

Am. J. Hum. Genet. 64:1524–1540, 1999

1524

Calpainopathy—A Survey of Mutations and Polymorphisms
I. Richard,1 C. Roudaut,1 A. Saenz,2 R. Pogue,3 J. E. M. A. Grimbergen,4 L. V. B. Anderson,3
C. Beley,1 A-M Cobo,2 C. de Diego,2 B. Eymard,5 P. Gallano,6 H. B. Ginjaar,4 A. Lasa,6
C. Pollitt,3 H. Topaloglu,7 J. A. Urtizberea,5 M. de Visser,8 A. van der Kooi,8 K. Bushby,3
E. Bakker,4 A. Lopez de Munain,2 M. Fardeau,5 and J. S. Beckmann1

1URA 1922 CNRS, Généthon, Evry, France; 2Department of Neurology and Experimental Unit, Hospital Ntra. Sra. Aranzazu, San Sebastian,
Spain; 3School of Biochemistry and Genetics, University of Newcastle upon Tyne, Newcastle upon Tyne; 4MGC-Department of Human and
Clinical Genetics, Leiden University Medical Center, Leiden; 5Institut de Myologie, INSERM U. 153, Hopital Salpétrière, Paris; 6Hospital Sant
Pau, Unitat Genetica Molecular, Barcelona; 7Medical Biology Department, Hacettepe University, Ankara; and 8Amsterdam Medical Center,
Amsterdam

Summary

Limb-girdle muscular dystrophy type 2A (LGMD2A) is
an autosomal recessive disorder characterized mainly by
symmetrical and selective atrophy of the proximal limb
muscles. It derives from defects in the human CAPN3
gene, which encodes the skeletal muscle–specific member
of the calpain family. This report represents a compi-
lation of the mutations and variants identified so far in
this gene. To date, 97 distinct pathogenic calpain 3 mu-
tations have been identified (4 nonsense mutations, 32
deletions/insertions, 8 splice-site mutations, and 53 mis-
sense mutations), 56 of which have not been described
previously, together with 12 polymorphisms and 5 non-
classified variants. The mutations are distributed along
the entire length of the CAPN3 gene. Thus far, most
mutations identified represent private variants, although
particular mutations have been found more frequently.
Knowledge of the mutation spectrum occurring in the
CAPN3 gene may contribute significantly to structure/
function and pathogenesis studies. It may also help in
the design of efficient mutation-screening strategies for
calpainopathies.

Introduction

Limb-girdle muscular dystrophy type 2A (LGMD2A [
MIM 253600]) is an autosomal recessive disorder char-
acterized by symmetrical and selective atrophy of the
pelvic, scapular, and trunk muscles; elevated serum cre-

Received January 25, 1999; accepted for publication April 14, 1999;
electronically published May 7, 1999.

Address for correspondence and reprints: Dr. Jacques S. Beckmann,
Généthon, 1 rue de l’internationale, 91000 Evry, France. E-mail:
beckmann@genethon.fr

q 1999 by The American Society of Human Genetics. All rights reserved.
0002-9297/99/6406-0006$02.00

atine kinase; and a necrotic regeneration pattern on mus-
cular biopsies. Classically, calf hypertrophy is rare and
always discrete, and there is no mental, cardiac, or facial
disturbance. In the vast majority of cases, the symptoms
arise in childhood, progress gradually, and, 10–20 years
after onset, patients are frequently unable to walk and
are confined to a wheelchair (Beckmann and Fardeau
1998). A limb-girdle muscle involvement can also be
recognized in a number of muscular disorders—in par-
ticular, in the other progressive muscular dystrophies
that also have been regrouped under the term
“LGMD2”; the latter comprise the four sarcoglycano-
pathies (LGMD2C–LGMD2F; (Roberds et al. 1994;
Bönnemann et al. 1995; Lim et al. 1995; Noguchi et al.
1995; Nigro et al. 1996), which are characterized by a
disruption of the sarcoglycan complex (Campbell 1995),
and three additional LGMD2 entities (LGMD2B,
LGMD2G, and LGMD2H [Bashir et al. 1994; Moreira
et al. 1997; Weiler et al. 1998]), which present no al-
terations in the sarcoglycan complex. The causative gene
for LGMD2B recently has been identified as coding for
dysferlin (Bashir et al. 1998; Liu et al. 1998).

The gene responsible for LGMD2A had been local-
ized, by linkage and linkage-disequilibrium analyses, to
chromosome 15q15.1-15.3, within a 1-cM region al-
most entirely covered by a 1.6-Mb YAC (Beckmann et
al. 1991; Fougerousse et al. 1994; Allamand et al. 1995).
Characterization of genes from this region led eventually
to the identification of the LGMD2A gene that encodes
calpain 3, a member of the calpain family (Chiannilk-
ulchai et al. 1995; Richard et al. 1995). The latter re-
groups nonlysosomal calcium-dependent cysteine pro-
teases, whose precise functional roles remain elusive.

The human calpain 3 gene (CAPN3) covers a genomic
region of ∼40 kb (Richard et al. 1995). It is expressed
predominantly in the skeletal muscle tissue, as a 3.5-kb
transcript, driving the translation of a 94-kD protein
containing a proteolytic domain, a Ca11-activated do-
main, and two other domains with no known homology


Richard et al.: Calpain 3 Mutations and Polymorphisms 1525

Table 1

Primer Pair Used for Detection of Mutations, by Allele-Specific PCR

MUTATION

PRIMER(S)a

(5′r3′) TOUCHDOWN PCR

Uppera Lower

Temperature
Range
(7C)

Formamide
Concentration

(%)

550DA ATAAGATGACTGCCTGCCAAC (1) TTCCTGTGAGTGAGGTCTCG 65–60
TAGATGACTGCCTGCCAC (2)

G222R CTACGAAGCTCTGAAAGGTG (1) GGCTTTCTTCATGATCTTGT 60–55 2
CTACGAAGCTCTGAAAGGTA (2)

IVS621GrA CTCTGGTTACTGCTCTACAA (1) AGCACGAAAAGCAAAGATAAA 63–58
CTCTGGTTACTGCTCTACAG (2)

R489W GCCCTGATGCAGAAGAACC (1) CCAGGAGCTCTGTGGGTCA 65–60 4
GCCCTGATGCAGAAGAACT (2)

R489Q GCCCTGATGCAGAAGAACCG (1) CCAGGAGCTCTGTGGGTCA 65–60 4
GCCCTGATGCAGAAGAACCA (2)

R572W AGGGGGAATTCATCCTCCG (1) TTCAACCTCTGGGAGTGGGCC 60–55 2
AGGGGGAATTCATCCTCTG (2)

R572Q AGGGGGAATTCATCCTCCG (1) TTCAACCTCTGGGAGTGGGCC 60–55 2
AGGGGGAATTCATCCTCCA (2)

S744G AAGAATGGGGTTGATTTGGAG GCATTTCGCATCTCGTAGCT (1) 65–60 2
GCATTTCGCATCTCGTAGCC (2)

R748Q AAGAATGGGGTTGATTTGGAGA TGCGTCGTTGACTGCATTTC (1) 60–55
TGCGTCGTTGACTGCATTTT (2)

R769Q AAGAATGGGGTTGATTTGGAGA ATGTGTTTGTCTGCGTACC (1) 65–60
ATGTGTTTGTCTGCGTACT (2)

1872CrT CACCAGAGCAAACCGTCCAC CAGGGCTTGTTTTGCCTTTG (1) 63–58 )
CAGGGCTTGTTTTGCCTTTA (2)

2069DCA CAAGGACCTGAAGACACACG (1) ACCCTGTATGTTGCCTTGG 65–60 2
CACAAGGACCTGAAGACACG (2)

2362AGrTCATCT CATCTGCTGCTTCGTTAG (1) GGAGATTATCAGGTGAGATGCC 60–55
TGCTGCTTCGTTTCATCT (2)

a For each mutation, two sequences are used—one corresponding to the normal allele (1) and the other corresponding to the mutated allele
(2).

(Sorimachi et al. 1989). In addition, CAPN3 has three
unique sequences—NS, IS1, and IS2. Additional tran-
scripts resulting from alternative splicing and alternative
promoter use have been characterized (Ma et al. 1998a,
1998b; Herasse et al., in press; F. Fougerousse, personal
communication).

Since the discovery of the molecular defect in
LGMD2A, we have collectively undertaken studies to
determine the nature, number, and distribution of
CAPN3 mutations and to assess potential genotype-phe-
notype correlations. To date, 97 distinct mutations have
been identified, 56 of which have not been described
previously.

Subjects and Methods

Description of Families and Preparation of Samples
from Patients

This report summarizes studies, performed in five lab-
oratories (Généthon [France], Hospital Ntra. Sra. Ar-
anzazu [Spain], Hospital Sant Pau [Spain], the DNA di-
agnostic laboratory [Netherlands], and University of

Newcastle upon Tyne [United Kingdom), of families
from diverse ethnic or geographic origins (including Bra-
zil, Bulgaria, Canada, France, Greece, Israel, Italy, Japan,
Lebanon, the Netherlands, Poland, Portugal, Russia,
Spain, Switzerland, Turkey, United Kingdom, United
States, and Vietnam). Families with LGMD that were
from of the U.S. Amish community, from Réunion Is-
land, and Basque country were distinguished from, re-
spectively, families in the remaining United States, Met-
ropolitan French (i.e., continental France), and Spanish
families, since each of them represents a particular ge-
netic isolate with a predominant founder effect. All pa-
tients presented in this report fulfill the clinical criteria
of a classic LGMD phenotype—namely, autosomal re-
cessive inheritance, progression of muscle weakness in
a limb-girdle distribution (with sparing of facial mus-
cles), and a clearly dystrophic muscle biopsy (Fardeau
et al. 1996a, 1996b). Calpain 3–deficient patients were
also ascertained by the Newcastle group, by demon-
stration of both reduced 94-kD-band labeling and pre-
served sarcoglycans and dystrophin on western blots of
skeletal-muscle biopsies (Anderson et al. 1998).


Table 2

Mutations in the CAPN3 Gene

Exon/Intron and Origin

No. of Families
(No. of Homozygous

Patients)
Positiona

(bp) Mutationb
Position in

Amino Acidc Mutation Reference

1:
Spain 1 10 GTCrATCd 4 V4I Hospital Sant Pau
Turkey 1 (1) 19–23 DGCATC 6–7 19–23DGCATC Dinçer et al. (1997), Richard et al. (1997)
Réunion Island 1 43 DG 15 43DG Généthon
Spain 1 60 DA 20 60DA Hospital Sant Pau
Italy 1 77 CCGrCTGd 26 P26L Généthon
France 1 224–225 insT 76 224–225insT Généthon
France 1 229 GACrAAC 77 D77N Généthon
Lebanon 1 (1) 257 TCTrTTT 86 S86F Richard et al. (1997)
France 1 (1) 277–300 24-bp change 93–100 DFPIQFVWK Généthon

2:
Brazil 1 (1) 328 CGArTGAd 110 R110X Richard et al. (1995)
Bulgaria 1 352 AGArGGA 118 R118G Généthon

3:
United States 1 402 DC 134 402DC Richard et al. (1997)
France 1 (1) 409 TGCrCGC 137 C137R Généthon
Polish 1 418–419 insC 140 418–419insC MGC Leiden
Vietnam 1 424 CAGrTAG 142 Q142X Hospital Ntra. Sra. Aranzazu
France 1 484 ATCrCTC 162 I162L Généthon

4:
France 1 545 CTGrCAG 182 L182Q Richard et al. (1995)
France 1 548 CCArCTA 183 P183L Généthon
France 5 550 DA 184 550DA Richard et al. (1995, 1997)
Greece 1 University of Newcastle
Italy 1 (1) Richard et al. (1997)
Netherlands 2 MGC Leiden
Russia 1 (1) Généthon
Turkey 5 (5) Dinçer et al. (1997)
UK 1 University of Newcastle
Réunion Island 1 551 ACGrATGd 184 T184M Généthon
United Kingdom 1 566 CTGrCCG !189 L189P University of Newcastle
Bulgaria 1 598–612 15-bp change 200–204 DFWSAL Généthon
Germany 1 Häffner et al. (1998)


IV
S4

2
7

:
B

as
q

u
e

1
5

,3
2

8
-b

p
ch

an
ge

IV
S4

1
4

0
4
D

5
3

2
8

G
én

ét
h

o
n

5
:

N
et

h
er

la
n

d
s

1
6

4
0

G
G

T
r

A
G

T
2

1
4

G
2

1
4

S
M

G
C

L
ei

d
en

F
ra

n
ce

1
6

4
3

T
C

C
r

C
C

C
2

1
5

S2
1

5
P

R
ic

h
ar

d
et

al
.

(1
9

9
7

)
U

n
it

ed
St

at
es

1
6

4
3

–6
6

3
2

1
-b

p
ch

an
ge

2
1

5
–2

2
1

D
SY

E
A

L
K

G
R

ic
h

ar
d

et
al

.
(1

9
9

7
)

U
n

it
ed

K
in

gd
o

m
1

6
4

9
G

A
A

r
A

A
A

2
1

7
E

2
1

7
K

U
n

iv
er

si
ty

o
f

N
ew

ca
st

le
B

as
q

u
e

4
6

6
4

G
G

G
r

A
G

G
2

2
2

G
2

2
2

R
U

rt
as

u
n

et
al

.
(1

9
9

8
)

Sp
ai

n
1

(1
)

H
o

sp
it

al
Sa

n
t

P
au

Sp
ai

n
1

6
7

6
G

A
G

r
A

A
G

2
2

6
E

2
2

6
K

H
o

sp
it

al
N

at
r.

Sr
a.

A
ra

n
za

zu
F

ra
n

ce
1

6
9

5
A

C
A

r
A

T
A

2
3

2
T

2
3

2
I

G
én

ét
h

o
n

F
ra

n
ce

2
7

0
1

G
G

G
r

G
A

G
2

3
4

G
2

3
4

E
R

ic
h

ar
d

et
al

.
(1

9
9

5
)

L
eb

an
o

n
1

(1
)

7
1

7
D

T
2

3
9

7
1

7
D

T
R

ic
h

ar
d

et
al

.
(1

9
9

7
)

F
ra

n
ce

2
7

5
9

–7
6

1
D

G
A

A
2

5
3

–2
5

4
D

K
2

5
4

G
én

ét
h

o
n

IV
S5

:
G

er
m

an
y

1
8

0
1
1

1
G

T
r

A
T

)
IV

S5
1

1
G

r
A

H
äf

fn
er

et
al

.
(1

9
9

8
)

6
:

F
ra

n
ce

2
8

8
7

D
A

2
9

6
8

8
6
D

A
G

én
ét

h
o

n
F

ra
n

ce
1

9
4

5
D

G
3

1
5

9
4

5
D

G
R

ic
h

ar
d

et
al

.
(1

9
9

5
)

IV
S6

:
R

éu
n

io
n

Is
la

n
d

1
6

(9
)

9
4

6
2

1
A

G
r

A
A

)
IV

S6
2

1
G

r
A

R
ic

h
ar

d
et

al
.

(1
9

9
5

),
P

en
is

so
n

-B
es

n
ie

r
et

al
.

(1
9

9
8

)
7

:
L

eb
an

o
n

1
9

5
6

C
C

G
r

C
T

G
d

3
1

9
P

3
1

9
L

R
ic

h
ar

d
et

al
.

(1
9

9
7

)
F

ra
n

ce
1

1
0

0
2

C
A

C
r

C
A

G
3

3
4

H
3

3
4

Q
R

ic
h

ar
d

et
al

.
(1

9
9

7
)

T
u

rk
ey

1
(1

)
1

0
0

6
T

A
C

r
A

A
C

3
3

6
Y

3
3

6
N

D
in

çe
r

et
al

.
(1

9
9

7
)

8
:

F
ra

n
ce

1
(1

)
1

0
6

1
G

T
G

r
G

G
G

3
5

4
V

3
5

4
G

R
ic

h
ar

d
et

al
.

(1
9

9
5

)
F

ra
n

ce
1

1
0

7
9

T
G

G
r

T
A

G
3

6
0

W
3

6
0

X
R

ic
h

ar
d

et
al

.
(1

9
9

5
)

Ja
p

an
2

(2
)

1
0

8
0

T
G

G
r

T
G

C
3

6
0

W
3

6
0

C
K

aw
ai

et
al

.
(1

9
9

8
)

IV
S9

:
F

ra
n

ce
1

1
1

9
4
2

9
A

r
G

)
IV

S9
2

9
A

r
G

G
én

ét
h

o
n

1
0

: S
p

ai
n

1
1

2
9

2
–1

2
9

3
in

sT
4

3
1

1
2

9
2

–1
2

9
3

in
sT

H
o

sp
it

al
N

at
r.

Sr
a.

A
ra

n
za

zu
F

ra
n

ce
1

1
3

0
9

C
G

C
r

T
G

C
d

4
3

7
R

4
3

7
C

H
o

sp
it

al
N

at
r.

Sr
a.

A
ra

n
za

zu
Is

ra
el

1
(1

)
1

3
1

8
C

G
G

r
T

G
G

d
4

4
0

R
4

4
0

W
R

ic
h

ar
d

et
al

.
(1

9
9

7
)

Sp
ai

n
1

1
3

1
9

–1
3

2
2

D
G

G
G

G
4

4
0

–4
4

1
1

3
1

9
D

G
G

G
G

H
o

sp
it

al
N

at
r.

Sr
a.

A
ra

n
za

zu
F

ra
n

ce
1

1
3

2
2

G
G

T
r

G
A

T
4

4
1

G
4

4
1

D
G

én
ét

h
o

n
N

et
h

er
la

n
d

s
1

1
3

3
3

G
G

A
r

A
G

A
d

4
4

5
G

4
4

5
R

M
G

C
L

ei
d

en
N

et
h

er
la

n
d

s
1

1
3

4
2

C
G

C
r

T
G

C
d

4
4

8
R

4
4

8
C

M
G

C
L

ei
d

en
Sp

ai
n

1
H

o
sp

it
al

N
at

r.
Sr

a.
A

ra
n

za
zu

F
ra

n
ce

2
(1

)
1

3
4

2
C

G
C

r
G

G
C

4
4

8
R

4
4

8
G

G
én

ét
h

o
n

F
ra

n
ce

1
1

3
4

3
C

G
C

r
C

A
C

d
4

4
8

R
4

4
8

H
G

én
ét

h
o

n
U

n
it

ed
St

at
es

1
G

én
ét

h
o

n

(c
o

n
ti

n
u

ed
)


Table 2 (continued)

Exon/Intron and Origin

No. of Families
(No. of Homozygous

Patients)
Positiona

(bp) Mutationb
Position in

Amino Acidc Mutation Reference

11:
United Kingdom 1 1373 DC 458 1373DC University of Newcastle
Spain 1 1435 AGCrGGC 479 S479G Hospital Natr. Sra. Aranzazu
United Kingdom 1 University of Newcastle
Basque 1 1456 CAGrGAG 486 Q486E Urtasun et al. (1998)
Basque 2 1465 CGGrTGGd 489 R489W Urtasun et al. (1998)
Réunion Island 2 1466 CGGrCAGd 489 R489Q Généthon
United States 1 1468 CGGrTGGd 490 R490W Richard et al. (1997)
France 1 Richard et al. (1995)
France 5 (2) 1469 CGGrCAGd 490 R490Q Généthon
Turkey 1 Dinçer et al. (1997)
France 1 1477 CGGrTGGd 493 R493W Généthon
Spain 1 Hospiatl Sant Pau
Italy 1 1486 GGGrAGG 496 G496R Richard et al. (1997)
Netherlands 1 1505 ATTrACT 502 I502T MGC Leiden

13:
France 1 1611 TACrTAA 537 Y537X Généthon
Turkey 1 (1) Dinçer et al. (1997), Richard et al. (1997)
France 1 1622 CGGrCAGd 541 R541Q Généthon
Switzerland 1 (1) 1699 GGGrTGG 567 G567W Richard et al. (1997)
France 2 (1) 1714 CGGrTGGd 572 R572W Richard et al. (1997)
Réunion Island 1 (1) 1715 CGGrCAGd 572 R572Q Richard et al. (1995)
France 1 Généthon
Europe 1 1743/1744 DTG 581–582 1743DTG Hospital Natr. Sra. Aranzazu

15:
Spain 1 1785/1788 DAAAG 595–596 1785DAAAG Hospital Sant Pau
Japan 1 (1) 1795/1796 insA 599 1795–1796insA Kawai et al. (1998)

16:
Italy 1 1817 TCGrTTGd 606 S606L Richard et al. (1997)
Europe 1 Hospital Natr. Sra. Aranzazu
France 2 1838 DA 613 1838DA Généthon
France 1 1865 DAG 622 1865DAG Généthon
Réunion Island 1 (1) 1872 GGCrGGTd 624 1872 CrT Richard and Beckmann (1995)
Netherlands 1 1913 CAGrCCG 638 Q638P MGC Leiden

17:
Bulgaria 1 1981/1984 DATAG 661–662 1981DATAG Généthon
France 1 1981 DA 661 1981DA Généthon
United Kingdom 1 1983 DA 661 1983DA University of Newcastle


IV
S1

7
:

Sp
ai

n
2

1
9

9
2
1

1
G

T
r

A
T

)
IV

S1
7
1

1
G

r
T

H
o

sp
it

al
N

at
r.

Sr
a.

A
ra

n
za

zu
U

n
it

ed
K

in
gd

o
m

1
1

9
9

3
2

1
A

G
r

A
T

)
IV

S1
7
2

1
G

r
T

U
n

iv
er

si
ty

o
f

N
ew

ca
st

le
1

9
: R

éu
n

io
n

Is
la

n
d

1
2

0
6

9
/2

0
7

0
D

A
C

6
9

0
2

0
6

9
D

C
A

R
ic

h
ar

d
et

al
.

(1
9

9
5

)
U

n
it

ed
K

in
gd

o
m

1
2

0
9

3
C

G
T
r

C
C

T
6

9
8

R
6

9
8

P
U

n
iv

er
si

ty
o

f
N

ew
ca

st
le

T
u

rk
ey

1
(1

)
2

1
0

5
G

C
G

r
G

T
G

d
7

0
2

A
7

0
2

V
D

in
çe

r
et

al
.

(1
9

9
7

),
R

ic
h

ar
d

et
al

.
(1

9
9

7
)

F
ra

n
ce

1
(1

)
G

én
ét

h
o

n
F

ra
n

ce
1

2
1

1
3

G
A

T
r

C
A

T
7

0
5

D
7

0
5

H
H

o
sp

it
al

N
at

r.
Sr

a.
A

ra
n

za
zu

G
re

ec
e

1
2

1
1

4
G

A
T
r

G
C

T
7

0
5

D
7

0
5

G
U

n
iv

er
si

ty
o

f
N

ew
ca

st
le

IV
S2

0
2

ex
o

n
2

1
:

Sp
ai

n
1

(2
1

8
5
2

1
2

)/
2

1
9

4
2

2
-b

p
ch

an
ge

)
V

(I
V

S2
0
2

1
2

)/
2

1
9

4
D

H
o

sp
it

al
N

at
r.

Sr
a.

A
ra

n
za

zu
IV

S2
0

:
F

ra
n

ce
1

2
1

8
5
2

2
A

G
r

G
G

)
IV

S2
0
2

2
A

G
r

G
G

G
én

ét
h

o
n

C
an

ad
a

1
(1

)
G

én
ét

h
o

n
2

1
: F

ra
n

ce
1

2
1

9
2

T
T

C
r

T
C

C
7

3
1

F
7

3
1

S
G

én
ét

h
o

n
R

éu
n

io
n

Is
la

n
d

3
2

2
3

0
A

G
C

r
G

G
C

7
4

4
S7

4
4

G
R

ic
h

ar
d

et
al

.
(1

9
9

5
),

P
en

is
so

n
-B

es
n

ie
r

et
al

.
(1

9
9

8
)

B
as

q
u

e
3

2
2

4
3

C
G

A
r

C
A

A
d

7
4

8
R

7
4

8
Q

U
rt

as
u

n
et

al
.

(1
9

9
8

)
Sp

ai
n

4
(2

)
H

o
sp

it
al

Sa
n

t
P

au
,

H
o

sp
it

al
N

tr
a.

Sr
a.

A
ra

n
za

zu
T

u
rk

ey
1

(1
)

D
in

çe
r

et
al

.
(1

9
9

7
),

R
ic

h
ar

d
et

al
.

(1
9

9
7

)
IV

S2
1

:
Sp

ai
n

1
2

2
6

3
1

2
G

T
r

G
A

)
IV

S2
1
1

2
T
r

A
H

o
sp

it
al

N
at

r.
Sr

a.
A

ra
n

za
zu

2
2

: U
n

it
ed

St
at

es
(A

m
is

h
)

1
9

(1
9

)
2

3
0

6
C

G
G

r
C

A
G

d
7

6
9

R
7

6
9

Q
R

ic
h

ar
d

et
al

.
(1

9
9

5
)

B
ra

zi
l

1
(1

)
R

ic
h

ar
d

et
al

.
(1

9
9

5
)

F
ra

n
ce

2
R

ic
h

ar
d

et
al

.
(1

9
9

5
)

B
ra

zi
l

1
2

3
1

3
–2

3
1

6
D

A
G

A
C

7
7

1
–7

7
2

2
3

1
3
D

A
G

A
C

R
ic

h
ar

d
et

al
.

(1
9

9
5

)
F

ra
n

ce
1

2
3

1
7

–2
3

2
1

A
A

A
C

A
r

T
7

7
3

–7
7

4
2

3
1

7
A

A
A

C
A

r
T

G
én

ét
h

o
n

N
et

h
er

la
n

d
s

1
2

3
1

9
–2

3
2

0
in

sA
7

7
4

2
3

1
9

–2
3

2
0

in
sA

M
G

C
L

ei
d

en
B

ra
zi

l
1

2
3

6
2

–2
3

6
3

A
G

r
T

C
A

T
C

T
7

8
8

2
3

6
2

A
G

r
T

C
A

T
C

T
R

ic
h

ar
d

et
al

.
(1

9
9

7
)

R
éu

n
io

n
Is

la
n

d
3

T
C

A
T

C
T

R
ic

h
ar

d
et

al
.

(1
9

9
7

)
B

as
q

u
e

3
9

(3
1

)
U

rt
as

u
n

et
al

.
(1

9
9

8
)

Sp
ai

n
8

(3
)

H
o

sp
it

al
Sa

n
t

P
au

,
H

o
sp

it
al

N
at

r.
Sr

a.
A

ra
n

za
zu

U
n

it
ed

St
at

es
2

R
ic

h
ar

d
et

al
.

(1
9

9
7

)

a
N

u
m

b
er

ed
o

n
th

e
b

as
is

o
f

th
e

cD
N

A
(E

M
B

L
ac

ce
ss

io
n

n
u

m
b

er
8

5
0

3
0

),
st

ar
ti

n
g

fr
o

m
A

T
G

.
b

M
u

ta
te

d
n

u
cl

eo
ti

d
es

ar
e

u
n

d
er

li
n

ed
.

c
N

u
m

b
er

ed
o

n
th

e
b

as
is

o
f

th
e

p
ro

te
in

se
q

u
en

ce
s,

st
ar

ti
n

g
fr

o
m

th
e

fi
rs

t
m

et
h

io
n

in
e

re
si

d
u

e.
d

M
o

d
ifi

ed
C

p
G

si
te

.


1530 Am. J. Hum. Genet. 64:1524–1540, 1999

Table 3

Nature and Frequency of the Polymorphisms Identified in the CAPN3 Gene

Location Name
Position

(bp) Event CpG? Frequencya

Promoter 2408TrC 2408 TrC No ND
Exon 1 T3T 9 ACCrACT No !.01
Exon 1 T32T 96 ACTrACC No .12
Exon 2 C106C 318 TGCrTGT Yes .04
Exon 2 E107K 319 GAGrAAG Yes .05
Exon 3 F165F 495 TTCrTTT No .01
Exon 3 Q166Q 498 CAGrCAA No !.01
Exon 5 A236T 708 GCArACA No .10
intron 9 IVS9226CrG 1194226 CrG No .07
Exon 13 I556I 1668 ATCrATT Yes !.01
Exon 18 L673L 2019 CTCrGTC No !.01
Intron 22 2380112DA 2380112 DA No .10

a Based on study of 1100 independent chromosomes; ND 5 not determined.

Altogether, 180 families were selected on the basis of
identification of variation in the CAPN3 gene. Several
series of control chromosomes were examined, including
those of the parents of patients, those of CEPH individ-
uals, and/or those of control individuals from matched
geographic origins. Genomic DNA was prepared from
either peripheral-blood lymphocytes or cultures of lym-
phoblastoid cell lines.

Identification of DNA Sequence Variants

There was no uniform strategy for variant identifi-
cation, since different protocols were used in each of the
centers involved. These protocols included the following:

1. Illegitimate transcription analyses of cDNA—These
were performed as in the study by Richard et al. (1995),
by use of total cellular RNA extracted from lympho-
blastoid cell lines, followed by a final characterization
by DNA sequencing.

2. Heteroduplex analysis—Ten microliters of the PCR
product were denatured at 94 ˚C for 5 min and were
allowed to renature at room temperature for >1 h. The
product was electrophoresed on Hydrolink gel prepared
according to the manufacturer’s recommendations. For
PCR products of ∼150 bp, the conditions of migration
are an overnight (∼16 h) run at 400 V. After the gel was
stained by ethidium bromide, the products are visualized
under UV light. Identification of the variant sequences
was performed by DNA sequencing.

3. SSCP—Five microliters of the PCR products were
denatured at 957C for 5 min and cooled on ice, with 10
ml of loading buffer (95% formamide and 0.1% brom-
ophenol blue). The samples were separated on 0.4-mm-
thick 6% acrylamide 5% glycerol gels in 0.5 #

. Gels were run at room temperature,Tris-borate EDTA
at 8 W for 16 h. The samples were transferred to nylon
filters and hybridized with the primers used in the PCR,

according to the ECL protocol (Amersham), as described
by Vignal et al. (1993). Identification of the variant se-
quences was performed by DNA sequencing.

4. Direct sequencing—One hundred nanograms of
DNA were amplified under buffer and cycle conditions
described by Richard et al. (1995). Each exon was am-
plified by specific primers, chosen in introns. The prod-
ucts of PCR were directly sequenced, with the same
primers, by dye-dideoxy sequencing, after purification
through either Microcon devices (Amicon) or polyacryl-
amide gel Biogel P-25 (Bio-Rad).

5. Allele-specific PCR—Oligonucleotides for muta-
tions 551DA, G222R, IVS621GrA, R489W, R489Q,
R572W, R572Q, S744G, R748Q, R769Q, 1872CrT,
2069DCA, and 2362AGrTCATCT were designed such
that their 3′ bases were at the point of the mutation (table
1). One hundred nanograms of DNA were amplified by
touchdown PCR in a 50-ml volume containing 1 mM
Tris-HCl, pH 8.8, 50 mM KCl, 1.5 mM MgCl2, 0.1%
Triton X-100, 200 mM of each dNTP, 100 ng of each
primer, and 2 units of Taq polymerase (Perkin-Elmer).
It was necessary to add various amounts of formamide,
to increase the specificity of some PCRs. After 5-min
denaturation at 967C, thermocycling was performed as
follows: 947C for 40 s and then 30-s annealing steps
starting 57C higher than the annealing temperature and
decreasing 17C every two cycles until the annealing tem-
perature was reached. These steps were followed by 20
cycles at 947C for 40 s and 30 s at the annealing tem-
perature. After PCR, the products were electrophoresed
on a 2% horizontal agarose gel stained with ethidium
bromide.

Haplotype Analysis

Short tandem repeats covering a region of 1 cM at
the LGMD2A locus at 15q15.2 (markers D15S514,
D15S779, D15S782, D15S780, and D15S778) were


Richard et al.: Calpain 3 Mutations and Polymorphisms 1531

Figure 1 Distribution of mutations along the CAPN3 transcript
and protein. The mutations are represented, according to their type,
along the schematic representation of the CAPN3 gene and protein.
The number of independent events is given either inside or above the
symbol of the mutation. The 24 exons are indicated by unblackened
boxes (with the exon numbers below them). The four domains (I–IV)
and the muscle-specific sequences (NS, IS1, and IS2) of calpain 3 also
are shown.

Table 4

Unclassified Variants

Exon/Intron Origin
No. of

Families
Position

(bp) Event
Amino Acid

Position
Putative
Effect

7 Spain, Netherlands 2 984 TGCrTGTa C328 Splicing?
10 Switzerland 1 (1) 1263 CTGrCTAa L421 Splicing?
IVS12 Japan 1 (1) 1525–1534 DC ) Splicing?
22 Turkey 1 2320 CACrGAC H774 HrD
23 United Kingdom 1 2393 GCArGAA A798 ArE

NOTE.—Data are as defined in the footnotes totable 2.
a Modified CpG site.

used to assign haplotypes, as described in a previous
publication (Richard et al. 1997).

Results

Altogether, 114 variants have been detected in the
CAPN3 gene (table 2). Ninety-seven mutations were
identified after screening of the CAPN3 gene in a total
of 180 families from 22 different ethnic or geographic
origins. Ninety-six mutations were characterized in our
laboratories, and the additional one (IVS511GrA) was
previously reported as having occurred in a German
family (Häffner et al. 1998). In most families, both mu-
tations were identified, with the exception of 24 families
for which only one allele each has been uncovered.
Twelve polymorphisms were uncovered in the course of
these studies (table 3). Five additional variants (table 4),
found only in patients, were not demonstrated to be
causative mutations and therefore could not be classified
as belonging to either of the two formerly described
groups.

Mutations Identified

A summary of all the CAPN3 mutations is presented
intable 2 and figure 1. The names of the mutations fol-
low published guidelines (Antonarakis 1998). The mu-
tations include 65 single-base-pair substitutions—which
resulted in a change of amino acid in 53 cases, in pre-
mature stop codons in 4 cases, and in splicing defects
in 8 cases. Twenty-three (35%) of these 65 point mu-
tations affect a CpG dinucleotide. An additional 32 mu-
tations were small insertions and/or deletions. Therefore,
∼45% (44/97) of the CAPN3 mutations are likely to
result in a truncated protein and, therefore, are likely to
be inactivating. For the missense mutations, the follow-
ing arguments are used to infer their pathogenic nature:
(i) the fact that they were not encountered in control
populations (in general, >100 independent chromo-
somes were examined) and (ii) the observation that all
but one of these mutations involve changes of amino
acids that are strictly conserved among the human,
mouse, rat, bovine, and porcine CAPN3 sequences, if

not among all the calpains (table 5). Furthermore, some
mutations were demonstrated to induce consequences
with regard to three biochemical characteristics of
CAPN3—namely, titin-binding ability, autolysis capac-
ity, and fodrinolysis capacity (Ono et al. 1998).

Interestingly, one of the splicing mutation (1872CrT)
changes the third base of a codon to a synonymous co-
don yet leads to both the creation of a novel internal
splice site and the subsequent loss of the end of the exon
and, eventually, to a premature in-phase stop codon (Ri-
chard and Beckmann 1995). Demonstration of the path-
ogenic nature of this sequence variant required the ex-
amination of the transcription product. The seven
remaining splice mutations are located within 5′

or 3′ splice sites. IVS511GrA was reported to induce
skipping of exon 5 (Häffner et al. 1998), and
IVS621AGrAA leads to the use of a cryptic splice site
upstream in the intron. The consequences, at the RNA
level, of the other splice-site mutations were not ex-


1532 Am. J. Hum. Genet. 64:1524–1540, 1999

Table 5

Conservation of Amino Acids at Locations of Missense Mutations and Nonclassified Variants of the CAPN3 Protein

MUTATION

AMINO ACIDS IN

CAPN3 ofaa Other Calpainsb
Calpain

Homologue(s)bHuman Mouse Rat Porcine Bovine Chicken Mammalian Nonmammalian

V4I V V V V ) V Not relevantc Not relevantc Not relevantc

P26L P P P A ) S Not relevantc Not relevantc Not relevantc

D77N D D D D ) D D D D
S86F S S S S ) S S, A A, S S, A
R118G R R R R ) R R R N, R, S
C137R C C C C ) C S, C S, N, C C, S, N
I162L I I I I ) I I I, V I
L182Q L L L L ) L L L L
P183L P P P P ) P P P P
T184M T T T T ) T T, I T, V T
L189P L L L L ) L L L L
G214S G G G G ) G G G G
S215P S S S S ) S S, C C, S C, S
E217K E E E E ) E E E, A E, Q
G222R G G G G ) G G G G, L
E226K E E E E ) E E E D, E
T232I T T T T ) T T T T
G234E G G G G ) G G G G, T
P319L P P P P P P Not relevantc Not relevantc Not relevantc

H334Q H H H H H H H H H, Y
Y336N Y Y Y Y Y Y Y Y Y
V354G V V V V V V I I, M I, V
W360C W W W W W W W W W, L
R437C R R R R ) R N, T, R, S E, Q, N, S, T E, D, N
R440W R R R R ) R R R, P L, R, P, Q
G441D G G G G ) G G G, N M, G, Q
G445R G G G G ) G G G G
R448H or G or C R R R R ) R R, L R, I Y, I, R
S479G S S S S ) S S, T S, T T
Q486E Q Q Q Q ) Q Q Q Q
R489W or Q R R R R ) R R R, I, T L, K, R
R490Q or W R R R R ) R R, Q R R
R493W R R R R ) R R, K R, K R
G496R G G G G ) G G G G
I502T I I I I ) I I I I
R541Q R R R R ) R R R R
G567W G G G G ) G G, A G, A, V G, S
R572Q or W R R R R ) R R R R
S606L S S S S ) S Not relevantc Not relevantc Not relevantc

Q638P Q Q Q K ) ) Not relevantc Not relevantc Not relevantc

R698P R R R R ) R R, K R, K, N K
A702V A A A A ) A D, N, S, H N, A, D, H D
D705H or G D D D D ) D D, E D, Q, T D
F731S F F F F ) F F, Y F, Y, I Y
S744G S S S S ) S S, A, T F, S, G, A S
R748Q R R R R ) R R R, D R
R769Q R R R R ) R R R R
H774D H H H Y ) N D, E, N, Q, K D, E Q
A798E A A A A ) A Q, A, C, S, L, I E, A, K, S Q

a Indicative of conservation among all known calpain 3 proteins; an ellipsis ()) denotes that no information is available.
b In each cell, amino acids are listed according to frequency in published sequences.
c With regard to amino acids of the three specific parts of calpain 3: NS, IS1 or IS2.

amined, because cells were not available from these pa-
tients. Nevertheless, present knowledge of splicing mech-
anisms could be used to infer putative consequences, in
accordance with the exon-definition model (Robberson

et al. 1990). IVS929ArG generates a new dinucleotide,
AG, that could potentially be used as a splicing anchor
(Smith et al. 1989); use of the latter would result
in a frameshift mutation, by addition of 8 bases


Richard et al.: Calpain 3 Mutations and Polymorphisms 1533

Table 6

Haplotype Analysis of Recurrent Mutations

MUTATION (CpG?) AND ORIGIN

HAPLOTYPEa

D15S514 D15S779 D15S782 D15S780 D15S778

G234E:
France ( )n 5 1 204 96 128 2 179
France ( )n 5 1 200 96 130 5 177

R490Q (yes):
France ( )n 5 2 202/206 96 130/128 3 181/187
France ( )n 5 3 202 82/94 122 4/5 183/179
Turkey ( )n 5 1 210 86 128 5 181

R572W (yes):
France ( )n 5 1 202 96 120 3/5 177
France ( )n 5 1 210 96 132 5 179

R572Q (yes):
Réunion Island ( )n 5 1 198 92 130 5 181
France ( )n 5 1 210 84 130 3 181

IVS2022AGrGG:
France ( )n 5 1 196/210 96/98 128/132 2 181
Canada ( )n 5 1 200 86 116 3 177

R748Q (yes):
Spain ( )n 5 2 202 90 128 3 183
Spain ( )n 5 2 210 84/96 130/128 3 181
Turkey ( )n 5 1 210 82 126 5 181

R769Q (yes):
Amish ( ), France ( )n 5 19 n 5 2 210/202 96 122 4 181
Brazil ( )n 5 1 202/210 84 126 4 181

a Only haplotypes corresponding to recurrent mutations are presented. Markers are listed ac-
cording to their linear map order. The disease locus is located between markers D15S779 and
D15S782. In situations in which the haplotype could not be reconstructed, the two alleles have been
given, separated by a virgule (/). For all markers except D15S780, the numbers represent allele size
(in bp); in the case of marker D15S780, the numbers represent allele designations as given in the
CEPH database.

in the mRNA. IVS1711GrT, IVS1721GrT, and
IVS2021AGrGG may lead to either single-exon skip-
ping or cryptic–splice-site utilization.

Most of the deletion/insertion mutations caused a
frameshift, leading to a secondary premature stop co-
don. Four of the mutations (DFPIQFVW, DFWSAL,
DSYEALKG, and DK254), however, were in-frame de-
letions that would theoretically result only in limited
amino acid deletions. Two mutations were slightly
more complex and resulted in the combination of
small insertions and deletions. 2317AAACArT consists
of a 5-bp deletion of nucleotides 2317–2321 of the cod-
ing region, coupled with a 1-bp insertion, and
2362AGrTCATCT consists of a 2-bp deletion coupled
with a 6-bp insertion. The net result in both cases is a
shift in the reading frame, followed by a premature stop
codon.

The first major genomic deletion in the calpain 3 gene
was identified in a Basque patient. Reverse transcrip-
tion–PCR of the illegitimate transcription products
showed that exons 5–7 were missing. Since no sequence
alteration was identified in the splice sites of exons 4–8,
a genomic deletion was suspected. Two PCR prim-
ers—one located in exon 4 and the other located in exon
8—that, in normal genomic DNA, were separated by

∼7 kb, led to the amplification of a fragment of ∼1.5
kb in DNA from this patient, whereas no amplification
was obtained in DNA from control individuals. Se-
quencing of the corresponding PCR product enabled us
to identify the mutation event as a 5,328-bp deletion
spanning exons 5–7. The 5′ breakpoint is in the right
arm of the 3′r5′-oriented Alu-Jb repeat in intron 4, and
the 3′ breakpoint maps to the left arm of the Alu-Sg
repeat in intron 7, which is oriented in the same direc-
tion. The deletion therefore seems to result from an un-
equal recombination event between two adjacent Alu
elements.

Polymorphisms

To date, 12 polymorphisms have been characterized
in the CAPN3 gene (table 3). They were excluded from
being LGMD2A mutations, because of their occurrence
in normal healthy populations and/or occurrence along
the same chromosome that has a mutation. The range
of frequencies of these polymorphisms varies between
!0.01% and 12% (table 3). None of these polymor-
phisms is in linkage disequilibrium with a particular mu-
tation. Three of the variants are due to CrT transitions.
Three polymorphisms map in noncoding regions. Seven


1534 Am. J. Hum. Genet. 64:1524–1540, 1999

Table 7

Detection of Variants by Modification of Restriction
Enzymes

MUTATION

ENZYMATIC SITE(S)a

Normal Allele Variant Allele

19DGCATC HhaI, HinPI, SfaNI AluI
43DG AlwNI, NspBII HaeII, Eco47III
60DA Ava I
P26L BcnI, NciI
T32T DdeI SecI
D77N AvaII
S86F MnlI
I162L AlwI, BstYI BanII, Bsp1286I
T184M MaeII
S215P AvaII, Sau96I
DK254 MboI
945DG NciI, BcnI AvaIb

P319L MspI, HpaII
H334Q StuI
A236T HphI
W360X NlaIV, SecI
W360C NlaIV, SecI EaeI, GdiII
IVS9226CrG Tth111II
L421L PstI
1292insT Tth111I
R437C NspII, HaeIII
R440W RsaI
Q486E SfaNI
R489W Cfr10I
R489Q Cfr10I NspBII
R493W AvaII
G496R MaeI DdeI
R572W BcnI, NciI
R572Q BcnI, NciI
1838DA AlwNI
1981DATAG BbvI, Fnu4HI
2000DA BglII
A702V HinPI, HhaI HgiaI
F731S XmnI
S744G AluI
2317AAACArT Tth111II BspHI, PleI
A798E NsiI BspHI

a Enzymatic sites corresponding to enzymes that cut too
frequently to be used in an analysis are not given.

b Present only on genomic DNA, not on cDNA.

Figure 2 Evolution curves for LGMD2A patients. Functional
stages are graded as I–VII, according to the Gardner-Medwin and
Walton (1974) scale. Only the curves for patients for whom two points
of the disease evolution are known are represented, whereas the cal-
culation of means and duration takes into account all available data.

exonic polymorphisms do not lead to an alteration of
the encoded amino acid, and the majority of them in-
volve the third base of a codon. The last two polymor-
phisms, E107K and A236T, which are present, respec-
tively, in exons 2 and 5, lead to an amino acid
substitution. It should be noted, however, that the minor
variants correspond to an amino acid that can also be
found at the equivalent positions in other members of
the calpain family: for E107, a lysine is present at the
equivalent position in the human nCL4 protein, and, for
A236, a threonine is present at the corresponding po-
sition in rat, mouse, and porcine calpain 3, as well as
in the human, mouse, rat, and chicken calpain 1 pro-
teins. This suggests that these changes may not have
major deleterious effects on the protein’s function(s).

Unclassified Variants

Also detected were five changes that could not be clas-
sified as either a polymorphism or a deleterious muta-
tion, even though they were observed only in patients
(table 4). One intronic change, leading to a deletion of
a C in intron 11 at position 234 of exon 12, was ob-
served in a homozygous state in a Japanese patient who
was homozygous for a chromosome 15 haplotype. Two
additional changes are single-base substitutions involv-
ing the third base of a codon: a TGCrTGT transversion
at position 984 in exon 7 (C984rT [C328C]) and a
CTGrCTA transversion at position 1263 in exon 10


Richard et al.: Calpain 3 Mutations and Polymorphisms 1535

Figure 3 Location of missense mutations and in-frame deletions along the CAPN3 protein

(G1263rA [L421L]). C984rT (C328) was observed in
three unrelated LGMD2 families of different geographic
origins but not in 1100 control individuals. We are con-
fident that two of these families are authentic LGMD2A
families, since one mutant allele already has been char-
acterized. No additional changes were identified in the
corresponding patients’ DNA, despite exhaustive screen-
ing. G1263rA (L421) was encountered only in a Swiss
family in which the segregation is consistent with the
chromosome 15 location of the disease locus.

All these changes may represent neutral polymor-
phisms; however, it is also possible that they represent
authentic LGMD2A mutations that affect the structure
of the transcript. Unfortunately, no biopsies or cell lines
were available for the corresponding patients. Hence,
we could not examine whether these changes have con-
sequences at the RNA level. We can, however, attempt

to predict their putative manifestations on the transcrip-
tion products. IVS12234DC may affect the branch site,
since it is located at the corresponding consensus posi-
tion. C984rT and G1263rA create potential splicing
sites, with scores of 59% and 80%, respectively (Shapiro
and Senapathy 1987). The last two unclassified variants,
H774D and A798E, lead to a change in the amino acid
composition; but the presence, at the corresponding po-
sition, of the variant amino acid in other members of
the calpain family is a strong argument in favor of their
representing rare innocuous polymorphisms (table 5).

Recurrent Mutations or Founder Effect

Twenty-nine mutations were each found in more
than one family; however, it is necessary to include a
correction for founder effect, since several mutations


1536 Am. J. Hum. Genet. 64:1524–1540, 1999

have been amplified in specific populations. This is
particularly true for the Amish community, in which
the R769Q mutation is encountered in a homozygous
state in all 19 families with LGMD2A that were
tested. A predominant mutation is also observed in
two other populations with high consanguinity.
IVS621GrA and 2362AGrTCATCT are present on
160% (25/39) and 76% (42/55) of the carrier chro-
mosomes in families from Réunion Island and in the
Basque isolate, respectively. Haplotype analyses by
microsatellite markers that flank the LGMD2A locus
confirmed that the mutations derived from the same
ancestral mutational event (Allamand et al. 1995; Ri-
chard et al. 1995; Urtasun et al. 1998). The same
analysis allowed us also to infer a relationship be-
tween families, even in the absence of any known ge-
nealogical or geographic link. In particular, the
550DA mutation observed in patients from eight dif-
ferent countries was found, each time the information
was available, to be associated with the same
LGMD2A marker haplotype. This is also the case for
(i) Y537X and A702V, which are found in Turkish
and metropolitan French families; (ii) R769Q, which
is present in both Amish and metropolitan French
families; and (iii) 2362AGrTCATCT, which is present
in 2 North American, 1 Brazilian, 3 Réunion Island,
and 49 Spanish families. In other cases, haplotype
data allowed us to rule out the possibility of a com-
mon founder effect and suggested, instead, that the
mutations represent events that have occurred several
times independently. Overall, if only the clearly in-
dependent events are considered, the number of re-
current mutations collapses to seven, representing 16
independent events (table 6); among these mutations,
five correspond to CrT transitions involving CpG
sites. Thus, altogether, 105 independent mutations
were observed.

Distribution of CAPN3 Mutations

Mutations are relatively evenly distributed over all
exons of the calpain 3 gene, with the exception of exons
9, 12, 14, 23, and 24, all of which are among the smallest
exons and in which no mutations have been found so
far (fig. 1). These exons are to be contrasted with exon
21, which carries an excess of mutations ( ). WhenP ! .01
the nature of the mutations is considered, exons 5, 11,
and 21 show an excess of missense mutations ( ,P ! .05
.001, and .05, respectively), and, when allowance is
made for the small numbers, exons 15, 17, and 22 may
show an excess of deletions or insertions. Interestingly,
a mutation cluster spanning a region of 11 amino acids
in exon 11 accounts for 18% (11/61) of the independent
missense events. This may be due to the elevated number
of CpG sites at this location. No differences in the dis-

tribution of mutations were observed between the four
calpain 3 domains.

Rapid Screening for Particular Mutations

In some cases, rapid testing for the presence of a par-
ticular mutation in a patient’s DNA may be important,
especially in the case of recurrent or population-specific
mutations. Direct sequencing may be the method of
choice, but alternatives allowing the analysis of a great
number of individuals exist. Some CAPN3 mutations
can be tested rapidly because they cause a change in a
restriction-enzyme site (table 7). Short deletions could
be tested directly by electrophoresis of the corresponding
PCR products. In addition, we have developed allele-
specific PCR assays (table 1) for the mutations present
in families from the Amish community, from Réunion
Island, and from the Basque country. This has allowed
us to test rapidly the presence of the corresponding mu-
tations in new patients with these origins. We also have
done this for the frequently observed mutation 550DA,
which is always associated with the same haplotype,
allowing to test candidate carriers presenting either this
particular haplotype or a related haplotype.

Phenotypic Characteristics of Patients with LGMD2A

Clinical information was available for 163 of the pa-
tients presented here. For many of them, clinical char-
acteristics have been reported elsewhere, and a typical
formula has been defined, first in patients from Réunion
Island (Fardeau et al. 1996b). This formula is charac-
terized mainly by a symmetrical, very selective atrophic
involvement of limb-girdle and trunk muscles, with the
gluteus maximus and thigh adductors being most af-
fected (Fardeau et al. 1996a, 1996b). The same pattern
of muscle involvement was also reported for 3/4 of the
examined metropolitan French patients, with occasion-
ally minor variations around this pattern. The same
holds for the majority of LGMD2A patients of Turkish
or Basque origin (Dinçer et al. 1997; Topaloglu et al.
1997; Beckmann and Fardeau 1998; Urtasun et al.
1998). Overall, there was a marked heterogeneity of
severity. The mean age at onset is 13.7 years (range 2–40
years old; fig. 2), and the mean age at loss of walking
ability is 17.3 years (range 5–39 years; fig. 2). No sex
difference was evident in either age at onset or evolution.

In an attempt to draw correlations between the clinical
severity and the nature of gene mutations, evolution
curves associating age at onset and functional stages
were separated and compared according to the nature
of the mutations (null/null, null/missense, and missense/
missense) (fig. 3). The age at onset in patients carrying
two null mutations is quite homogeneous and occurs
generally at age 15 years. In contrast, age at onset in
patients who either had two missense mutations or were


Richard et al.: Calpain 3 Mutations and Polymorphisms 1537

compound heterozygotes for one missense mutation and
one null mutation is much more variable and can occur
as early as age 2.5 years or as late as age 40 years.
Furthermore, patients carrying two null mutations lost
walking ability at age !40 years, whereas patients with
either one missense and one null mutation or two mis-
sense mutations had broader variability in the evolution
curves. These data further confirm and extend prior ob-
servations regarding the relative severity of null muta-
tions versus missense mutations.

Discussion

In this report, we have presented all the mutations
and polymorphisms identified so far in CAPN3, the gene
involved in LGMD2A. A total of 114 variants have been
identified in the calpain 3 gene: 97 mutations, 12 poly-
morphisms, and 5 variants that could not be classified
as either mutation or polymorphism. Until one assesses
the impact of the latter on mRNA processing and sta-
bility, it may be difficult to ascertain the neutrality or
nonpathogenicity of these variants. The precedent dem-
onstration of the pathogenic character of the 1872CrT
mutation, which does not change the nature of the en-
coded amino acid (Richard and Beckmann 1995), con-
firms that one needs to interpret with caution the ap-
parently “neutral” mutations.

The 97 mutations reported thus far, 56 of which are
newly described, include 4 nonsense, 32 deletion/inser-
tion, 8 splice-site, and 53 missense mutations. If we
consider the additional, unclassified IVS12234DC,
C495rT, and G1263rA variants as causative muta-
tions, then the number of mutations characterized thus
far would total 100. The mutations are distributed along
the length of the CAPN3 gene, with a slight mutational
“hot spot” in exon 21. If we consider missense mutations
alone, exon 11 shows a notable excess of mutations.
This exon constitutes a good primary target for screening
of mild calpainopathies, in light of the degree of disease
severity of carriers of either two missense mutations or
a missense mutation associated with a null mutation.

Most mutations (68/97 [70%]) represent private var-
iants, although particular mutations were found more
frequently. Among these particular mutations, 11 (38%)
are associated with a founder effect, 7 (24%) are really
recurrent, and no information was obtained for the re-
maining mutations. CAPN3 mutations were identified
in 181 families originating from 19 countries, further
demonstrating the worldwide distribution of the disease.
For 24 families, only one mutant allele was identified,
despite the fact that all coding regions in these patients
have been examined. These observations therefore sug-
gest that mutations may occasionally lie in an essential
noncoding regions, such as the promoter region or an
intron. We also have described the first large genomic

deletion of the calpain 3 gene, which is due to an unequal
recombination between two intragenic Alu elements.
The methodologies usually used for mutation screening
are not oriented toward identification of such altera-
tions. It would be interesting, given the relatively high
Alu-sequences content of the calpain 3 region, to verify
what fraction of the 23 unidentified mutations can be
explained by such events. For this purpose, it may be
useful to resort to use of additional procedures, such as
Southern blot.

Diagnostic Applications

After the initial confirmation of CAPN3 as the
LGMD2A gene, the primary clinical motivation for per-
forming a systematic mutation analysis was to provide
accurate and unambiguous LGMD2A diagnosis for pa-
tients with progressive muscular dystrophies. In partic-
ular, the identification of mutations in a family allowed
the latter to be classified as belonging to the LGMD2A
group. This is of importance in light of (i) the genetic
heterogeneity of LGMD2 (which has at least eight caus-
ative genes, LGMD2A–LGMD2H) and (ii) the difficulty
in clinically distinguishing one entity from another. Even
if it is now possible, with the help of antisarcoglycans
antibodies, to test for defects in the corresponding pro-
teins and to distinguish sarcoglycanopathies from the
other progressive muscular dystrophies, unambiguous
diagnosis still rests on the identification of the underlying
mutations. Hence, genetic and immunochemical analy-
ses provide important means by which to refine and
simplify diagnosis.

The wide spectrum of CAPN3 mutations, the relative
large size of the CAPN3 gene, and, finally, the fact that
most mutations seem to be private and relatively evenly
distributed over most calpain 3 exons create significant
practical problems for diagnosis of calpainopathies. Be-
cause of all these characteristics, no single mutation-
detection method appears to be ideal, and the challenge
of mutation identification remains important. The as-
sessment of the calpain 3 protein in muscle biopsies, by
either western blot or immunohistochemistry, may be-
come an additional tool with which to investigate cal-
painopathies, although it should be remembered that the
autoproteolytic nature of calpain 3 could render this
diagnostic test difficult to interpret (Anderson et al.
1998; Beckmann and Fardeau 1998). Nevertheless, the
potential of this diagnostic strategy, which has been fol-
lowed by the Newcastle team, is clearly demonstrated
in the present study.

Since the realization of an objective molecular-genetic
diagnosis still presents a formidable challenge, it could
be helpful, whenever possible, before a direct mutation
search is performed, to further classify the family—or,
at least, to exclude particular loci—by linkage analysis.


1538 Am. J. Hum. Genet. 64:1524–1540, 1999

To streamline the genetic analyses, we have developed
a fluorescent-marker kit, using a set of markers brack-
eting the LGMD2A–LGMD2F disease loci (Richard et
al., in press). In an additional step, haplotype analyses
can be used to infer a common ancestral origin and thus
to point to a specific mutation(s). The geographic origins
of the patients is another element that can direct the
mutation screening.

We also have reported herein the polymorphisms iden-
tified in the study of our patient’s cohort, since it is
essential, in a diagnostic test, to distinguish between
them and the morbid DNA-sequence anomalies and,
thereby, to prevent erroneous diagnosis. This is of par-
ticular importance in the detection of mutations in exons
1 and 22 and in intron 22, which contain polymor-
phisms that can reach 10%–12% heterozygosity in the
population examined. The inclusion of appropriate con-
trols in the mutation analysis of this exon will help us
to distinguish between these neutral polymorphisms and
causative mutations.

LGMD2 Phenotype

The availability of molecular diagnosis has enabled
precise symptomatology, by identification of clues al-
lowing the recognition of specific clinical features of cal-
painopathies, compared with other progressive muscular
dystrophies caused by defects in structural proteins. Mo-
lecular classification of patients as belonging within the
different LGMD2 forms helps us to validate the specific
topography and characteristics of muscle involvement
(for a precise description of LGMD2A patients, see (Far-
deau et al. 1996a, 1996b; Dinçer et al. 1997; Topaloglu
et al. 1997; Beckmann and Fardeau 1998; Urtasun et
al. 1998). To sum up, the typical formula with regard
to calpain 3–deficient patients includes, in the early stage
of the disease, a predominant wasting of muscles from
the posterior compartments of the limbs, clearly visible
by both clinical examination and computed-tomography
(CT) scan, whereas in sarcoglycanopathies there is a
marked quadriceps femoris involvement (Eymard et al.
1997). The differences, in the distribution of muscle
wasting, between the different forms renders CT-scan
analysis an important element in a clinical diagnosis. An
additional sign that may help us to distinguish calpain-
opathies from sarcoglycanopathies is the fact that, in the
former, calf hypertrophy is less frequent, whereas ma-
croglossia is never seen.

Both the characterization of mutations and the study
of phenotype/genotype correlations should allow us to
elucidate the molecular basis for the expression of dis-
parate phenotypes among calpain 3–deficient patients.
No evident correlations between the mutations and the
clinical manifestations of LGMD2A has been estab-
lished, either between the families or even within the
families, although, in general, null mutations result in

clinical consequences that are more severe than those
that result from missense mutations. Furthermore, the
study of this large cohort of calpain 3–deficient patients
has failed to show that the patient’s gender has any
influence on either age at onset or disease evolution.

Locus-Specific Database

Given both the widespread geographic distribution of
this disease and the growing number of reported cases,
it seems important, in order to help investigators and
clinicians in their diagnostic process, to maintain and
update a CAPN3-mutation database. With this goal in
mind, we are in the process of constructing a locus-
specific database Website in accordance with HUGO
guidelines for content and structure of mutations da-
tabases, and we encourage others to join us and to par-
ticipate in a collective effort to regroup all CAPN3 mu-
tations in a publicly available database. Meanwhile,
some of these data are already accessible in the Leiden
Muscular Dystrophy pages.

Acknowledgments

We would like to thank the LGMD2A families. We are grate-
ful to all of those clinicians who have collected samples, in
particular Drs. O. F. Brouwer (LUMC, Leiden), J. M. Bur-
gunder (Neurologie Inselpital, Bern), P. Dinçer (Hacettepe Uni-
versity, Ankara), and C. E. Jackson (Henri Ford Hospital, De-
troit) and Profs. G. Lefranc (University of Montpellier, France),
C. Legum (Ichilov Hospital, Tel Aviv), and L. Merlini (Istituto
Rizzoli, Bologna). This study was supported by grants from
the Association Française contre les Myopathies, by Marato
TV3 grant 1012/97, and by FIS grant 98/0040-2,1 from the
Spanish Ministry of Health. A.S. is a predoctoral fellow from
the Basque Ministry of Health.

Electronic-Database Information

Online Mendelian inheritance in Man (OMIM), http://
www3.ncbi.nlm.nih.gov:80 (for LGMD2A [MIM 253600])

Leiden Muscular Dystrophy pages, http://www.dmd.nl

References

Allamand V, Broux O, Richard I, Fougerousse F, Chiannilk-
ulchai N, Bourg N, Brenguier L, et al (1995) Preferential
localization of the limb-girdle muscular dystrophy type 2A
gene in the proximal part of a 1-cM 15q15.1-q15.3 interval.
Am J Hum Genet 56:1417–1430

Anderson LVB, Davison K, Moss JA, Richard I, Fardeau M,
Tomé FMS, Hübner C, et al (1998) Characterization of mon-
oclonal antibodies to Calpain 3 and protein expression in
muscle from patients with limb-girdle muscular dystrophy
type 2A. Am J Pathol 153:1169–1179

Antonarakis SE (1998) Recommendations for a nomenclature
system for human gene mutations: Nomenclature Working
Group. Hum Mutat 11:1–3

Bashir R, Britton S, Strachan T, Keers S, Vafiadaki E, Lako


Richard et al.: Calpain 3 Mutations and Polymorphisms 1539

M, Richard I, et al (1998) A gene related to the C. elegans
spermatogenesis factor fer-1 is mutated in limb-girdle mus-
cular dystrophy type 2B. Nat Genet 20:37–42

Bashir R, Strachan T, Keers S, Stephenson A, Mahjneh I, Mar-
coni G, Nashef L, et al (1994) A gene for autosomal recessive
limb-girdle muscular dystrophy maps to chromosome 2p.
Hum Mol Genet 3:455–457

Beckmann JS, Fardeau M (1998) Limb-girdle muscular dys-
trophies. In: Emery AEH (ed) Neuromuscular disorders:
clinical and molecular genetics. John Wiley & Sons, Chich-
ester, UK, pp 123–156

Beckmann JS, Richard I, Hillaire D, Broux O, Antignac C,
Bois E, Cann H, et al (1991) A gene for limb-girdle muscular
dystrophy maps to chromosome 15 by linkage analysis. C
R Acad Sci III 312:141–148

Bönnemann CG, Modi R, Noguchi S, Mizuno Y, Yoshida M,
Gussoni E, McNally EM, et al (1995) b-Sarcoglycan (A3b)
mutations cause autosomal recessive muscular dystrophy
with loss of the sarcoglycan complex. Nat Genet 11:
266–273

Campbell KP (1995) Three muscular dystrophies: loss of cy-
toskeleton-extracellular matrix linkage. Cell 80:675–679

Chiannilkulchai N, Pasturaud P, Richard I, Auffray C, Beck-
mann JS (1995) A primary expression map of chromosome
15q15 region containing the recessive form of limb-girdle
muscular dystrophy (LGMD2A) gene. Hum Mol Genet 4:
717–726

Dinçer P, Leturcq F, Richard I, Piccolo F, Yalnizoglu D, de
Toma C, Broux O, et al (1997) A biochemical, genetic and
clinical survey of autosomal recessive limb-girdle muscular
dystrophies in Turkey. Ann Neurol 42:222–229

Eymard B, Romero NB, Leturcq F, Piccolo F, Carrié A, Jean-
pierre M, Collin H, et al (1997) Primary adhalinopathy (a-
sarcoglycanopathy): clinical, pathological and genetic cor-
relation in twenty patients with autosomal recessive
muscular dystrophy. Neurology 48:1227–1234

Fardeau M, Eymard B, Mignard C, Tomé FMS, Richard I,
Beckmann JS (1996a) Chromosome 15-linked limb girdle
muscular dystrophy: clinical phenotypes in Réunion island
and french metropolitan communities. Neuromusc Disord
6:447–453

Fardeau M, Hillaire D, Mignard C, Feingold N, Mignard D,
de Ubeda B, Collin H, et al (1996b) Juvenile limb-girdle
muscular dystrophy: clinical, histopathological, and genetic
data from a small community living in the Reunion Island.
Brain 119:295–308

Fougerousse F, Broux O, Richard I, Allamand V, Pereira de
Souza A, Bourg N, Brenguier L, et al (1994) Mapping of a
chromosome 15 region involved in limb-girdle muscular
dystrophy. Hum Mol Genet 3:285–293

Gardner-Medwin D, Walton JN (1974) The clinical exami-
nation of the voluntary muscles. In: Walton JN (ed) Dis-
orders of the voluntary muscles, 3d ed. Churchill Living-
stone, Edinburgh, pp 517–560

Häffner K, Speer A, Hübner C, Voit T, Oexle K (1998) A small
in-frame deletion within the protease domain of muscle-
specific calpain, p94 causes early-onset limb-girdle muscular
dystrophy. Hum Mutat Suppl 1:S298-S300

Herasse M, Ono Y, Fougerousse F, Kimura E, Beley C, Mon-
tarras D, Pinset C, et al. Expression and functional char-
acteristics of calpain 3 isoforms generated through tissue-

specific transcriptional and post-transcriptional events. Mol
Cell Biol (in press)

Kawai H, Akaike M, Kunishige M, Inui T, Adachi K, Kimura
C, Kawajiri T, et al (1998) Clinical, pathological and mo-
lecular features of limb-girdle muscular dystrophy
(LGMD2A) with new calpain 3 gene mutations in seven
patients from three Japanese families. Muscle Nerve 21:
1493–1501

Lim LE, Duclos F, Broux O, Bourg N, Sunada Y, Allamand
V, Meyer J, et al (1995) b-Sarcoglycan (43DAG): charac-
terization and role in limb-girdle muscular dystrophy linked
to chromosome 4q12. Nat Genet 11:257–265

Liu J, Aoki M, Illa I, Wu C, Fardeau M, Angelini C, Serrano
C, et al (1998) Dysferlin, a novel skeletal muscle gene, is
mutated in Miyoshi myopathy and limb-girdle muscular
dystrophy. Nat Genet 20:31–36

Ma H, Fukiage C, Azuma M, Shearer TR (1998a) Cloning
and expression of mRNA for calpain Lp82 from rat lens:
splice variant of p94. Invest Ophthalmol Vis Sci 39:454–461

Ma H, Shih M, Hata I, Fukiage C, Azuma M, Shearer TR
(1998b) Protein for Lp82 calpain is expressed and enzy-
matically active in young rat lens. Exp Eye Res 67:221–229

Moreira ES, Vainzof M, Marie SK, Sertié AL, Zatz M, Passos-
Bueno MR (1997) The seventh form of autosomal recessive
limb-girdle muscular dystrophy is mapped to 17q11-12. Am
J Hum Genet 61:151–159

Nigro V, de Sà Moreira E, Piluso G, Vainzof M, Belsito A,
Politano L, Puca AA, et al (1996) Autosomal recessive limb-
girdle muscular dystrophy, LGMD2F, is caused by a mu-
tation in the d-sarcoglycan gene. Nat Genet 14:195–198

Noguchi S, McNally EM, Ben Othmane K, Hagiwara Y, Mi-
zuno Y, Yoshida M, Yamamoto H, et al (1995) Mutations
in the dystrophin-associated protein g-sarcoglycan in chro-
mosome 13 muscular dystrophy. Science 270:819–822

Ono Y, Shimada H, Sorimachi H, Richard I, Saido TC, Beck-
mann JS, Ishiura S, et al (1998) Functional defects of a
muscle-specific calpain, p94, caused by mutations associated
with limb-girdle muscular dystrophy type 2A. J Biol Chem
273:17073–17078

Penisson-Besnier I, Richard I. Beckmann JS, Fardeau M (1998)
Phenotypic variations of calpain deficiency in two siblings.
Muscle Nerve 21:1078–1080

Richard I, Beckmann JS (1995) How neutral are synonymous
codon mutations? Nat Genet 10:259

Richard I, Brenguier L, Dinçer P, Roudaut C, Bady B, Bur-
gunder J-M, Chemaly R, et al (1997) Multiple independent
molecular etiology for limb-girdle muscular dystrophy type
2A patients from various geographical origins. Am J Hum
Genet 60:1128–1138

Richard I, Bourg N, Marchand S, Alibert O, Eymard B, van
der Kooi AJ, Jackson CE, et al. A diagnostic fluorescent
marker kit for six limb-girdle muscular dystrophies. Neu-
romusc Disord (in press)

Richard I, Broux O, Allamand V, Fougerousse F, Chiannilk-
ulchai N, Bourg N, Brenguier L, et al (1995) Mutations in
the proteolytic enzyme, calpain 3, cause limb-girdle mus-
cular dystrophy type 2A. Cell 81:27–40

Robberson BL, Cote GJ, Berget SM (1990) Exon definition
may facilitate splice site selection in RNAs with multiple
exons. Mol Cell Biol 10:84–94

Roberds SL, Leturcq F, Allamand V, Piccolo F, Jeanpierre M,


1540 Am. J. Hum. Genet. 64:1524–1540, 1999

Anderson RD, Lim LE, et al (1994) Missense mutations in
the adhalin gene linked to autosomal recessive muscular dys-
trophy. Cell 78:625–633

Shapiro MB, Senapathy P (1987) RNA splice junctions of dif-
ferent classes of eukaryotes: sequence statistics and func-
tional implications in gene expression. Nucleic Acid Res 15:
7155–7174

Smith CWJ, Porro EB, Patton JG, Nadal-Ginard (1989) Scan-
ning from an independently specified branch point defines
the 3′ spice site of mammalian exon. Nature 342:243–247

Sorimachi H, Imajoh-Ohmi S, Emori Y, Kawasaki H, Ohno
S, Minami Y, Suzuki K (1989) Molecular cloning of a novel
mammalian calcium-dependent protease distinct from both
m- and mu-type: specific expression of the mRNA in skeletal
muscle. J Biol Chem 264:20106–20111

Topaloglu H, Dinçer P, Richard I, Leturcq F, Akçoren Z, Ur-
tizberea JA, Alehan D, et al (1997) Calpain deficiency causes

a mild muscular dystrophy in childhood. Neuropediatrics
28:212–216

Urtasun M, Saenz A, Roudaut C, Poza JJ, Urtizberea JA, Cobo
AM, Richard I, et al (1998) Limb-girdle muscular dystrophy
in Guipuzcoa (Basque country, Spain). Brain 121:
1735–1747

Vignal A, Gyapay G, Hazan J, Nguyen S, Dupraz C, Cheron
N, Becuwe N, et al (1993) Non-radioactive multiplex pro-
cedure for genotyping of microsatellite markers. In: Adolph
KW (ed) Methods in molecular genetics. Academic Press,
San Diego, pp 211–221

Weiler T, Greenberg CR, Zelinski T, Nylen E, Coghlan G,
Crumley MJ, Fujiwara TM, et al (1998) A gene for auto-
somal recessive limb-girdle muscular dystrophy in Manitoba
Hutterites maps to chromosome region 9q31-33: evidence
for another limb-girdle muscular dystrophy locus. Am J
Hum Genet 63:140–147