creation of a new splice site,
upstream of the normal site, leading to a shortened mRNA
product and to the appearance of an out of phase stop codon
18 nucleotides after the new splice site (Table 3). Hence, five
distinct mutations have now been identified in this small
These findings allowed us to attempt a genotype/phenotype
correlation. Evaluation of the clinical progression curves in
terms of the specific type of mutation clearly shows that
different mutations may result in both different ages of
onset and degrees of severity. Homozygous patients carrying
presumed null mutations showed a more severe profile than
patients carrying a missense mutation, with heterozygous
individuals showing intermediate profiles (Fig. 10).
The muscular -disorder observed in a small community of
French origin in Reunion Island specifically involves limb-
girdle and trunk muscles, in a symmetrical fashion, with a
very selective pattern of involvement within the different
muscle groups. Clinical onset is most often between the ages
of 8 and 12 years, but with a larger range of 3-30 years. It
is characterized by weakness of pelvic girdle muscles. The
process is mainly atrophic, calf hypertrophy being very rarely
noticed. Contractures are initially limited to calf muscles.
by guest on May 26, 2012
Juvenile limb-girdle muscular dystrophy
2. 2 3 22
Family 11 F a m i'y 12
i A 1. i
A i A i i A i
1 •" 2 3
9 10 12 1 1
1 A U ATI
O 1 D
1. 4 i A
Fig. 8 Genealogic pedigrees of the Reunion Island LGMD families. The families numbers refer to the numbers used in Beckmann el al.
(1991). The numbers of the 20 original cases are identified. Additional cases are indicated by *. + refers to a case of cystic fibrosis.
Facial muscles are almost constantly spared. Ocular and
velopharyngeal muscles are never involved. There is a slow,
steady deterioration of the muscle strength, leading to loss
of ambulation in the early 20s for 50% of the patients. Serum
creatine kinase is markedly increased in the early phases of
the disease, then decreases to normal values in the late stages.
These clinical and biological features are in close accordance
with the criteria most usually reserved for the description of
limb-girdle muscular dystrophies (Walton and Nattrass, 1954;
Munsat, 1977; Bushby, 1992, 1995). They correspond in the
pattern of muscular involvement and in the profile of
evolution, to the initial description given, at least for some
of his cases, by Erb (1884), who proposed to use the term
'a juvenile form of progressive muscular dystrophy'. The
striking homology in the selective pattern of peri-scapular
muscle involvement between Erb's patients (Erb, 1884;
Fig. 1) and the Reunion Island patients was seen in the
photographs of their silhouettes.
by guest on May 26, 2012
M. Fardeau et al.
Table 3 Characteristics of the different mutations found in 'Reunion Island' LGMD patients
Effect of mutation
S 744 G
*Homozygosity by haplotypes and/or mutations; ^nucleotide position refer to CANP3 gene structure (Richard et al., 1995).
ENTRE-DEUX • • • •
Fig. 9 Geographical distribution of the patients on the R6union
Island map. Concentration of the families around Saint-Pierre
Much confusion arose later in the definition of LGMD.
With the development of electrophysiological and histo-
pathological techniques, it appeared that many muscular
disorders of neuropathic, metabolic, congenital or inflam-
matory origin may roughly mimic the distribution of the
muscular deficit of LGMD (Brooke, 1977). It was thus
important to stress the very peculiar sequence of muscle
involvement which characterizes LGMD, the generally
moderate involvement of the diaphragm and respiratory
muscles, the normality of heart function and the absence of
any intellectual disturbances. Electromyographic exam-
inations revealed small, polyphasic potentials without any
sign of neuropathic involvement, and motor and sensory
conduction velocities were consistently normal.
Despite the difficulties in their identification, cases of
LGMD have occasionally been reported, particularly in some
inbred communities in Switzerland (the Bern area) (Pfa'ndler,
1950; Moser et al., 1966) and in the Amish and Hutterite
communities living in North America (Kloepfer and Talley,
1957; Jackson and Carey, 1961; Jackson and Strehler, 1968;
Shokeirand Kobrinsky, 1976; Shokeir and Rozdilsky, 1985).
A significant number of familial cases with a predominant
limb-girdle distribution and autosomal recessive transmission
were also reported in Tunisia (Ben Hamida and Fardeau,
1980; Ben Hamida et al., 1983) and then in other peri-
Mediterranean or Middle-Eastern countries (Dubowitz, 1980;
Salih et al., 1983). However, in these areas, the SCARMD
differed from Erb's initial description and from the Reunion
Island cases in their clinical pattern which was close to the
Duchenne/Becker phenotype, with a greater frequency of calf
hypertrophy (in >50% cases), more marked contractures, an
earlier age of onset and a slighly more severe evolution.
Similar cases were also reported in other countries (Somer
et al., 1985; Zatz et al., 1989; Me Guire and Fischbeck,
1991; Passos-Bueno et al., 1993). Large families in which
affected members were identified as LGMD, with a relatively
later onset and a more benign evolution were detected
elsewhere (Mahjneh et al., 1992). Clinical diagnosis of
LGMD was also made in patients belonging to large pedigrees
exhibiting an autosomal dominant transmission (De Coster
et al., 1974; Chutkow et al., 1986; Gilchrist et al., 1988).
Histopathological and molecular genetic studies has greatly
helped to clarify the nosological situation within the
The Reunion Island LGMD is pathologically defined by a
necrotic-regenerative histopathological pattern, with a large
variation in size of the muscle fibres and the presence of
many lobulated fibres in the end-stage of the disease.
Dystrophin was found to be normal both by immuno-
cytochemical and biochemical
(dystrophin-related protein) was also found to be normal, as
well as adhalin (the 50 kDa component of the dystrophin-
associated glycoprotein complex) (Ervasti and Campbell,
1993). These negative results allowed us to distinguish
between LGMD and other muscular dystrophies, which had
for long time been frequently difficult to diagnose, i.e. the
Becker muscular dystrophies or Duchenne/Becker hetero-
zygotes with a marked clinical expression (Hoffman et al.,
1992). It also enabled us to distinguish LGMD from the
SCARMD in which adhalin deficiency was demonstrated
(Matsumura et al., 1992). Recently, two SCARMD loci have
been mapped, one on the chromosome 13q (Ben Othmane
by guest on May 26, 2012
Juvenile limb-girdle muscular dystrophy 305
Fig. 10 Correlation between clinical severity and the gene mutations. Continuous lines correspond to null mutations previously identified
(Richard et al., 1995); the bold continuous line corresponds to a new null mutation discovered in Family 15; lines constructed of plus
signs (+ + + +) correspond to missense mutations, in homo- (+ + +) or heterozygous conditions (+ + +).
et al., 1992), the other on 17q. The latter was demonstrated
to be the gene encoding the 50 kDa adhalin, and missense
mutations were identified within this gene (Roberds et al.,
Genetic and epidemiological data
The geographical concentration of the patients in a small
area of R6union Island, and the historical and socio-economic
reasons of segregation were highly conducive to the formation
of a genetic isolate. An autosomal recessive mode of
inheritance was supported by the fact that both sexes were
affected while the parents were clinically unaffected.
Analysis of the genealogy of the pedigrees confirmed the
high degree of consanguinity within this population, with
marked relationships between the different identified families.
It is noticeable that prevalence figures of LGMD in Reunion
Island were very close to that found in the Bern area (Moser
et al., 1966) and higher than the classical figures (20X10"6)
given by Morton and Chung (1959).
This highly probable genetic homogeneity of R6union
Island community was considered as a favourable condition
for linkage studies, and significant lod-scores were obtained
for chromosome 15q. It is noticeable that positive linkage
on this chromosome was also obtained in the Old Order
Amish communities (Young et al., 1992) and in Brazilian
families (Passos-Bueno et al., 1993).
In contrast, it was demonstrated that the gene for LGMD
in some affected Brazilian families (Passos-Bueno et al.,
1993) did not map to chromosome 15 and, recently, LGMD
in a large inbred Palestinian community was demonstrated
to be linked to chromosome 2p (Bashir et al., 1994). As for
the dominant form of LGMD, it has been shown to be linked
to chromosome 5q (Speer et al., 1992) in one large North
All these data clearly demontrate the genetic heterogeneity
of LGMD, and a new classification has been proposed
(Bushby, 1995). This underlines the major importance of a
precise analysis of the phenotypic expression of these
different limb-girdle muscular dystrophies, and to compare
them with those which are genetically well defined, such as
those observed in the Rdunion Island community. In this
group, the pattern of muscular involvement does not vary
significantly: the very precise and selective involvement of
skeletal muscles, the timing and location of contractures, the
rarety of calf hypertrophy, provide a picture quite reproducible
from one patient to another and from one family to another.
However, the evolution, with respect to severity, varies: the
earlier the age at clinical onset, the more severe the
progression. Furthermore, it is worth noticing that each
mutation class seems to form a rather homogeneous group.
If seems that patients lacking the muscle-specific calpain
activity, as judged on the basis of the nature of the mutations,
are more severely affected than those that may have a residual
activity. It will be of interest to test whether the presence of
the protein can be demonstrated amongst carriers of missense
mutations, since these could influence its turnover rate. Also,
it will be interesting to characterize the eventual phenotypic
variability in calpain-linked families already identified outside
the R6union Island community (M.F., J.B. and I.R., study in
Finally, the identification of multiple mutations within the
Rdunion Island LGMD patients was totally unexpected in
what was considered to be a highly inbred community. In
two families, compound heterozygotes can be found (each
spouse marrying into the family contributing a different
CANP3 allele). This phenomenon of multiple mutations is
still an unexplained enigma, as it is difficult to reconcile this
observation with the very low prevalence of LGMD in the
general population (Emery, 1991) These results may require
a reevaluation of the prevalence of this dystrophy and of the
different mutations in the general population. They may also
suggest a more complex genetic process, for instance, that
by guest on May 26, 2012
M. Fardeau et al.
LGMD follows a digenic inheritance pattern (Richard
et al., 1995).
All these data and hypotheses highlight once more the
importance of precise clinico-genetic correlations in the
heterogeneous group of limb-girdle muscular dystrophies.
Clearly, the identification of the 'LGMD2A' locus has greatly
helped with the nosological definition of this clinical entity,
confirming the accuracy of Erb's initial descriptions.
We wish to thank the following: A. Rouche, P. Bozin and
M. Roumane, from INSERM 153, for photographic work,
image analysis and preparation
V. Allamand, F. Fougerousse,
Chiannilkulchai, from the Ge'ne'thon laboratory, for molecular
genetics studies; T. Khurana and K. P. Campbell for the gift
of antibodies; K. Bushby for critical reading of the manuscript;
all the neurologists, paediatricians and geneticists, from Saint-
Pierre and Saint-Denis-de-la-R6union, for helping with the
clinical studies; M. Begue, physiotherapist from Le Tampon
and Mrs Schmidt, psychologist from Saint-Pierre Hospital,
for the evaluation of the patients; and B. Remi (Reunion
Island University) for the geographical study. Finally, we
would like especially to thank Idriss Cadjee of the
'Association R6unionnaise contre la Myopathie', and all the
patients and families without whose participation this work
would never have seen the light. This work was supported
by INSERM (R6seau de Recherche Clinique) and by grants
from Association Franchise contre les Myopathies (A.F.M.).
Allamand V, Broux O, Richard I, Fougerousse F, Chiannilkuchai
N, Bourg N, et al. Preferential localization of the Limb-Girdle
Muscular Dystrophy type 2A gene in the proximal part of a lcM
15ql5-l-ql5-3. Am J Hum Genet 1995; 56: 1417-30.
Bashir R, Strachan T, Keers S, Stephenson A, Mahjneh I, Marconi
G, et ai. A gene for autosomal recessive limb-girdle muscular
dystrophy maps to chromosome 2p. Hum Mol Genet 1994; 3:455-7.
Batten FE. The myopathies or muscular dystrophies. Q J Med 1909;
Beckmann JS, Richard I, Hillaire D, Broux O, Antignac C, Bois E,
et al. A gene for limb-girdle muscular dystrophy maps to
chromosome 15 by linkage. C R Acad Sci III 1991; 312: 114-18.
Bell J. On pseudohypertrophic and allied types of progressive
muscular dystrophy. In: Fischer RA, editor. The treasury of human
inheritance, Vol. 4, Pt. 4. London: Cambridge University Press,
Ben Hamida M, Fardeau M. Severe, autosomal recessive, limb-
girdle muscular dystrophies frequent in Tunisia. In: Angelini C,
Danieli GA, Fontanari D, editors. Muscular dystrophy research:
advances and new trends. Amsterdam: Excerpta Medica, 1980:
dystrophy affecting both sexes and frequent in Tunisia. Muscle
Nerve 1983; 6: 469-80.
Ben Othmane K, Ben Hamida M, Pericak-Vance M, Ben Hamida
C, Blel S, Carter SC, et al. Linkage of Tunisian autosomal recessive
Duchenne-like muscular dystrophies to the pericentromeric region
of chromosome 13q. Nat Genet 1992; 2: 315-17.
Bradley WG. The limb-girdle syndromes. In: Vinken PJ, Bruyn
GW, Ringel SP, editors. Handbook of clinical neurology, Vol. 40.
Amsterdam: North-Holland, 1979: 433-69.
Brooke MH. A clinician's view of neuromuscular diseases.
Baltimore: Williams and Wilkins, 1977.
Bushby K. Report on the 12th ENMC sponsored international
workshop—the 'limb-girdle' muscular dystrophies. Neuromuscul
Disord 1992; 2: 3-5.
Bushby KMD. Diagnostic criteria for the limb-girdle muscular
dystrophies: report of the ENMC consortium on limb-girdle
dystrophies. Neuromuscul Disord 1995; 1: 71-4.
Bushby KMD, Gardner-Medwin D. The clinical, genetic and
dystrophin characteristics of Becker muscular dystrophy. I. Natural
history [published erratum appears in J Neurol 1993; 240: 453]. J
Neurol 1993; 240: 98-104.
Chelly J, Concordet JP, Kaplan JC, Kahn A. Illegitimate
transcription: transcription of any gene in any cell type. Proc Natl
Acad Sci USA 1989; 86: 2617-21.
Chiannilkulchai N, Pasturaud P, Richard I, Auffray C, Beckmann
JS. A primary expression map of the chromosome 15ql5 region
containing the recessive form of limb-girdle muscular dystrophy
(LGMD 2A) gene. Hum Mol Genet 1995; 4: 717-25.
Chutkow JG, Heffner RR Jr, Kramer AA, Edwards JA. Adult-onset
autosomal dominant limb-girdle muscular dystrophy. Ann Neurol
1986; 20: 240-8.
De Coster W, De Reuck J, Thiery E. A late autosomal dominant
form of limb-girdle muscular dystrophy. A clinical, genetic, and
morphological study. Eur Neurol 1974; 12: 159-72.
De Ubeda B. Formule clinique et dvolution longitudinale de la
Dystrophie Musculaire des Ceintures 're'unionnaise', Vol. 1. Tours:
Faculty of Medicine, Med Thesis, 1992.
Dubowitz V. Rapidly progressive limb girdle muscular dystrophy
in childhood. In: Angelini C, Danieli GA, Fontanari D, editors.
Muscular dystrophy research: advances and new trends. Amsterdam:
Excerpta Medica, 1980: 129-33.
Dubowitz V, Brooke M. Muscle biopsy: a modem approach. London:
Emery AEH. Population frequencies of inherited neuromuscular
diseases—a world survey. [Review]. Neuromuscul Disord 1991; 1:
Erb WH. Ueber die 'juvenile Form' der progressiven Muskelatrophie
und ihre Beziehungen zur sogenannten Pseudohypertrophie der
Muskeln. Dt Arch Klin Med 1884; 34: 467-519.
Erb W. Dystrophia muscularis progressiva. Dt Z Nervheilk 1891;
1: 13-94, 173-261.
Ben Hamida M, Fardeau M, Attia N. Severe childhood muscular Ervasti JM, Campbell KP. A role for the dystrophin-glycoprotein
by guest on May 26, 2012
Juvenile limb-girdle muscular dystrophy 307
complex as a transmembrane linker between laminin and actin. J
Cell Biol 1993; 122: 809-23.
Fardeau M, Hillaire D, Mignard C, Mignard D, Collin H. Limb-
girdle muscular dystrophies frequent in Reunion Island [abstract].
Proceedings of the XIVth World Congress of Neurology; 1989 Oct
22-27; Neurology India 1989; 37 Suppl.
Fardeau M, Matsumura K, Tome" FMS, Collin H, Leturcq F,
Kaplan JC, et al. Deficiency of the 50 kDa dystrophin-associated
glycoprotein (adhalin) in severe autosomal recessive muscular
dystrophies in children native from European countries. C R Acad
Sci III, Sciences de la vie/Life Sciences 1993; 316: 799-804.
Fougerousse F, Broux O, Richard I, Allamand V, de Souza AP,
Bourg N, et al. Mapping of a chromosome 15 region involved in
limb girdle muscular dystrophy. Hum Mol Genet 1994; 3: 285-93.
Gardner-Medwin D, Walton JN. The clinical examination of the
voluntary muscles. In: Walton JN, editor. Disorders of voluntary
muscles. 3rd ed. Edinburgh: Churchill Livingstone, 1974: 517-60.
Gilchrist JM, Pericak-Vance M, Silverman L, Roses AD. Clinical and
genetic investigation in autosomal dominant limb-girdle muscular
dystrophy. Neurology 1988; 38: 5-9.
Haldane JBS. The relative importance of principal and modifying
genes in determining some human diseases. J Genetics 1941; 41:
Hoffman EP, Kunkel LM. Dystrophin abnormalities in Duchenne/
Becker muscular dystrophy. [Review]. Neuron 1989; 2: 1019-29.
Hoffman EP, Arahata K, Minetti C, Bonilla E, Rowland LP.
Dystrophinopathy in isolated cases of myopathy in females.
Neurology 1992; 42: 967-75.
Jackson CE, Carey JH. Progressive muscular dystrophy: autosomal
recessive type. Pediatrics 1961; 28: 77-84.
Jackson CE, Strehler DA. Limb-girdle muscular dystrophy: clinical
manifestations and detection of preclinical disease. Pediatrics 1968;
Kloepfer HW, TaHey C. Autosomal recessive inheritance of
Duchenne-type muscular dystrophy. Acta Genet (Basel) 1957; 7:
Landouzy L, Dejerine J. De la myopatfiie atrophique progressive.
Myopathie h6r£ditaire, sans neuropathic ddbutant d'ordinaire dans
l'enfance, par la face. Rev Med 1885; 5: 81-117, 253-366.
Landouzy L, Dejerine J. Nouvelles recherches cliniques et anatomo-
pathologiques sur la myopathie atrophique progressive. A propos
de six observations nouvelles, dont une avec autopsie. Rev Med
1886; 6: 977-1027.
Levison H. Dystrophia musculorum progressiva. Acta Psychiat
Neurol Scandinav 1951; Suppl 76: 1-176.
Leyden E. KJinik der RUckenmarks-krankheiten. Berlin: Hirschwald,
Mahjneh I, Vannelli G, Bushby K, Marconi GP. A large inbred
Palestinian family with two forms of muscular dystrophy.
Neuromuscul Disord 1992; 2: 277-83.
Matsumura K, Tom6 FMS, Collin H, Azibi K, Chaouch M, Kaplan
JC, et al. Deficiency of the 50K dystrophin-associated glycoprotein
in severe childhood autosomal recessive muscular dystrophy. Nature
1992; 359: 320-2.
McGuire SA, Fischbeck KH. Autosomal recessive Duchenne-like
muscular dystrophy: molecular and histochemical results. Muscle
Nerve 1991; 14: 1209-12.
Medical Research Council. Aids to the invesitgation of peripheral
nerve injuries. Medical Research Council War Memorandum no. 7.
2nd ed. London: HMSO, 1943.
MObius PJ. Ueber die hereditSren Nervenkrankheiten. Samml Klin
Votr 1879; 171: 1505-31.
Morton NE, Chung CS. Formal genetics of muscular dystrophy.
Am J Hum Genet 1959; 11: 360-79.
Moser H, Wiesmann U, Richterich R, Rossi E. Progressive
Muskeldystrophie. VIII. HaUfigkeit, Klinik und Genetik der Typen
I und H. Schweiz Med Wochenschr 1966; 96: 169-74, 205-11.
Munsat TL. Classification of human dystrophies and pathogenesis
of human muscular dystrophies. Proceedings of the Vth International
Congress on Neuromuscular Diseases. Amsterdam: Excerpta
Medica; 1977: 23-31.
Passos-Bueno MR, Richard I, Vainzof M, Fougerousse F,
Weissenbach J, Broux O, et al. Evidence of genetic heterogeneity
in the autosomal recessive adult forms of limb-girdle muscular
dystrophy following linkage analysis with 15q probes in Brazilian
families. J Med Genet 1993; 30: 385-7.
Pfandler U. Eine einfach rezessive Form der Dystrophia musculorum
progressiva mit einer Sippenstammtafel aus dem Emmental
(Schweiz.). Deutsch Med Wchnschr 1950; 75: 1221-5.
Richard I, Broux O, Allamand V, Fougerousse F, Chiannilkulchai
N, Bourg N, et al. A novel mechanism leading to muscular
dystrophy: mutations in calpain 3 cause limb-girdle muscular
dystrophy type 2A. Cell 1995; 81: 27-40.
R&berds SL, Leturcq F, Allamand V, Piccolo F, Jeanpierre M,
Anderson RD, et al. Missense mutations in the adhalin gene linked
to autosomal recessive muscular dystrophy. Cell 1994; 78: 625-33.
Romero NB, Tome' FMS, Leturcq F, el Kerch F, Azibi K, Bachner
L, et al. Genetic heterogeneity of severe childhood autosomal
recessive muscular dystrophy with adhalin (50 kDa dystrophin-
associated glycoprotein) deficiency. C R Acad Sci III. Sciences de
la vie/Life Sciences 1994; 317: 70-6.
Salih MAM, Omer MIA, Bayoumi RA, Karrar O, Johnson M.
Severe autosomal recessive muscular dystrophy in an extended
Sudanese kindred. Dev Med Child Neurol 1983; 25: 43-52.
Shokeir MH, Kobrinsky NL. Autosomal recessive muscular
dystrophy in Manitoba Hutterites. Clin Genet 1976; 9: 197-202.
Shokeir MHK, Rozdilsky B. Muscular dystrophy in Saskatchewan
Hutterites. Am J Hum Genet 1985; 22: 487-93.
Somer H, Voutilainen A, Knuutila S, Kaitila I, Rapola J, Leinonen
H. Duchenne-like muscular dystrophy in two sisters with normal
karyotypes: evidence for autosomal recessive inheritance. Clin Genet
1985; 28: 151-6.
Speer MC, Yamaoka LH, Gilchrist JH, Gaskell CP, Stajich JM,
Vance JM, et al. Confirmation of genetic heterogeneity in limb-
by guest on May 26, 2012
308 Download full-text
M. Fardeau et al.
girdle muscular dystrophy: linkage of an autosomal dominant form
to chromosome 5q. Am J Hum Genet 1992; 50: 1211-17.
Stevenson AC. Muscular dystrophy in Northern Ireland. I. An
account of the condition in fifty-one families. Ann Eugenics 1953;
Stevenson AC. Muscular dystrophy in Northern Ireland. II. An
account of nine additional families. Ann Human Genet 1955; 19:
Tyler F, Wintrobe MM. Studies in disorders of muscle I. The
problem of progressive muscular dystrophy. Ann Int Med 1950; 32:
Walton JN, Nattrass FJ. On the classification, natural history and
treatment of the myopathies. Brain 1954; 77: 169-231.
Young K, Foroud T, Williams P, Jackson CE, Beckmann JS, Cohen
D, et al. Confirmation of linkage of limb-girdle muscular dystrophy,
type 2, to chromosome 15. Genomics 1992; 13: 1370-1.
Zatz M, Passos-Bueno MR, Rapaport D. Estimate of the proportion
of Duchenne muscular dystrophy with autosomal recessive
inheritance. Am J Med Genet 1989; 32: 407-10.
Received April 26, 1995. Accepted July 22, 1995
by guest on May 26, 2012