Heterogeneity of Nemaline Myopathy Cases

with Skeletal Muscle ␣-Actin

Gene Mutations
Pankaj B. Agrawal, MD, MMSc,1–3 Corinne D. Strickland, MS,1 Charles Midgett, BS,1 Ana Morales, MS,1
Daniel E. Newburger,1 Melisa A. Poulos, BA,1 Kinga K. Tomczak, MD,1 Monique M. Ryan, MB, BS,1,3
Susan T. Iannaccone, MD,4 Tom O. Crawford, MD,5 Nigel G. Laing, PhD,6 and Alan H. Beggs, PhD1,3

Nemaline myopathy (NM) is the most common of several congenital myopathies that present with skeletal muscle
weakness and hypotonia. It is clinically heterogeneous and the diagnosis is confirmed by identification of nemaline
bodies in affected muscles. The skeletal muscle ␣-actin gene (ACTA1) is one of five genes for thin filament proteins
identified so far as responsible for different forms of NM. We have screened the ACTA1 gene in a cohort of 109
unrelated patients with NM. Here, we describe clinical and pathological features associated with 29 ACTA1 mutations
found in 38 individuals from 28 families. Although ACTA1 mutations cause a remarkably heterogeneous range of phe-
notypes, they were preferentially associated with severe clinical presentations (p < 0.0001). Most pathogenic ACTA1
mutations were missense changes with two instances of single base pair deletions. Most patients with ACTA1 mutations
had no prior family history of neuromuscular disease (24/28). One severe case, caused by compound heterozygous
recessive ACTA1 mutations, demonstrated increased ␣-cardiac actin expression, suggesting that cardiac actin might par-
tially compensate for ACTA1 abnormalities in the fetal/neonatal period. This cohort also includes the first instance of an
ACTA1 mutation manifesting with adult-onset disease and two pedigrees exhibiting potential incomplete penetrance.
Overall, ACTA1 mutations are a common cause of NM, accounting for more than half of severe cases and 26% of all NM
cases in this series.
Ann Neurol 2004;56:86 –96

The congenital myopathies are rare neuromuscular dis-
orders characterized by nonprogressive or slowly pro-
gressive muscle weakness. They often are associated
with structural abnormalities of the contractile appara-
tus or other architectural abnormalities within muscle
fibers. Given significant phenotypic overlap, diagnosis
of the specific forms of congenital myopathy generally
begins with the histological characterization of these ar-
chitectural changes in a muscle biopsy specimen. In
general, clinical expression of nemaline myopathy
(NM) is characterized by weakness of the face, neck
flexors, and proximal muscles of the limbs.1,2 The pre-
sentation may range from stillbirth or infantile demise
to a mild skeletal myopathy presenting in later life.3– 6
The characteristic pathological feature of NM is a
rod-like structure, or “nemaline body,” seen on light
microscopy with Gomori trichrome staining. Ultra-
structurally, these electron-dense nemaline bodies orig-
inate from the Z-disc of sarcomeres. Other nonspecific
pathological findings frequently found in NM include
fiber size variation, fiber-type disproportion, and type I
fiber predominance. More rarely, dystrophic features
such as degenerating and regenerating fibers, internal
nuclei, and increased connective tissue may be ob-
served.1,7
NM is genetically, as well as phenotypically, hetero-
geneous. Causative mutations have now been described
in five genes encoding different skeletal muscle thin fil-
ament proteins.2 Nebulin gene (NEB) mutations are
likely responsible for most autosomal recessive cases8,9;
however, the large size of this gene presents an obstacle
to the efficient detection of pathogenic mutations. Mu-
tations in NEB usually present with typical congenital
NM,8 but severe and mild presentations have also been

From the 1Genomics Program and Division of Genetics and 2Divi-
sion of Neonatology, Children’s Hospital; 3Harvard Medical
School, Boston, MA; 4Department of Pediatric Neurology, Scottish
Rite Hospital for Children, Dallas, TX; 5Department of Neurology,
Johns Hopkins University School of Medicine, Baltimore, MD;
6
Centre for Neuromuscular and Neurological Disorders, University
of Western Australia, Australian Neuromuscular Research Institute,
and Centre for Medical Research, West Australian Institute of Med-
ical Research, QEII Medical Centre, Nedlands, Western Australia,
Australia.
Received Feb 20, 2004, and in revised form Apr 13. Accepted for
publication Apr 13, 2004.
See Appendix on page 93.
Published online Jun 28, 2004, in Wiley InterScience
(www.interscience.wiley.com). DOI: 10.1002/ana.20157
Address correspondence to Dr Beggs, Genetics Division, Children’s
Hospital Boston, 300 Longwood Avenue, Boston, MA 02115.
E-mail: beggs@enders.tch.harvard.edu

86 © 2004 American Neurological Association

Published by Wiley-Liss, Inc., through Wiley Subscription Services
described.10 The tropomyosin 3 gene (TPM3) has been
implicated in rare cases of dominant and recessive
nemaline myopathy.11–13 A single instance of mutation
also has been identified in the gene for slow troponin
T (TNNT1), causing a unique form of recessive nema-
line myopathy in the Amish population.14 Most re-
cently, dominant mutations in the tropomyosin 2 gene
(TPM2) have been reported in two patients.15 Overall,
mutations in the TNNT1, TPM2, and TPM3 genes
are rare, likely accounting for less than 5% of all cases.
Mutations recently have been reported in the skeletal
muscle ␣-actin (ACTA1) gene in patients with nema-
line myopathy16,17 as well as in patients with “actin
myopathy,” a related condition characterized by abnor-
mal accumulations of actin filaments within muscle fi-
bers.16 The clinical severity and mode of inheritance in
NM patients with ACTA1 mutations is highly variable
(reviewed by Sparrow and colleagues18). Most reported
ACTA1 mutations have been sporadic, but several fam-
ilies manifesting dominant inheritance have been ob-
served.
This study was undertaken to describe the clinical
features, pathology, and molecular findings of nemaline
myopathy caused by mutations in ACTA1. Here, we
report the relationship of genotype to phenotype in 38
individuals with ACTA1 mutations from 28 families
derived from a large, mostly North American, cohort.
In this group, ACTA1 mutations are responsible for
26% (28/109) of NM cases with a broad range of clin-
ical and pathological phenotypes, including, for the
first time to our knowledge, several cases illustrating
probable incomplete penetrance.
Patients and Methods
Patient Ascertainment and Classification
One hundred nine patients with clinicopathological diag-
noses of NM were screened for mutations in the ACTA1
gene. The diagnosis was established by muscle biopsy, and all
cases were considered to represent primary NM because of
absence of findings indicative of other disorders in which
nemaline bodies may be found.19,20 For purposes of pheno-
typic characterization, six clinical categories of NM have
been defined using criteria based on age of onset and severity
of involvement.4,19 Congenital joint contractures or bone
fractures, an immediate need for assisted breathing beginning
at birth, or absent spontaneous movements at birth, define
the “severe” congenital form of NM. Patients with the “in-
termediate” congenital form are able to breath independently
at birth but manifest early weakness by failure to develop, or
loss of, early motor milestones, loss of independent respira-
tion, or both. The “typical” congenital form presents in in-
fancy or early childhood with weakness involving proximal
muscles, facial, bulbar, neck, and respiratory muscles. Mile-
stones often are delayed but reached, and patients may re-
main ambulant well into adulthood. The “mild” or
childhood-onset form starts in childhood with a milder
course, whereas the “adult-onset” form, by definition, pre-
sents later. Atypical cases not meeting any of the above cri-
teria are classified as the sixth, “other” form.
Patients represent a mixture of ethnic backgrounds arising
from referral by clinicians throughout North and South
America. This study population has been described previ-
ously as the “North American” cases, characterized by Ryan
and colleagues.4,7 Nowak and colleagues previously published
molecular information on four with ACTA1 mutations,16
and partial data for others have been included in a recent
review.18 The research protocol received prior approval from
the Children’s Hospital Boston Institutional Review Board,
and appropriate informed consent was obtained from all
study participants.
Statistical analysis was performed using STATA 7 software
with categorical variables analyzed by ␹2 test to calculate p
values.
Mutation Analysis
Peripheral blood DNA was extracted using “Puregene” kits
(Gentra Systems, Minneapolis, MN). Polymerase chain reac-
tion (PCR) primers were synthesized and used essentially as
described by Nowak and colleagues.16 Alternate primers for
exon 3 were 3Falt (5⬘-GGGTGCGTGGTGTCTCGGCT-
CT-3⬘) and 3Ralt (5⬘-GGGGCAGCGGGCACTCA-3⬘),
and for exon 5, 5Falt (5⬘-CGGCCCCTGAGTGAG-3⬘) and
5Ralt (5⬘-GGCGAGGGCGAGC-3⬘). Four percent dimeth-
ylsulphoxide was added to reactions amplifying exon 7. PCR
primer annealing temperatures were as follows: 5⬘ untrans-
lated region/exon 1, exon 2, exon 3 (3Falt/3Ralt), exon 4,
exon 5, 60°C; exon 3, exon 6, 64°C; exon5 (5Falt/5Ralt),
exon 7, 56°C. PCR products were purified using either the
High Pure PCR purification kit (Roche Molecular Bio-
chemicals, Indianapolis, IN) or the Microcon PCR purifica-
tion kit (Millipore, Bedford, MA), directly sequenced using
ABI Big Dye Terminator chemistry and analyzed on an ABI
377 or 3730 DNA analyzer (Applied Biosystems, Foster
City, CA). Sequencing primers were same as for PCR except
in cases of heterozygosity for intronic insertion/deletion poly-
morphisms when the alternate primers described above were
used. The numbering of altered nucleotide bases was as-
signed where A of the ATG initiation codon was base 1 of
the ACTA1 cDNA sequence (GenBank accession no.
J00068), but aspartic acid at residue 3 is numbered amino
acid 1, because of posttranslational processing that removes
the first two amino acids.
Putative mutations were confirmed on replicate DNA
samples, and DNA from all available first-degree relatives
was evaluated for the same change. In addition, each candi-
date mutation was screened for in a panel of unaffected con-
trol DNAs using single-strand conformational polymorphism
analysis21 or denaturing high-performance liquid chromatog-
raphy using a Transgenomic Wave System (Transgenomic,
Omaha, NE).
Immunohistochemistry and Immunoblotting
Indirect immunofluorescence was performed essentially as
described.22 All patients younger than 1 year at time of bi-
opsy, for whom high-quality tissue specimens were available,
were analyzed. Antibodies used included mouse monoclonal
anti-␣-cardiac actin (Mab Ac1-20.4.2) at 1 to 40 dilution
Agrawal et al: Actin ACTA1 in Nemaline Myopathy 87
(Progen, Heidelberg, Germany), rabbit anti–pan-actin anti-
body (AA20-33) at 1 to 200 dilution (Sigma, St. Louis,
MO), and rabbit anti ␣-actinin-2 antibody (4B) at 1 to
200.23 Secondary antibodies were Alexa Fluor 488 goat anti–
mouse and Alexa Fluor 594 donkey anti–rabbit antibodies at
1 to 40 dilutions (Molecular Probes, Eugene, OR). For im-
munoblotting, 10␮g of lysates were resolved on 4 to 20%
sodium dodecyl sulfate–polyacrylamide gels and electroblot-
ted as described.24 Mouse monoclonal anti–␣-sarcomeric ac-
tin (clone 5C-5) (1:1,000 dilution; Sigma) and mouse anti-
cardiac actin (Ac1-20.4.2) were incubated at 1 to 1,000
dilutions and detected as described.24 To enhance the spec-
ificity of staining for cardiac actin, 1M NaCl was added to
the anticardiac actin primary antibody incubations, followed
by phosphate-buffered saline washes as suggested by the
manufacturer.
Results
We identified 29 different ACTA1 mutations in 28 of
109 fully screened probands. Thirty-eight members of
these 28 families carried these mutations including
three apparently unaffected carriers of dominant muta-
tions described below. The clinical features, classifica-
tion, and genotype for each NM case with ACTA1
mutation are summarized in Appendix. As previously
observed in an overlapping, but distinct collection of
mutations,18 locations of these 29 ACTA1 mutations
were distributed throughout the six coding exons of the
gene without any apparent hotspots (Fig 1).
The Relationship of Clinical Phenotype to the
ACTA1 Mutation
Of the 109 NM patients screened in this series, 25
manifested the severe congenital form. The remaining
84 included predominantly intermediate congenital
cases (23) and typical congenital cases (50) (Table 1).
ACTA1 mutations were responsible for 14 of the 25
cases of severe congenital NM (56%). The remaining
14 ACTA1 mutations were found in 84 nonsevere
cases (17%). Overall, ACTA1 mutations were more
likely to cause the severe form of NM ( p ⬍ 0.0001).
Mortality in severe NM cases due to ACTA1 muta-
tion was high, at 64% (9/14), with all deaths occurring
before the first birthday. This likely reflects the severity
of weakness rather than a specific feature of ACTA1
Fig 1. Distribution of mutations among exons of ACTA1 gene (introns not to scale). Mutations resulting in severe (S), intermedi-
ate (I), typical (T), congenital or adult (A) clinical presentations are indicated above. Recessive mutations are indicated with an
asterisk. Exon numbers and sizes are indicated below. 5⬘ and 3⬘ untranslated region regions are indicated as unfilled boxes at ei-
ther end.
88 Annals of Neurology Vol 56 No 1 July 2004
mutation, however, as the corresponding mortality for
non-ACTA1 mutation cases with severe NM was sim-
ilarly high (7/11), with all deaths in the first year of
life. There were only two deaths among patients with
nonsevere forms of NM, with one death each from the
ACTA1 and non-ACTA1 mutation groups.
Mutations and Mode of Inheritance
Of the 29 mutations, there were 27 missense and two
single-base deletion mutations (see Appendix). None of
these mutations were seen in at least 100 (range, 100 –
200) normal North American chromosomes, providing
evidence for a causative role in the disease. Each of the
missense changes was predicted to alter a highly con-
served amino acid in the ␣-skeletal actin protein. The
two deletions are predicted to result in production of a
truncated protein in one case (Patient 117-1) and an
addition of 42 amino acids at the carboxy terminus in
the other case (Patient 188-1).
Four of the 28 NM families with ACTA1 mutations
(families 80, 90, 104, and 120) in this series displayed
clear autosomal dominant patterns of inheritance. In
each case, the proband manifested a typical congenital
form of NM. For Families 80 and 90, the probands
and affected parents were similarly affected with typical
congenital NM. Some intrafamilial variability was ob-
served in Family 104, with four affected family mem-
bers exhibiting the typical congenital form of nemaline
myopathy whereas one has mild NM, and Family 120
in which an 8-year-old daughter was more severely af-
fected than her father. Two additional families (108,
188) demonstrated potential incomplete penetrance
with probands having the disease and other clinically
unaffected family members carrying the same muta-
tion.
Of the 22 remaining clinically sporadic cases, there
was one unambiguous instance of autosomal recessive
inheritance in a patient (117-1) who was compound
heterozygous for a missense and a deletion change (see
Appendix). Testing of his clinically normal parents
showed the father to be a carrier of the single base pair
deletion and the mother to be a carrier of the missense
change. The remainder were all heterozygotes for pre-
Table 1. Clinical Categorization and Inheritance Patterns of NM Probands with ACTA1 Mutations
Total NM NM Cases
Clinical Cases with ACTA1
Category Studied Mutations
Severe 25 14
Intermediate 23 3 13
Typical 50 10
Mild/childhood 2 0
Adult onset 3 1
Unknown/other 6 0
Total 109 28
NM ⫽ nemaline myopathy; AR ⫽ autosomal recessive; AD ⫽ autosomal domoinant.
sumed dominant missense changes. In 16 patients,
dominant pathogenesis could be confirmed by the ab-
sence of mutation in either parent (indicating de novo
mutation and/or gonadal mosaicism in one parent).
For the remaining five cases, either or both parents
could not be tested because of nonavailability of DNA.
The 21 probands, excluding the autosomal recessive
case, had clinical presentations ranging from severe to
typical congenital onset. Thirteen of these probands
had severe, one had intermediate, and seven had typical
congenital NM. The clinical outcome was poor for se-
vere and intermediate cases: 9 of 13 severe cases and
the intermediate case were all deceased by 1 year of
age. All seven typical congenital NM cases with ages
ranging from 2.5 years to 38 years were alive, ambula-
tory, and needed no respiratory assistance.
Pathological Findings
Detailed pathology data and or slides were available for
25 of the 28 probands in whom ACTA1 mutations
were identified (see Appendix). With one exception
(Patient 226-1), all patients had nemaline bodies on
Gomori trichome staining and/or on electron micros-
copy. Patient 226-1 presented with congenital contrac-
tures, profound weakness, and respiratory insufficiency
and a mutation at ACTA1 residue 146 (Gly146Asp).
This patient’s biopsy was characteristic for actin myop-
athy,25 with presence of small masses of Z-band mate-
rial resembling mininemaline bodies on electron mi-
croscopy. In all remaining cases, nemaline bodies were
mostly sarcoplasmic or subsarcolemmal, appearing as
extensions of the Z-disc on electron microscopy and
often staining positively with anti–␣-actinin-2 antisera
by immunofluorescence analysis (Fig 2). Of the cases
with nemaline bodies, one previously reported16 case
with severe disease (Patient 86-1) had both intranuclear
and sarcoplasmic nemaline bodies, whereas the rest had
sarcoplasmic bodies only. Nemaline bodies were
present in variable numbers, and in variable propor-
tions of fibers, ranging from 10% to virtually all fibers.
Most cases (87%) had abundant nemaline bodies in
No. of ACTA1 Mutations by
% Cases with Mode of Inheritance
ACTA1
Mutations Sporadic AR AD
56 13 1 0
3 0 0
20 6 0 4
0 0 0 0
33 1 0 0
0 0 0 0
26 23 1 4
numbers that do not correlate well with disease sever-
ity.
Common findings on hematoxylin/eosin staining
and myosin ATPase staining were fiber size variation,
mild to severe atrophy of type 1, type 2 or both types
of fibers, type 1 fiber predominance, and occasional
hypertrophy of type 1 or 2 fibers. None of the cases
had the “nontyping” pattern of pathology defined by
Ryan and colleagues7 and Sanoudou and colleagues,26
nor did any exhibit type 2 fiber predominance which
has been reported in 4% of unselected NM cases.7
Moderate to severe increases of endomysial and or peri-
mysial connective tissues were noted in two cases
(117-1 and 308-1). Severe cases with ACTA1 muta-
tions were associated significantly with marked variabil-
ity in fiber size ( p ⫽ 0.006) and severe atrophy of
some fibers ( p ⫽ 0.015). Conversely, typical cases usu-
ally exhibited only mild to moderate variability in fiber
size. However, because severe cases were invariably in-
fants at time of biopsy, whereas typical cases were often
older, we cannot exclude age at biopsy as a confound-
ing variable affecting this association. In general, nema-
line bodies resulting from actin mutations were ultra-
structurally indistinguishable from those in cases
without ACTA1 mutations. Other EM findings, in-
cluding Z-line streaming and myofibrillar disorganiza-
tion, were seen in both ACTA1 and non-ACTA1 mu-
tated cases.
Immunolocalization and Isoform Composition of
Sarcomeric Actins
Although skeletal muscle ␣-actin (ie, the ACTA1 gene
product) is the predominant sarcomeric actin isoform
in skeletal muscle, ␣-cardiac actin (the ACTC gene
product) is also present, particularly in the fetal and
postnatal periods and in regenerating fibers at all
ages.27,28 Because ACTC expression might compensate
for, or otherwise modulate, the effects of ␣-skeletal ac-
tin abnormalities, we assessed expression of these pro-
teins in muscles from eight patients with, and three
without, ACTA1 mutations (Table 2; Figs 2, 3). All
Agrawal et al: Actin ACTA1 in Nemaline Myopathy 89
Fig 2. Imunofluorescence histochemistry of sarcomeric proteins in four nemaline myopathy patients 108-1 (A–C), 221-1 (D–F),
226-1 (G–I), and 117-1 (J–L) carrying ACTA1 mutations. Muscle biopsy specimens were stained with ␣-cardiac actin (A, D, G,
J), pan-actin (B, E, H, K), and ␣-actinin-2 (C, F, I, L). Muscle from Patient 108-1 shows uniformly low levels of cardiac actin
expression in all fibers (A); Patient 221-1 demonstrates several small fibers with intense staining for ␣-cardiac actin suggestive of
regenerative activity (D). Numerous ␣–actinin-2 containing rods are also evident (F). Biopsy from the actin myopathy case (226-1)
shows several fibers with intense subsarcolemmal staining for ␣-cardiac actin (G), a pattern also seen with anti–pan-actin and
␣–actinin-2 antibodies (H, I). Muscle biopsy from the recessive 117-1 is unique as all fibers stain intensely for ␣-cardiac actin ( J).
Similar staining is seen with anti–pan-actin antibodies (K). Occasional atrophic fibers with collections of ␣–actinin-2–positive
nemaline bodies are also seen (L). Scale bar in A ⫽ 10␮m (A–L)

cases analyzed were biopsied at younger than 1 year of
age. Double-label indirect immunofluorescence stain-
ing with cardiac actin-specific and anti–pan-actin anti-
bodies showed several distinct patterns of ␣-cardiac ac-
tin expression. Two patients with ACTA1 mutations
(108-1 and 18-3) with intermediate and typical clinical
presentations exhibited uniform, low-level expression of
␣-cardiac actin in all fibers without evidence for any
regeneration (eg, see Fig 2A–C). Two non-ACTA1
mutation NM cases (164-1, 256-1), as well as an un-

90 Annals of Neurology Vol 56 No 1 July 2004

Table 2. Immunohistochemical Findings on Staining Muscle Biopsies for Cardiac Actin, All Actins, and ␣-Actinin-2 in Nemaline
Myopathy Patients with and without ACTA1 Mutations
␣-actinin-2 Cardiac-actin
(fibers with
␣-actinin-2–
positive No. positive
ID ACTA1 Fiber Size nemaline regenerating No. positive Pan-actin Age at
No. Specimen Mutation Variability bodies) (small) nonregenerating Staining intensity pattern Specimen site biopsy

8-2 T10 Yes Marked None 23% 19% Variable Uniform Quadriceps 18 days
221-1 T164 Yes Moderate 50–60%, 10% 30% Variable Uniform Gastroneimus 8 mo
prominent
103-1 T71 Yes Moderate Indistinct 16% 70% Variable Not uni- Quadriceps 3 days
form
95-1 T77 Yes Moderate 80%, Promi- 17% 2% Variable Occasional Unknown 10 mo
nent intense
108-1 T80 Yes Minimal Few, indistinct 0 100% Low Uniform Quadriceps 7.5 weeks
117-1 T93 Yes Marked Occasional 100% 100% High Uniform Quadriceps 5 weeks
226-1 T156 Yes Marked None 100% 100% Increased inten- Uniform Quadriceps 5.5 mo
sity subsar-
colemmal re-
gion
18-3 T66 Yes Minimal Not done 0 100% Low Uniform Unknown 6 mo
164-1 T115 No Minimal 10%, Indis- 0 50% Low Uniform Quadriceps 7 days
tinct
174-1 T124 No Significant 75%, Promi- 9% 50% Variable Uniform Quadriceps 8 mo
nent
256-1 T182 No Moderate Few, granular 0 90% Low Not done Quadriceps 14 days
223-1 T149 Unaffecteda Mild None 0 50% Low Uniform Quadriceps 1st week
of life

a
Muscle sample from a control patient with no muscle disease.

affected control case (223-1), also exhibited similar
findings. In contrast, muscles from another group of
ACTA1 mutated cases (8-2, 221-1, 103-1, 95-1), rang-
ing in clinical presentation from typical to severe, con-
tained 10 to 25% small intensely staining fibers, indic-
ative of active regeneration (eg, see Fig 2D–F). The
third non-ACTA1 mutation case (174-1) had similar
findings to this group. There was no obvious correla-
tion between cardiac actin expression and severity of
disease in this admittedly small group. All these mus-
cles exhibited uniform, sarcomeric pan-actin staining in
all muscle fibers with no evidence for abnormal accu-
mulations or localization of actin proteins.
The remaining two cases with ACTA1 mutations
each had a unique patterns of actin expression. Muscle
from Patient 226-1, biopsied at 5.5 months of age,
contained many fibers showing intense ring-like, sub-
Fig 3. Western blot analysis of total actin (A) and ␣-cardiac
actin (B) in two control muscles and muscle biopsies from
nemaline myopathy patients (patient identifications indicated
above). Relative loading in each lane is indicated by Coomas-
sie blue–stained gel section shown in (C).
sarcolemmal staining with anti–pan-actin antibodies,
consistent with the pathological diagnosis of actin my-
opathy (Fig 2G–I). Remarkably, a similar pattern was
seen using specific anti–␣-cardiac actin antisera, sug-
gesting that both mutant ␣-skeletal actin and nonmu-
tated ␣-cardiac actin proteins were present in the ab-
normal filamentous accumulations. Muscle from the
recessively inherited ACTA1 NM patient (117-1) at 5
weeks of age was also remarkable. This biopsy alone
exhibited intense staining for cardiac actin in every
myofiber (see Fig 2J–L). Western blot analysis for sar-
comeric actins and cardiac actin confirmed the appar-
ent increase in cardiac actin expression in muscle from
this patient (see Fig 3). No additional abnormalities in
quantity or quality of actin isoforms were noted.
Discussion
We identified 29 ACTA1 mutations causing NM in 28
of 109 probands screened, suggesting that ACTA1 mu-
tations account for approximately a quarter of NM
cases worldwide. This is higher than the 19% (11/59)
and 15% (5/35) in previously studied cohorts16,17;
however, as ACTA1 mutations are more common
among cases of severe NM, observed differences in
ACTA1 mutation frequency may represent ascertain-
ment biases in phenotypic severity of the referral pop-
ulation.
The primary component of muscle thin filaments is
F-actin, formed by the self-association of G-actin
monomers. Many areas of the actin monomer are
known to be involved in the polymerization to F-actin
or interaction with different actin-binding proteins, in-
cluding myosin, tropomyosin, troponin, and nebulin,

Agrawal et al: Actin ACTA1 in Nemaline Myopathy 91

and interaction with nucleotides and cations.29 Muta-
tions in each of these areas were found in patients with
different levels of severity, and, overall, no specific cor-
relations between mutation location and disease sever-
ity were observed (see Appendix).
Most identified ACTA1 mutations are de novo dom-
inant, recognized by the appearance of heterozygosity
for genetically lethal missense mutations that are not
present in the parents. These are most likely dominant
negative mutations that impair local structure and
function but do not significantly alter the overall struc-
ture of the molecule, thereby retaining the monomers
capacity to interact with other molecules in a deleteri-
ous manner. Classic dominant inheritance (ie, parent
to child transmission) was observed within four kin-
dreds in which the expressed phenotype did not com-
promise the reproductive viability. In these families,
the phenotypes were largely consistent between affected
individuals. There were, however, two families with ap-
parent instances of incomplete penetrance. Family 188
segregated a unique mutation predicted to eliminate
the normal translational stop codon, resulting in pro-
duction of a larger protein with 42 additional amino
acids. One might expect this to result in severe disrup-
tion of F-actin polymerization; however, the late onset
of this proband’s (188-1) disease and the lack of clin-
ical phenotype in his mutation-carrying father (188-2)
suggest that this protein defect results in only a mild
pathophysiological outcome. Although paternal so-
matic mosaicism cannot be ruled out to explain the
father’s lack of disease, this mechanism cannot explain
the findings for the other family (108) with potential
incomplete penetrance. In this instance, incomplete
penetrance is suggested by the presence of heterozygos-
ity for ACTA1 Val134Ala in both the proband with
intermediate congenital NM (108-1) and his clinically
unaffected mother (108-7) and maternal grandfather
(108-4). The conclusion that this represents incom-
plete penetrance rests on the assumption that
Val134Ala is the pathogenic mutation responsible for
the proband’s disease. Although, like most ACTA1 mu-
tations, this change has not been seen in other cases of
NM, there is no precedent for any ACTA1 missense
change ever being found in hundreds of unaffected
control subjects screened over the past 5 years (Sparrow
and colleagues18; N. G. Laing and A. H. Beggs, un-
published observations). Furthermore, a neighboring
mutation, Met132Val, found independently in two pa-
tients with mild and typical disease, respectively, re-
cently was shown in vitro to inhibit actin polymeriza-
tion and produce faster sliding thin filaments.30
Whether the phenotypic heterogeneity associated with
these mutations can be attributed to differences in gene
expression, message survival, or to other genetic or en-
vironmental influences is unknown. Ultimately, func-
tional studies of Val134Ala mutant actin and/or con-
92 Annals of Neurology Vol 56 No 1 July 2004
struction of appropriate animal models31 may be
necessary to finally resolve these questions.
The presence of one pair of recessively inherited
NM mutations (Patient 117-1), combining a missense
mutation, Glu259Val, with a deletion/truncation mu-
tation, both inherited from unaffected parents, illus-
trates the extreme heterogeneity associated with
␣-skeletal actin defects. This was one of three reces-
sive cases reviewed by Sparrow and colleagues.18 In-
terestingly, one of the others, first reported by Nowak
and colleagues, represented an independent example
of Glu259Val, in which this mutation was recessively
inherited in compound heterozygous state with an-
other missense change (Leu94Pro) in a patient with
severe disease.16 The functional properties of residue
259 are unknown, although mutations of Glu259 in
yeast produce a recessive phenotype.32 The second
mutation, in Patient 117-1, is a predicted null allele,
expected to produce a truncated and likely unstable
partial peptide. Skeletal actin abundance, assessed
nonquantitatively by western blotting, was not detect-
ably altered in this proband’s muscle biopsy. This,
along with the normal strength of the father who car-
ries the deletion/truncation mutation, supports the idea
that haploinsufficiency for ACTA1 is clinically silent.
Presumably deletion/truncation mutations severely dis-
rupt the structure of the actin molecule, leading to
rapid degradation and nonavailability of this actin vari-
ant, which otherwise might act in a dominant-negative
fashion. We hypothesize that this may be compensated
for by increased expression of the normal allele and/or
sufficient stability of the normal protein to allow at-
tainment of normal protein levels.
An unexpected complexity of NM muscle pathology
is demonstrated by immunohistochemical analysis of
cardiac actin expression in infants with ACTA1 muta-
tions biopsied in infancy.27,28 Previously, Ilkovski and
colleagues did not see any apparent compensatory in-
creases in cardiac actin (ACTC) gene expression in five
patients with ACTA1 mutations.17 We demonstrated
similar noncompensation of cardiac actin for skeletal
muscle actin in our dominant ACTA1 mutation cases.
The one instructive exception involves the recessive
compound heterozygote (Patient 117-1), where the
combination of missense and probable protein null
ACTA1 deletion/frameshift mutation was associated
with cardiac actin overexpression evident by both west-
ern blot analysis and immunohistochemistry. A relative
preservation of sarcomeric organization in a subset of
myofibers in this biopsy (data not shown) suggests that
retention of upregulated fetal ACTC expression in this
infant’s muscle likely complemented the loss of func-
tional skeletal actin, allowing fetal survival and devel-
opment till birth.
We have now identified 28 probands with mutations
in the ACTA1 gene of 109 patients screened. These
results suggest that approximately a quarter of NM
cases may be caused by mutations in the ACTA1 gene.
ACTA1 mutations are associated with very heteroge-
neous clinical, pathological, and genetic findings. Our
data demonstrate that ACTA1 mutations are overrep-
resented among severe NM cases. Given the relatively
small size of the ACTA1 gene, and the great heteroge
neity of NM cases with ACTA1 mutations, we suggest
that ACTA1 should be the first gene screened in cases
of NM and other undefined myopathies characterized
by sarcomeric disruption and/or Z-line abnormalities.
Finally, although rare, incomplete penetrance of
ACTA1 mutations should now be considered when
counseling patients with NM.

Appendix. Clinical, Pathological, and Molecular Findings in Nemaline Myopathy Patients with ACTA1 Mutations
Amino Acid Disease Presenting Respiratory Feeding
Patient Sex Mutation Substitution Severitya Inheritanceb Symptomsc Issues Issues Status Pathological Findingsd

8-2 M c.949C⬎G Asn280Lys SC De novo No spontaneous Needed in- Gastrostomy Died at 9 mo Marked fiber-size variabil-
movements, poor termittent tube and of aspiration ity, numerous atrophic
respiratory efforts ventilatory fundopli- pneumonia fibers including clusters
at birth support cation of atrophic fibers, nu-
merous nemaline bodies
18-3 M c.654G⬎A Gly182Asp TC Isolated Failure to thrive and Frequent Gastrostomy Improvement Type I fiber predominance,
hypotonia at 3 pneumonia tube from after in- scattered atrophic fibers,
weeks, delayed during 2–18 mo fancy, now numerous nemaline bod-
motor milestones infancy 3 yr ies, majority myofibrils
intact
25-1 M c.966A⬎G Asp286Gly SC Isolated Hypotonia, joint Ventilatory Exclusively Died at 6 days Marked fiber size variabil-
contractures, femur support tube fed ity, severe atrophy both
fracture, no move- until death fiber types, numerous
ment, no respira- nemaline bodies, severe
tory effort at birth disruption of myofibrils.
35-1 F c.656C⬎G Arg183Gly SC De novo No spontaneous Needed in- Exclusively Died at age 1 Type I fiber predominance,
movements, poor termittent tube fed yr marked fiber size vari-
feeding, hypotonia, ventilatory ability, mildly increased
at birth support connective tissue, abun-
dant nemaline bodies
80-1 M c.231A⬎G Gln41Arg TC AD Delayed milestones, None None Alive and am- Mild fiber size variation,
hypotonia, weak- bulatory at many sarcoplasmic nema-
ness presenting in 56 yr line bodies.
infancy
80-2 M c.231A⬎G Gln41Arg TC AD Recurrent pulmonary None None Alive and am- Mild fiber size variation,
infections, failure bulatory at atrophy of mainly type I
to thrive, hypoto- 23 yr fibers, sarcoplasmic rods,
nia from infancy mild disorganization of
myofibrils
86-1 M c.227C⬎T His40Tyr SC De novo Hypotonia, contrac- Supplemental Mostly tube Died at 2 mo Moderate number of intra-
tures at birth oxygen for fed cytoplasmic and intranu-
cyanosis, clear nemaline bodies,
dyspnea mild increase in connec-
tive tissue, disorganized
myofibrils
90-1 M c.1226A⬎C Lys373Gln TC AD Hypotonia at birth Restrictive Tube fed Alive and am- Biopsy done in 1960s, re-
and failure to lung dis- for several bulatory at port unavailable
thrive, recurrent ease easy mos 34, surgery
pneumonia fatiguabil- for scoliosis
ity at 18 yr
90-2 M c.1226A⬎C Lys373Gln TC AD Presented in infancy None Slow feeder Alive and am- Not done
with delayed mo- bulatory at
tor milestones 7 yr
95-1 F c.338A⬎G Thr77Ala TC De novo Poor suck at birth, None Gastrostomy Alive and am- Type I fiber predominance,
failure to thrive, tube and bulatory at marked fiber-size vari-
delayed motor fundopli- 2.5 yr ability, numerous nema-
milestones cation at line bodies, myofibrils
9 mo relatively intact
103-1 M c.327A⬎T His73Leu SC De novo Hypotonia, contrac- Ventilatory Exclusively Died at 5 days Marked fiber size variabil-
tures, no spontane- support tube fed ity, type I fibers not pre-
ous movements, until death dominant, small,
no respiratory ef- rounded and atrophic
fort at birth fibers, numerous nema-
line bodies.
104-1 F c.846A⬎G Gln246Arg TC AD Difficult steps and None None Alive and am- Not done
frequent falls start- bulatory at
ing 2 yr age 9 yr
104-2 F c.846A⬎G Gln246Arg TC AD Difficult steps and None None Alive and am- Type I fiber predominance,
frequent falls in bulatory at mild fiber-size variability,
early childhood 41 yr numerous nemaline bod-
ies
104-3 F c.846A⬎G Gln246Arg TC AD Difficult steps, fre- None None Alive and am- Biopsy findings similar to
quent falls, weak- bulatory at 104-2
ness in early child- 40 yr
hood
104-4 F c.846A⬎G Gln246Arg TC AD Difficult steps and None None Alive and am- Not done
frequent falls in bulatory at
early childhood 20 yr
104-5 F c.846A⬎G Gln246Arg M AD Mild leg weakness in None None Alive and am- Not done
childhood bulatory at
15 yr

Agrawal et al: Actin ACTA1 in Nemaline Myopathy 93

Appendix. Continued

Amino Acid Disease Presenting Respiratory Feeding

Patient Sex Mutation Substitution Severitya Inheritanceb Symptomsc Issues Issues Status Pathological Findingsd

108-1 M c.510T⬎C Val134Ala IC AD Failure to thrive, hy- Nighttime Gastrostomy Alive and am- Uniform fiber size, numer-
potonia, cyanosis ventilatory feeds bulatory at ous nemaline bodies,
in infancy support 6 yr relatively intact myofi-
brils
108-4 F c.510T⬎C Val134Ala U AD Slow runner in child- None None Alive and well Not done
hood, slender at 34 yr
adult
108-7 M c.510T⬎C Val134Ala U AD Tall, thin, long face, None None Alive and well Not done
basketball player at 63 yr
115-1 F c.212G⬎C Val35Leu SC De novo Hypotonia, no spon- Needs me- Gastrostomy Unable to roll Type I fiber predominance,
taneous movement chanical feeds over or sit, atrophic type II fibers,
at birth ventilation can hold her numerous nemaline bod-
⬎20hr/day head alive at ies
18 mo
117-1 M c539delG/ trunc188/ SC AR Hypotonia, contrac- Ventilatory Tube fed Died at 2 mo Marked fiber size variation,
c.885A⬎T Glu259Val tures, no spontane- support exclusively mild-moderate increase
ous movements, until death in connective tissue, scat-
no respiratory ef- tered atrophy of both
fort at birth type fibers, numerous
nemaline bodies, in-
creased glycogen
120-1 M c.300T⬎C Ilc64Asn TC AD Feeding problems in Recently di- None Alive and am- Few nemaline bodies in
infancy, weakness agnosed bulatory at sarcoplasm, mostly intact
and low muscle with sleep 48 yr myofibrillar arrangement.
bulk apnea
120-3 F c.300T⬎C Ilc64Asn TC AD Failure to thrive, None Slow feeder Alive and am- Not done
weakness frequent bulatory at
falls during in- 8 yr has
fancy bilateral foot
drop
166-1 F c.645A⬎G Asp179Gly SC De novo Severe hypotonia, Ventilator Gastrostomy Alive at 2 yr, Marked fiber-size variabil-
inability to dependent tube for severely ity, diffuse and wide-
breathe or feed since feeding hypotonic spread both fiber type
independently, no birth; with no atrophy with round fi-
movements at tracheos- head con- bers, abundant nemaline
birth tomy trol bodies
181-1 F c.911G⬎T Gly268Cys TC Isolated Hypotonic at birth, None Poor suck, Alive and am- Not available
delayed motor slow bulatory at
milestones feeder 7 yr, wors-
ening lor-
dosis
186-1 F c.788A⬎G Met227Val IC De novo Hypotonia, delayed Needs CPAP Gastrostomy Alive and am- Not done
motor milestones, support tube for bulatory at
poor feeding since 2 yr feeding 3 yr
age
186-2 F c.788A⬎G Met227Val IC De novo Hypotonia, delayed Needs CPAP Gastrostomy Alive and am- Type I fiber predominance,
motor milestones, support tube for bulatory at numerous nemaline
poor feeding since 2 yr feeding 3 yr bodies, areas of Z-line
age streaming
188-1 M c.1134delG Gly342Ala ⫹ 75 A AD Weakness on exercise None Difficulty Alive and am- Significant disruption of
at 42 yr, now swallow- bulatory at myofibrillar architecture
difficulty climbing ing 55 yr with due to nemaline bodies,
stairs, swallowing increasing focal accumulations of
and has head weakness mitochondria
drop
188-2 M c.1134delG Gly342Ala ⫹ 75 A AD Unaffected None None Died at 85 yr Not done
of unre-
lated cause
199-1 M c.790G⬎C Met227Ile TC Isolated Onset at 2 yr with Easy fatiga- None Alive and am- Mild fiber size variation,
difficulty running, bility bulatory at subsarcolemmal nema-
climbing stairs 38 yr, sur- line bodies, slight in-
gery for crease in connective
scoliosis at tissue
18 yr
220-1 F c.332A⬎C Ile75Leu SC De novo No spontaneous Ventilator Gastrostomy Alive at 1.8 yr, Not available
movement, no dependent tube for not able to
respiratory effort since feeding hold head/
at birth birth sit
221-1 M c.944T⬎C Tyr279His IC De novo Hypotonia, feeding Apnea, bra- Tube for Died at 9 mo Type I fiber predominance,
difficulty dycardias feeding abundant cytoplasmic
since age rods, streaming and
3 wk multifocal enlargement
of Z lines
226-1 F c.546G⬎A Gly146Asp SC De novo Contractures, mark- Intermittent Tube feeds Died at 6 mo Type I fiber predominance,
edly reduced ventila- and scattered atrophy, small
movements, hy- tion for gtube masses of Z-band mate-
potonia at birth apnea and feeds rial resembling mini-
pneumo- nemaline bodies on
nia EM. Strong staining
with actin immunoper-
oxidase seen as subsar-
colemmal ring distribu-
tion in most fibers

94 Annals of Neurology Vol 56 No 1 July 2004

Appendix. Continued

Amino Acid Disease Presenting Respiratory Feeding

Patient Sex Mutation Substitution Severitya Inheritanceb Symptomsc Issues Issues Status Pathological Findingsd

234-1 F c.861G⬎A Gly251Asp SC De novo No spontaneous Needs me- Tube fed Alive at 8 yr, Type I fiber predominance,
movements chanical since unable to 50% fibers abnormally
ventila- birth hold head small, several nemaline
tion 24 or sit with- bodies, variable disorga-
hr/day out support nization of myofibrils
since 8
mo
239-1 M c.780A ⬎ G Glu224Gly TC De novo Presented in infancy None None Alive at 3 yr Less than 10% fibers con-
with hypotonia, tain nemaline bodies
failure to thrive,
limb girdle and
bulbar weakness
307-1 M c.327A⬎G His73Arg SC Isolated Severe hypotonia, Ventilator Tube fed Alive at 11 yr Marked variation in fiber
insufficient respi- dependent since with severe size, several rounded fi-
ration since birth birth weakness bers, abundant nemaline
bodies
308-1 F c.222C⬎T Pro38Leu SC De novo Severe hypotonia, Ventilatory Tube fed Died at 15 Marked fiber size variabil-
insufficient respira- support until days ity, diffuse atrophy, ex-
tion, no spontane- until death death tensive perimysial and
ous movements at endomysial fibrosis, nu-
birth merous nemaline bodies
349-1 M c.356G⬎A Glu83Lys TC De novo Hypotonia noted at None None Alive and am- Not done
3 mo bulatory at
3 yr
349-2 M c.356G⬎A Glu83Lys TC De novo Hypotonia at birth, None None Alive and am- Marked type 1 fiber pre-
had delayed mile- bulatory at dominance, moderate
stones, sat at 9 mo 3 yr fiber-size variability, scat-
and walked at 16 tered and grouped atro-
mo phy, numerous nemaline
bodies
350-1 F c.698G⬎A Gly197Ser SC De novo Hypotonia, no respi- Ventilator Tube fed Died at 3 mo Marked type I fiber pre-
ratory efforts dependent since dominance, hypotrophic
since birth birth both type fibers, abun-
dant nemaline bodies

a
SC ⫽ severe congenital; IC ⫽ intermediate congenital; TC ⫽ typical congenital; IC ⫽ intermediate congenital; M ⫽ mild (childhood or
juvenile onset), A ⫽ adult onset, U ⫽ unaffected.
bAD ⫽ autosomal dominant; AR ⫽ autosomal recessive; de novo ⫽ indicates sporadic mutation ruled out in both parents; isolated, sporadic

mutation, not ruled out in both parents because of DNA nonavailability.

c
CPAP ⫽ continuous positive airway pressure.
d
EM ⫽ electron microscopy.

3. Jungbluth H, Sewry CA, Brown SC, et al. Mild phenotype of

This work was supported by NIH (National Institute of Arthritis
and Musculoskeletal and Skin Diseases R01-AR44345, A.H.B.), the
Muscular Dystrophy Association of the USA (A.H.B.), the Joshua
Frase Foundation (A.H.B.), the Lee and Penny Anderson Family
Foundation (A.H.B.), and the Australian National Health and Med-
ical Research Council Fellowship (139170 and 139039, N.G.L.).
DNA sequencing was performed by the Children’s Hospital Boston
Genomics Program, Mental Retardation Research Center core DNA
sequencing facility supported by the NIH (National Institute of
Child Health and Human Development grant, P30-HD18655).
We thank the many referring physicians and other health care pro-
fessionals for their outstanding help and support enrolling study
subjects and contributing clinical and pathological data. Special
thanks to the patients and their families for helping advance our
knowledge about nemaline myopathy. We also thank D. Sanoudou
for valuable suggestions and N. Dexter for genetic counseling.
References
1. North KN, Laing NG, Wallgren-Pettersson C. Nemaline
myopathy: current concepts. The ENMC International Consor-
tium and Nemaline Myopathy. J Med Genet 1997;34:
705–713.
2. Sanoudou D, Beggs AH. Clinical and genetic heterogeneity in
nemaline myopathy—a disease of skeletal muscle thin filaments.
Trends Mol Med 2001;7:362–368.
nemaline myopathy with sleep hypoventilation due to a muta-
tion in the skeletal muscle alpha-actin (ACTA1) gene. Neuro-
muscul Disord 2001;11:35– 40.
4. Ryan MM, Schnell C, Strickland CD, et al. Nemaline
myopathy: a clinical study of 143 cases. Ann Neurol 2001;50:
312–320.
5. Lammens M, Moerman P, Fryns JP, et al. Fetal akinesia se-
quence caused by nemaline myopathy. Neuropediatrics 1997;
28:116 –119.
6. Wallgren-Pettersson C, Pelin K, Hilpela P, et al. Clinical and
genetic heterogeneity in autosomal recessive nemaline myop-
athy. Neuromuscul Disord 1999;9:564 –572.
7. Ryan MM, Ilkovski B, Strickland CD, et al. Clinical course
correlates poorly with muscle pathology in nemaline myopathy.
Neurology 2003;60:665– 673.
8. Pelin K, Hilpela P, Donner K, et al. Mutations in the nebulin
gene associated with autosomal recessive nemaline myopathy.
Proc Natl Acad Sci USA 1999;96:2305–2310.
9. Pelin K, Donner K, Holmberg M, et al. Nebulin mutations in
autosomal recessive nemaline myopathy: an update. Neuromus-
cul Disord 2002;12:680 – 686.
10. Wallgren-Pettersson C, Donner K, Sewry C, et al. Mutations in
the nebulin gene can cause severe congenital nemaline myop-
athy. Neuromuscul Disord 2002;12:674 – 679.
11. Laing NG, Wilton SD, Akkari PA, et al. A mutation in
the alpha tropomyosin gene TPM3 associated with autoso-
mal dominant nemaline myopathy. Nat Genet 1995;9:75–
79.

Agrawal et al: Actin ACTA1 in Nemaline Myopathy 95

12. Tan P, Briner J, Boltshauser E, et al. Homozygosity for a non-
sense mutation in the alpha-tropomyosin slow gene TPM3 in a
patient with severe infantile nemaline myopathy. Neuromuscul
Disord 1999;9:573–579.
13. Wattanasirichaigoon D, Swoboda KJ, Takada F, et al. Mutations
of the slow muscle alpha-tropomyosin gene, TPM3, are a rare
cause of nemaline myopathy. Neurology 2002;59:613– 617.
14. Johnston JJ, Kelley RI, Crawford TO, et al. A novel nemaline
myopathy in the Amish caused by a mutation in troponin T1.
Am J Hum Genet 2000;67:814 – 821.
15. Donner K, Ollikainen M, Ridanpaa M, et al. Mutations in the
beta-tropomyosin (TPM2) gene—a rare cause of nemaline my-
opathy. Neuromuscul Disord 2002;12:151–158.
16. Nowak KJ, Wattanasirichaigoon D, Goebel HH, et al. Mutations
in the skeletal muscle alpha-actin gene in patients with actin my-
opathy and nemaline myopathy. Nat Genet 1999;23:208–212.
17. Ilkovski B, Cooper ST, Nowak K, et al. Nemaline myopathy
caused by mutations in the muscle alpha-skeletal-actin gene.
Am J Hum Genet 2001;68:1333–1343.
18. Sparrow JC, Nowak KJ, Durling HJ, et al. Muscle disease
caused by mutations in the skeletal muscle alpha-actin gene
(ACTA1). Neuromuscul Disord 2003;13:519 –531.
19. Wallgren-Pettersson C, Laing NG. Report of the 70th ENMC
International Workshop: nemaline myopathy, 11–13 June
1999, Naarden, The Netherlands. Neuromuscul Disord 2000;
10:299 –306.
20. Feinberg DM, Spiro AJ, Weidenheim KM. Distinct light mi-
croscopic changes in human immunodeficiency virus-associated
nemaline myopathy. Neurology 1998;50:529 –531.
21. Duggal P, Vesely MR, Wattanasirichaigoon D, et al. Mutation
of the gene for IsK associated with both Jervell and Lange-
Nielsen and Romano-Ward forms of Long-QT syndrome. Cir-
culation 1998;97:142–146.
22. North KN, Beggs AH. Deficiency of a skeletal muscle isoform of
alpha-actinin (alpha-actinin-3) in merosin-positive congenital
muscular dystrophy. Neuromuscul Disord 1996;6:229 –235.
96 Annals of Neurology Vol 56 No 1 July 2004
23. Chan Y, Tong HQ, Beggs AH, Kunkel LM. Human skeletal
muscle-specific alpha-actinin-2 and -3 isoforms form ho-
modimers and heterodimers in vitro and in vivo. Biochem Bio-
phys Res Commun 1998;248:134 –139.
24. Tomczak KK, Marinescu VD, Ramoni MF, et al. Expression
profiling and identification of novel genes involved in myogenic
differentiation. FASEB J 2004;18:403– 405.
25. Goebel HH, Anderson JR, Hubner C, et al. Congenital myop-
athy with excess of thin myofilaments. Neuromuscul Disord
1997;7:160 –168.
26. Sanoudou D, Frieden LA, Haslett JN, et al. Molecular classifi-
cation of nemaline myopathies: “nontyping” specimens exhibit
unique patterns of gene expression. Neurobiol Dis 2004;15:
590 – 600.
27. Franke WW, Stehr S, Stumpp S, et al. Specific immunohisto-
chemical detection of cardiac/fetal alpha-actin in human cardi-
omyocytes and regenerating skeletal muscle cells. Differentia-
tion 1996;60:245–250.
28. Collins T, Joya JE, Arkell RM, et al. Reappearance of the mi-
nor alpha-sarcomeric actins in postnatal muscle. Am J Physiol
1997;273:C1801–C1810.
29. Sheterline P, Clayton J, Sparrow JC. Actin. 4th ed. Oxford:
Oxford University Press, 1998.
30. Marston S, Mirza M, Abdulrazzak H, Sewry C. Functional
characterisation of a mutant actin (Met132Val) from a patient
with nemaline myopathy. Neuromuscul Disord 2004;14:
167–174.
31. Corbett MA, Robinson CS, Dunglison GF, et al. A mutation
in alpha-tropomyosin(slow) affects muscle strength, maturation
and hypertrophy in a mouse model for nemaline myopathy.
Hum Mol Genet 2001;10:317–328.
32. Wertman KF, Drubin DG, Botstein D. Systematic mutational
analysis of the yeast ACT1 gene. Genetics 1992;132:337–
350.