Brain Advance Access published June 23, 2003

DOI: 10.1093/brain/awg202 Brain (2003), 126, Page 1 of 11

Clinical, electrophysiological and morphological

®ndings of Charcot±Marie±Tooth neuropathy with
vocal cord palsy and mutations in the GDAP1 gene
Teresa Sevilla,1 Ana Cuesta,4 MarõÂa Jose Chumillas,2 Fernando Mayordomo,3 Laia Pedrola,4
Francesc Palau4 and Juan J. VõÂlchez1

Departments of 1Neurology, 2Clinical Neurophysiology and Correspondence to: Dr Teresa Sevilla, Servicio de
3Experimental Cellular Pathology, University Hospital La NeurologõÂa, Hospital Universitari La Fe, Avenida
Fe, and 4Laboratory of Genetics and Molecular Medicine, Campanar 21, 46009 Valencia, Spain
Instituto de Biomedicina, Consejo Superior de E-mail: teresevillaus@yahoo.com

Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016

Investigaciones Cientõ®cas (CSIC), Valencia, Spain

Summary
Three Spanish families with an autosomal recessive
severe hereditary motor and sensory neuropathy, show-
ing mutations in the ganglioside-induced-differentiation-
associated protein 1 (GDAP1) gene in the Charcot±
Marie±Tooth (CMT) type 4A locus were studied. The
disorder started in the neonatal period or early infancy
with weakness and wasting of the feet and, sub-
sequently, involvement of the hands, causing severe
disability. By the late teens, some patients developed a
hoarse voice and vocal cord paresis. Peripheral motor
nerve conduction velocity (MNCV) could not be meas-
ured in many cases because of the absence of muscle
response due to distal atrophy. However, latencies to
proximal muscles were in the normal range; median
MNCV was >40 m/s in those cases in which it could be
measured. Sural nerve biopsy from two patients showed
a pronounced depletion of myelinated ®bres, regenera-
tive clusters and signs of axonal atrophy. Additionally,
a small proportion of thin myelinated ®bres and pro-
liferation of Schwann cells forming onion bulb struc-
tures were also found. Unmyelinated ®bre population
was markedly increased. These ®ndings are indicative
of a predominant axonal degeneration with some
demyelinating features. These Spanish families share in
the severe CMT clinical phenotype with some Tunisian
families who also presented mutations in the GDAP1
gene and to which the CMT4A locus was originally
assigned. However, our families differ in the presence
of laryngeal involvement and values of MNCV and
pathological features are more in line with CMT2 type.
The possibility that GDAP1 gene mutations could be
expressed under different phenotypes is a question to be
resolved.

Keywords: Charcot±Marie±Tooth disease type 2; hereditary motor and sensory neuropathy type II; vocal cord paresis,
autosomal recessive; GDAP1 gene

Abbreviations: CMAP = compound muscle action potential; GDAP1 = ganglioside-induced differentiation-associated

protein 1; CMT = Charcot±Marie±Tooth; DL = distal latency; GJB1 = gap junction protein beta-1; HMSN = hereditary
motor and sensory neuropathy; MRC = Medical Research Council; MCV = motor conduction velocity; MNCV = motor
nerve conduction velocity; MPZ = myelin protein zero; NCS = nerve conduction studies; OB = onion bulbs;
SNCV = sensory nerve conduction velocity; SNAPs = sensory nerve action potentials.

Introduction
Hereditary motor and sensory neuropathy (HMSN) or
Charcot±Marie±Tooth disease (CMT) was classically
grouped into two main categories according to electro-
physiological and nerve biopsy ®ndings: (i) CMT1 showing a
median nerve motor conduction velocity (MCV) of <38 m/s
and nerve ®bre demyelination with proliferation of Schwann
Brain 126 ã Guarantors of Brain 2003; all rights reserved
cells forming onion bulbs (OB); and (ii) CMT2 with normal
or near normal conduction velocities and pathological signs
of axonal degeneration and regeneration (Harding and
Thomas, 1980a; Dyck and Lambert, 1968a, b). Other authors
de®ned an intermediate group with conduction velocities
ranging between 27 and 35 m/s, which did not gain wide
Page 2 of 11 T. Sevilla et al.
acceptance at ®rst. Dejerine±Sottas syndrome or CMT3
category was applied to early childhood-onset cases with
sporadic or recessive transmission, presenting very slow
motor nerve conduction velocity (MNCV) and severe
demyelinating features (Dyck et al., 1993). A dominant
X-linked category was also recognized and was included in
the CMT1 group (Rozear et al 1987; Hahn et al., 1990).
Examples of autosomal recessive CMT either demyelinating
or axonal forms were also reported (Harding and Thomas,
1980b; Ouvrier et al., 1981; GabreeÈls-Festen et al., 1991).
Genetic linkage studies discovered that HMSN categories
were heterogeneous and that the CMT1 and CMT2 forms
were sub-classi®ed according to loci ascription (Reilly 2000;
Boerkoel et al., 2002a). Furthermore, a CMT4 group emerged
to include autosomal recessive CMT forms (Ben Othmane
et al., 1993) and was rapidly enlarged with various
subcategories (Bolino et al., 1996; Kalaydjieva et al., 1996;
LeGuern et al., 1996; Delague et al., 2000). The discovery of
the gene products lead to a rational approach to the study of
the pathogenesis of the various CMT forms (Shy et al., 2002)
but, at the same time, it has added complexity to the
traditional concepts used to classify HMSN. For example, it
has been shown that the Dejerine±Sottas phenotype may
result from mutation of different genes (Hayasaka et al.,
1993; Ionasescu et al., 1995; Bort et al., 1998; Timmerman
et al., 1999). Conversely, mutations in certain genes like Gap
junction protein beta-1 (GJB1) gene or myelin protein zero
(MPZ) may produce demyelinating or axonal phenotypes
(De Jonghe et al., 1999; Misu et al., 2000; Dubourg et al.,
2001; Young et al., 2001; Boerkoel et al., 2002b).
Recently, two research groups simultaneously described
CMT patients presenting mutations in the GDAP1 gene. In
one case (Baxter et al., 2002), the mutations were found in
Tunisian families and had been previously reported as
CMT4A form and de®ned as a severe demyelinating
neuropathy. The other group studied (Cuesta et al., 2002)
consisted of Spanish families who also presented a severe
clinical phenotype, but whose nerve conduction velocities
and pathological pattern were consistent with an axonal
neuropathy. Clinical, electrophysiological and nerve biopsy
®ndings of the GDAP1 gene-related neuropathy in patients
from these Spanish families are reported in detail in this
study.
Patients and methods
Clinical and electrophysiological examinations were carried
out on 18 affected and non-affected members of three
families: LF249, LF20 and LF38. Neuropathic symptoms and
de®cits were assessed by two of us (T.S. and J.J.V.). Genetic
studies con®rmed that all patients were either homozygous or
compound heterozygous for mutations in the GDAP1 gene
(Cuesta et al., 2002).
Nerve conduction studies (NCS) were tested with surface
electrodes. Amplitudes of compound muscle action potential
(CMAPs), distal latency (DL) and conduction velocity were
recorded whenever possible from median, ulnar, peroneal,
tibial and axillary nerves using conventional methods.
Furthermore, motor nerve conduction studies of more
proximal upper limb muscles like the palmaris longus muscle
for the median nerve and ¯exor carpis ulnaris for the ulnar
nerve were also tested. CMAP and DL from the diaphragm
muscle were recorded by using phrenic nerve stimulation in
the neck (Bolton, 1993). Recordings of sensory nerve action
potentials (SNAPs) from median and ulnar nerves were
performed orthodromically, but sural nerve was tested
antidromically. Concentric needle electromyography was
performed in the proximal and distal muscles of the upper
and lower limbs.
Sural nerve biopsy was performed at ankle level after
written consent. The sample was ®xed in 2.5%
Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016
glutaraldehyde±1% paraformaldehyde in 0.1M sodium
cacodilate and post-®xed in 1% osmium tetraoxide, dehy-
drated in graded acetones and embedded in epoxy resin.
Semi-thin sections stained with toluidine blue were prepared
for evaluation under a light microscope. Ultrathin cut samples
were contrasted with uranyl acetate and lead citrate for ultra
structural study. Morphometry of myelinated ®bres was
performed on high-resolution (1600 3 1200 pixels) micro-
graphic images obtained with a Polaroid DMC digital camera.
Several pictures were obtained from each nerve fascicle; care
was taken to ensure they did not contain overlapping ®elds.
Measurement data were collected by means of Scion image
analyst software (http://www.scioncorp.com). A threshold
grey-scale level for myelin and axon pro®les was de®ned and
used to trace the border of the axon and outer edge of myelin;
the software could then calculate their respective areas and
diameters. To minimize the effects of shrinkage or irregular
shape of the myelin ring, the mean of major and minor ellipse
diameter was chosen. G-ratio was calculated by the quotient
axon diameter/myelinated ®bre diameter. OB formation was
considered as supernumeraryÐmore than one- layer of
Schwann cells completing a full circle. Regenerative cluster
was de®ned as two or more closely grouped small myelinated
®bres in a delimited circular area. Quanti®cation was
expressed in terms of density or number of axons per mm2
and also as the percentage of myelinated ®bres involved in
OB or the cluster with respect to the total count. Electron
microphotographs were taken randomly. Abnormal forma-
tions like OB, cluster or demyelinated axons were assessed at
low power magni®cation (33000). Pictures at higher
magni®cation (39000) were digitalized and used for
quanti®cation of unmyelinated ®bres. Presence of structural
abnormalities of myelin, axon, abnormal deposits in Schwann
cell cytoplasm or macrophages, as well as the quantity of
endoneural collagen was evaluated by inspection at the
appropriate ampli®cation. Microphotograph measurement
calibration was performed by mean of stage micrometer,
while electron micrograph magni®cation was calibrated by
rule diffraction cross grating replica, both from Agar
Scienti®c Ltd (Stansted, Essex, UK).
 GADP1 neuropathy and vocal cord palsy Page 3 of 11

Results weakness was a symptom reported during those ®rst years.

Clinical features
Fig. 1 represents the pedigrees of the three families and
Table 1 summarizes clinical data.
Kindred 1 (LF249)
The disorder in this family is inherited as an autosomal
recessive form. There are three brothers affected out of six
siblings (Fig. 1). The father died at 56 years of cancer; he had
no neurological abnormalities at that time. The mother was
clinically and electrophysiologically unaffected.
Patients started to walk at a normal age. Gait was clumsy
and they had frequent falls. Each patient underwent ortho-
paedic surgery in his very ®rst years of life before the
Patients were still able to walk at the end of their ®rst decade,
but all of them became chair-bound after the age of 12 years.
Most patients were unable to remember at what speci®c age
their voice changed, but it was usually in the second decade.
Intelligence was normal in all patients.
Clinical assessment was scored according to MRC
(Medical Research Council, UK) scale. All patients showed
complete paralysis in the muscles below the knee and marked
weakness in the thigh muscles (3/5 according to the MRC
scale). Complete paralysis of the distal muscles groups was
found (0/5 according to the MRC scale) in the upper limbs
while ¯exion and extension of elbow was possible against
mild resistance (4/5 according to the MRC scale). Strength in
shoulder muscles was normal except in proband one. Spinal
deformity was present in all patients and chest deformity in

Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016

diagnosis of a peripheral neuropathy was established. Hand
the proband only. Pes cavus, claw hands and contractures
were also prominent. Cranial nerves were not involved except
for hoarseness. Muscle stretch re¯exes were absent. All
modalities of sensation were affected. Pinprick and vibration
were absent at hallux, ankle and knee; joint position sense
was also absent at distal interphalangeal joints and ankle.
Pinprick and vibration were also diminished in hands,but
joint position sense was preserved. There was no clinical
variability between the affected members (Table 1). The CSF
protein determined in the proband resulted normal.

Kindred 2 (LF20)
There are two brothers affected out of four siblings in this
family (Fig. 1). Parents were clinically unaffected. The
electrodiagnostic examination was normal in both parents.
The propositus is a 39-year-old man. He was a ¯oppy
infant with delayed motor milestones who started walking at
the age of 2 years and began to have frequent falls. He had to
use orthopaedic devices. Dif®culty manipulating objects with
Fig. 1 Pedigree of families. the hands was a prominent symptom from infancy.

Table 1 Clinical features

Case F1-II-2 F1-II-4 F1-II5 F2-II-3 F2-II-4 F3-IV-9 F3-IV-14 F3-IV-10 F3-IV-11

Year of birth 1958 1962 1965 1965 1970 1946 1955 1949 1959
Age of walking 12 months 15 months 12 months Delayed 18 months Delayed Normal Normal Normal
Age at onset 18 months 2 years Birth Birth Birth 2 years 2 years 2 years 2 years
Proximal UL weakness + ++ + + + + + + +
Distal UL weakness +++ +++ +++ +++ +++ +++ +++ +++ +++
Proximal LL weakness ++ ++ ++ ++ ++ ++ + ++ ++
Distal LL weakness +++ +++ +++ +++ +++ +++ +++ +++ +++
Hoarseness Yes Yes Yes Yes Yes No No Yes No
Sensory loss in hands P, V, T P, V, T P, V, T P, V, T P, V, T P, V, T P, V, T P, V, T P, V, T
Sensory loss in feet All All All All All All All All All
Re¯exes Absent Absent Absent Absent Absent Absent Absent Absent Absent

Muscle weakness in upper limbs (UL): + = strength 4/5 on MRC scale; ++ = strength <4/5 on MRC scale; +++ = complete paralysis.
Muscle weakness in lower limbs (LL): + = 4/5 on MRC scale;, ++ = < 4/5 on MRC scale; +++ = complete paralysis. F1 = family LF249;
F2 = family LF20; F3 = family LF38; Sensory changes: P, V, T = decreased pinprick, vibration and touch; all = absent pinprick,
vibration, touch and position sense.
Page 4 of 11 T. Sevilla et al.
Intelligence was normal. He continued walking with the
assistance of crutches until the age of 9 years. At this age, he
became wheelchair dependent with marked weakness and
atrophy in upper and lower limbs. Clinical examination was
nearly identical to members of family LF249 (Table 1).
Hoarse voice had been present since 14 years of age. An
otolaryngologist studied the vocal cords and found paresis in
both cords, the left one being more affected. Symptoms and
clinical examination of his brother were identical (Table 1).
Kindred 3 (LF38)
This is a large family with nine affected members from four
consanguineous marriages (Fig. 1). We have examined only
four patients from this family. Their clinical data are
summarized in Table 1. They are similar to those described
for members of the other families, but the disease evolved at a
slower pace. Indeed, the members of this family only needed
crutches around age 18 years and become chair-bound around
30 years. One member of this family could even walk a few
metres with supports at 43 years. Two members had had
hoarseness since they were aged 30 years.
Electrophysiological studies
Table 2 summarizes EMG and NCV studies of the three
families. EMG records from members of families LF249 and
LF20 who were tested by us showed absence of motor unit
action potentials (MUAPs) during voluntary contraction and
®brillation density reduced in all lower limb and distal upper
limb muscles. In proximal upper limb muscles such as the
deltoid and biceps, needle EMG showed evidence of chronic
denervation with spontaneous ®brillation potentials, large
polyphasic potentials and reduced recruitment pattern.
Complex repetitive discharges were recorded in some
instances.
MNCVs were impossible to obtain in many instances. In
affected members from families LF249 and LF20, peroneal,
tibial, femoral, median, and ulnar CMAPs were not obtained
on recording both proximal and distal innervated muscles.
Axillary and phrenic nerve latencies were preserved in all
tested cases, but some cases showed low CMAPs (Table 2). In
two members of the LF38 family whose median motor NCV
Table 2 Electrophysiological data
Axillary Phrenic
CMAP DL CMAP DL
LF20-II-4 13.3 3.7 0.1 6.5
LF249-II-3 4.5 3.2 0.1 4.8
LF38-IV-1 NP NP NP NP
LF38-V-14 NP NP NP NP
Normal MCVs: motor median and ulnar nerves >51 m/s. Normal CMAP: axillary >6 mV; phrenic >0.3 mV; median >9 mV; ulnar
>7.7 mV. Normal DLs: axillary <5.3 ms;, phrenic <7.9 ms; median <4.1 ms; ulnar < 3.3 ms. NP = not performed; NR = no response.
could be obtained, it was in the normal or near-normal range
(Table 2).
Sural SNAP were absent in all patients. Amplitude of
SNAPs and sensory nerve conduction velocities (SNCVs)
from median nerve was obtained in two patients (LF249 II-4
and LF20 IV-14). The values of SNAPs were 1.3 and 0.2 mV,
respectively (normal >16.5 mV); and SNCVs were 37 and
46.3 m/s, respectively (normal >43 m/s).
Sural nerve biopsies
Sural nerve biopsies were performed in probands from LF20
and LF249 families when they were 22 and 19 years old,
respectively. A control nerve was obtained from a 27 year-old
multiorgan donor without neuropathic or systemic disease
Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016
history. Morphometric data are reported in Table 3. Semi-thin
sections revealed a marked loss of myelinated ®bres in both
nerves; this was more pronounced in the LF20 proband
patient (Fig. 2). Histograms of myelinated ®bre size distri-
bution (Fig. 3) showed a severe reduction of ®bres of all sizes.
No ®bres >7 mm were observed in either nerve, yet both
nerves displayed a population of very small myelinated ®bres
<2 mm, which were not present in the control nerve and
probably represented regenerating sprouts.
OB formation and regenerative cluster were occasionally
present in both nerves, being rather more prominent in the
LF249 proband nerve (Table 3). Most OB surrounded
regenerative clusters; less often, they contained only a single
myelinated axon (Figs 2 and 4). In detailed electron
microscopic views, OBs were mainly made up of concen-
trically proliferated Schwann cells processes adopting a
crescent shape. In general, the number of OB layers was small
but thick in most cases, containing abundant Schwann cell
cytoplasm and enclosing numerous unmyelinated axons
(Fig. 5A). Basal membrane layers were not perceived.
Myelin thickness was proportional to axon size in the
majority of ®bres from both nerves according to g-ratio
®gures (Table 3). However, the nerve of the LF20 proband
presented a considerable proportion (20%) of myelinated
®bres with a g ratio <0.4, indicating axonal atrophy; in turn,
the nerve of the LF249 proband showed an increase (15%) of
hypomyelinated ®bres with g-ratio >0.7.
Median Ulnar
CMAP MCV DL CMAP MCV DL
NR ± ± NR ± ±
NR ± ± NR ± ±
1.4 41 5.1 0.8 42 4.4
0.8 44.6 3.9 0.9 43.7 2.9
 GADP1 neuropathy and vocal cord palsy Page 5 of 11

Myelinated ®bres often adopted a distorted (not round)
shape like boomerang forms and occasionally displayed
internal or external myelin folding; only a tomaculum-like
formation was seen. These ®ndings were more prominent in
the nerve of the LF20 patient and they were considered as
indicative of myelin adaptation to axonal atrophy.
Neither demyelinated axons nor abnormalities in myelin
compaction were observed. Only rare ®bres were seen to
undergo active axonal degeneration (Fig. 5B).
Most of the intrafascicular compartment of the nerve was
occupied by collections of Schwann cells embedded in dense
endoneural collagen deposits. These collections of Schwann
cells sometimes showed a minor degree of concentric
orientation, indicating their derivation from former OB
(Fig. 6A), but this was not evident in the majority of cases
(Fig. 6B). The density of unmyelinated ®bres was higher
(Table 3), although the counting undoubtedly included
numerous non-myelinated regenerating sprouts derived
from myelinated ®bres.
In summary, there were some quantitative and qualitative
differences in both nerve biopsies. Nerve LF20 proband
presented extreme depletion of myelinated ®bres and many of
the remaining ®bres showed signs of axonal atrophy. Nerve
LF249 proband showed a less severe axonal loss, but at the

Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016

Table 3 Morphometric ®ndings in sural nerve
LF20-II-3 LF249-II-2 Control

Myelinated ®bre density (number/mm2) 998 1211 9003

Regenerative cluster:
density (number/mm2) 105 185 ±
% of myelinated ®bres 10.5 15 ±
Onion bulbs:
density (number/m2) 89 123 ±
% of myelinated ®bres 9 10 ±
g-ratioa:
<0.4 (%) 22.6 5 3
>0.7 (%) 2.5 15 5
Unmyelinated ®bre density (number/mm2) 53328 66600 28400
aAxon diameter/total ®bre diameter

Fig. 2 Semi-thin transverse section through sural nerve from Patient LF20 proband (A) and Patient LF249 proband (B)showing a
pronounced depletion of myelinated ®bres. Remaining ®bres are of very small size sometimes assembled in regenerative clusters. *Note
the proliferation of Schwann cells in circular fashion forming OB structures, particularly around cluster (black arrowhead). Some ®bres
are thinly myelinated (open arrowheads). Bar = 10 mm.
Page 6 of 11 T. Sevilla et al.

Fig. 3 Histograms of sural nerve myelinated ®bre size distribution of Patients LF20 and LF249 (®lled bars) compared with a control (open
bars). Patients show a remarkable global reduction of myelinated ®bres and absence of ®bres with diameter size >7 mm. Also note the
presence of very small ®bres (1±2 mm) not present in the control subjects.

Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016

Fig. 4 Low magni®cation electron micrograph of a transverse section through the sural nerve from Patient LF249 proband showing a few
myelinated ®bres often adopting irregular shapes. Some Schwann cell processes proliferate around a single myelinated ®bre forming an
incipient onion bulb (open arrowhead) and proliferate more profusely around a cluster (black arrowhead). Bar = 2 mm.

expense of regenerative sprouting, and a noticeable propor-
tion of hypomyelinated ®bres and abnormal Schwann cell
proliferation forming OB as well.
Discussion
The patients described in this report suffer from a chronically
progressive motor and sensory neuropathy beginning in early
childhood and resulting in severe disability at the end of their
®rst decade in LF249 and LF20 families, and at the third
decade in the large LF38 family. All of them presented
mutations in the GDAP1 gene at the CMT4A locus:
homozygotes for Q163X mutation in LF38 or compound
heterozygotes for Q163X and S194X in LF249 and for
Q163X and T288fsX290 in LF20 (Cuesta et al., 2002). Their
clinical picture is similar to a series of Tunisian families also

Fig. 5 Higher magni®cation electron micrograph from the specimen shown in Fig. 4. (A) A cluster of small myelinated ®bres is
surrounded by a concentric array of Schwann cell processes with which numerous unmyelinated axons are associated. (B) A myelinated
®bre undergoing active axonal degeneration along with several groups of unmyelinated ®bres fully encircled by a single Schwann cell.
Bars = 2 mm.
harbouring mutations in the same gene and previously
reported as corresponding to the CMT4A category (Baxter
et al., 2002). However, our Spanish patients have two
particular features: appearance of hoarseness due to vocal
cord paresis, and electrophysiological and pathological
evidence of an axonal neuropathyÐthereby differing from
the concept of CMT4A as a severe demyelinating neuropathy
(Ben Othmane et al., 1993).
Hoarseness was a frequent symptom. It appeared late in the
course of the disease in many affected members of our three
families. Moreover, subclinical phrenic nerve impairment
was also detected in some cases. Involvement of recurrent
laryngeal and phrenic nerves does not seem to be a speci®c
hallmark of GDAP1-related neuropathy. Vocal cord involve-
ment has been reported to occur in different types of
hereditary neuropathies like certain forms of distal motor
neuropathy (Young and Harper,1980), CMT2C (Dyck et al.,
1994; Yoshioka et al., 1996) and some CMT1 cases
associated with either peripheral myelin protein (PMP22)
(Thomas et al., 1997) or early growth response 2 (EGR2)
gene mutations (Pareyson et al., 2000). It has been proposed
that those nerves are rather long and the progression of a
length-dependent severe neuropathy might explain their
involvement (Thomas et al., 1997); other factors that may
predispose to this particular regional involvement are not yet
known.
Median NCV appeared >40 m/s in cases where it could be
recorded. In other patients, it was not obtained due to a severe
distal muscular atrophy of hand muscles, but proximal motor
latencies were always preserved. Assuming that nerve
GADP1 neuropathy and vocal cord palsy Page 7 of 11
Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016
conduction velocity is uniform along whole nerve in CMT
demyelinating neuropathies (Kaku et al., 1993; Krajewski
et al., 2000), our data are indicative of an axonal neuropathy
or CMT2 type.
In a detailed analysis of the Tunisian patients harbouring
GDAP1 gene mutations, the reported MNCV ranged between
27±35 m/s (Baxter et al., 2002), which appears less slower
than would be expected in a severe infantile hypo or
demyelinanting neuropathy (GabreeÈls-Festen et al., 1990).
Such values can be considered in the intermediate range that
may be displayed by distinct CMT forms; some considered
primarily demyelinating like GJB1 related cases (Nicholson
and Nash, 1993; Dubourg et al., 2001) and others reported as
axonal forms such as the recessive russe type neuropathy
(Thomas et al., 2001). In the latest example, loss or damage to
large calibre axons could be responsible for slowing conduc-
tion velocity to such a degree. The fact that some Tunisian
patients showed a marked decrease of CMAP (0.3±0.2 mV)
(Hentati et al., 2001) may indicate that axonal loss could have
had some in¯uence on nerve conduction results. In cases like
this, testing of proximal motor latencies can give substantial
information about the matter.
Sural nerve biopsy data from two of our patients showed
that axonal loss was the most prominent ®nding. The presence
of axonal atrophy and ®bres undergoing axonal degeneration
corroborate `axonal' as the main pathological feature. The
surviving myelinated ®bres corresponded to the small size
population showing large ®bre vulnerability and, probably,
replacement by regenerating ®bres; in fact, cluster formation
of small myelinated ®bres were quite prominent, particularly
 Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016

T. Sevilla et al.
Page 8 of 11

in the nerve of the LF20 propositus. The notable increase of
myelinated ®bre population is also indicative of regenerating
axonal sprout not reaching the myelinated condition. In
addition both nerves, particularly that of the LF249 proband,
showed a proportion of thinly myelinated ®bres and OB
formation that may be considered as demyelinating features.
However, although teased ®bre studies were not performed,
the absence of demyelinated axon in transverse nerve sections
precludes demyelination and secondary remyelination as the
most probable source of hypomyelinated ®bres. Many of
these ®bres may correspond to myelinated axonal sprouts or
they could even result from a developmental myelination
arrest.
OBs were quantitatively less frequent and qualitatively
different to those observed in typical CMT1 neuropathies
(Ouvrier et al., 1981; GabreeÈls-Festen et al., 1992; Thomas
et al., 1997). In our patients, most OBs were related to a
regenerating cluster. The presence of OB formation is
indicative of Schwann cell dysfunction and it usually appears
in long-lasting demyelinating effects. Additionally, in experi-
mental xenograft models, it has been demonstrated that
abnormal Schwann cells tend to form OB when they try to
myelinate axonal sprouting (Sahenk et al., 1999).
The nerve biopsy report of the Tunisian patients mentions a
signi®cant decrease in myelinated ®bre density, predomin-
antly in those of a large size. Apart from the presence of
segmental demyelination in teasing preparation, the presence
of hypomyelinated ®bres and OB formation are also
mentioned, but no quanti®cation or detailed descriptions are
available.
According to the histopathological data, the GADP1
alterations can interfere with axonal survival and induce an
abnormal Schwann cell behaviour with minor or mild
repercussion in myelin compaction. GDAP1 gene encodes a
358-amino acid protein that is expressed in the CNS and PNS
(Baxter et al., 2002; Cuesta et al., 2002), but its exact cell
location and function is not yet well understood; it probably
intervenes in axon±Schwann cell interaction (Shy et al.,
2002).
Some reports have recently been published describing
families from different ethnic origin presenting several
mutations at the GDAP1 gene. Nelis et al. (2002) described
three families whose members affected displayed median
MNCV >40 m/s except for one case who showed a MCV of
20 m/s; yet their distal CMAP amplitude was unknown and
soon after it could not be obtained, suggesting an advanced
axonal decay. Senderek et al. (2003) reported two patients
from different families showing MCV in the intermediate
range. The descriptions of pathological nerve features in both
Fig. 6 Survey electron micrograph of transverse section through sural nerve biopsy from patients LF20 proband (A) and LF249 proband
(B). In A, groups of Schwann cell processes with a residual partial concentric arrangement indicating their former onion bulb derivation
(black arrowhead). In B, collections of normal-appearing amyelinated ®bres along with a cluster of non-myelinated regenerative axon
(white arrowhead) adjacent to a small myelinated ®bre. There is extensive endoneural collagenization. Bars = 2 mm.
GADP1 neuropathy and vocal cord palsy Page 9 of 11
studies were similar to those found in our cases: loss of large
myelinated ®bres, presence of regenerative cluster and
variable appearance of thin myelinated ®bres and Schwann
cell proliferation with OB formation.
Boerkoel et al. (2003) reported four families. Three of
these were American Hispanic, sharing similar haplotypes
and harbouring the Q163X mutation found in our Spanish
families, suggesting a possible common founder mutation. As
in our patients, MCV could not be obtained in most cases and,
when it could be measured, was in the near normal range. The
histopathological ®ndings were also similar. It is also of
interest to mention that two of their families (one harbouring
the Q163X mutation) showed vocal cord paresis
In conclusion, according to the data available, GDAP1
mutations lead to severe early-onset neuropathy with remark-
able axonal degeneration and a variable degree of demyeli-
Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016
nating features. Vocal cord palsy may sometimes occur in a
length-related fashion. Electrophysiologically this neuro-
pathology may appear with near normal MCV or with mild
slowing in the intermediate range. These features are known
to occur with neuropathies associated with GJB1 mutations
resulting in CMTX (Dubourg et al., 2001; Boerkoel et al.,
2002a). A variable phenotypic presentation also occurs with
MPZ mutations but, in this case, the typical CMT1 phenotype
contrasts with a CMT2 form that appears in relation to
speci®c mutations (De Jonghe et al., 1999; Misu et al., 2000;
Young et al., 2001). At present, there is no reason to think that
a similar allelic distinction may happen with the Q163X
mutation present in all the Spanish and American Hispanic
families who apparently show a greater predisposition to
vocal cord palsy. Otherwise, more extensive clinical, bio-
logical and experimental studies are required to understand
the pathogenesis of these diseases.
Acknowledgements
We wish to thank the patients and their relatives for their
collaboration and EncarnacioÂn Garcia for her help with
sample patients. Dr Joaquin Piquero performed electrophy-
siological studies on a patient from family LF38. This work is
supported by grants from the ComisioÂn Interministerial de
Ciencia y TecnologõÂa (CICYT), Fondo de InvestigacioÂn
Sanitaria (FIS) and Fundacio `la Caixa', Spain.
References
Baxter RV, Ben Othmane K, Rochelle JM,Stajich JE, Hulette C,
Dew-Knight S, et al. Ganglioside-induced differentiation-associated
Page 10 of 11 T. Sevilla et al.
protein-1 is mutant in Charcot±Marie±Tooth disease type 4A/8q21.
Nature Genet 2002; 30: 21±2.
BenOthmane K, Hentati F, Lennon F, Ben Hamida C, Blel S, Roses
AD, et al. Linkage of a locus (CMT4A) for autosomal recessive
Charcot±Marie±Tooth disease to chromosome 8q. Hum Mol Genet
1993; 2: 1625±8.
Boerkoel CF, Takashima H, Lupski JR. The genetic convergence of
Charcot±Marie±Tooth disease types 1 and 2 and the role of genetics
in sporadic neuropathy. Curr Neurol Neurosci Rep 2002a; 2: 70±7.
Boerkoel CF, Takashima H, Garcia CA, Olney RK, Johnson J,
Berry K, et al. Charcot±Marie±Tooth disease and related
neuropathies: mutation distribution and genotype-phenothype
correlation. Ann Neurol 2002b; 51: 190±201.
Boerkoel CF, Takashima H, Nakagawa M, Izumo S, Armstrong D,
Butler I, et al. CMT4A: identi®cation of a Hispanic GDAP1 founder
mutation. Ann Neurol 2003; 53: 400±5.
Bolino A, Brancolini V, Bono F, Bruni A, Gambardella A, Romeo
G, et al. Localization of a gene responsible for autosomal recessive
demyelinating neuropathy with focally folded myelin sheaths to
chromosome 11q23 by homozygosity mapping and haplotype
sharing. Hum Mol Genet 1996; 5: 1051±4.
Bolton CF. AAEM Minimonograph: 40. Clinical neurophysiology
of the respiratory system. Muscle Nerve 1993; 16: 809±18.
Bort S, Sevilla T, GarcõÂa-Planells J, Blesa D, Paricio N, VõÂlchez JJ,
et al. Dejerine-Sottas neuropathy associated with de novo S79P
mutation of the peripheral myelin protein 22 (PMP22) gene. Hum
Mutat 1998; Suppl. 1: S95±8.
Cuesta A, Pedrola L, Sevilla T, GarcõÂa-Planells J, Chumillas MJ,
Mayordomo F, et al. The gene encoding ganglioside-induced
differentiation-associated protein 1 is mutated in axonal Charcot±
Marie±Tooth type 4A disease. Nature Genet 2002; 30: 22±5.
De Jonghe P, Timmerman V, Ceuterick C, Nelis E, De Vriendt E,
Lofgren A, et al. The Thr124Met mutation in the peripheral myelin
protein zero (MPZ) gene is associated with a clinically distinct
Charcot±Marie±Tooth phenotype. Brain 1999; 122: 281±90.
Delague V, Bareil C, Tuffery S, Bouvagnet P, Chouery E, Koussa S,
et al. Mapping of a new locus for autosomal recessive
demyelinating Charcot±Marie±Tooth disease to 19q13.1±13.3 in a
large consanguineous Lebanese family: exclusion of a MAG as a
candidate gene. Am J Hum Genet 2000; 67: 236±43.
Dubourg O, Tardieu S, Birouk N, Gouider R, LeÂger JM, Maisonobe
T, et al. Clinical, electrophysiological and molecular genetic
characteristics of 93 patients with X-linked Charcot±Marie±Tooth
disease. Brain 2001; 124: 1958±67
Dyck PJ, Lambert EH. Lower motor and primary sensory neuron
diseases with peroneal muscular atrophy. I. Neurologic, genetic, and
electrophysiologic ®ndings in hereditary polyneuropathies. Arch
Neurol 1968a; 18: 603±18.
Dyck PJ, Lambert EH. Lower motor and primary sensory neuron
diseases with peroneal muscular atrophy. II. Neurologic, genetic,
and electrophysiologic ®ndings in various neuronal degenerations.
Arch Neurol 1968b; 18: 619±25.
Dyck PJ, Chance P, Lebo R, Carney JA. Hereditary motor and
sensory neuropathies. In: Dyck PJ, Thomas PK, Grif®n JW, Low
PA, Poduslo JF, editors. Peripheral neuropathy. 3rd ed.
Philadelphia: W. B. Saunders; 1993. p. 1094±36.
Dyck PJ, Litchy WJ, Minnerath S, Bird TD, Chance PF, Schaid DJ,
et al. Hereditary motor and sensory neuropathy with diaphragm and
vocal cord paresis. Ann Neurol 1994; 35: 608±15.
GabreeÈls-Festen AA, Joosten EM, GabreeÈls FJ, Stegeman DF, Vos
AJM, Busch HF. Congenital demyelinating motor and sensory
neuropathy with focally folded myelin sheaths. Brain 1990; 113:
1629±43.
GabreeÈls-Festen AA, Joosten EM, GabreeÈls FJ, Jennekens FG,
Gooskens RH, Stegeman DF. Hereditary motor and sensory
neuropathy of neuronal type with onset in early chidhood. Brain
1991; 114: 1855±70.
GabreeÈls-Festen AA, Joosten EM, GabreeeÈls FJ, Jennekens FG,
Janssen-van Kempen TW. Early morphological features in
Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016
dominantly inherited demyelinating motor and sensory neuropathy
(HMSN type I). J Neurol Sci 1992; 107: 145±54.
Hahn AF, Brown WF, Koopman WJ, Feasby TE. X-linked
dominant hereditary motor and sensory neuropathy. Brain 1990;
113: 1511±25.
Harding AE, Thomas PK. The clinical features of hereditary motor
and sensory neuropathy types I and II. Brain 1980a; 103: 259±80.
Harding AE, Thomas PK. Autosomal recessive forms of hereditary
motor and sensory neuropathy (types I and II). J Neurol Neurosurg
Psychiatry 1980b; 43: 669±78.
Hayasaka K, Himoro M, Sawaishi Y, Nanao K, Takahashi T,
Takada G, et al. De novo mutation of the myelin P0 gene in
Dejerine Sottas disease (hereditary motor and sensory neuropathy
type III). Nature Genet 1993; 5: 266±8.
Hentati F, Ben Othmane K, Barhoumi C, Amouri R, Ben Hamida
Ch, Ben Hamida M. Genotype phenotype relation analysis in
autosomal recessive Charcot±Marie±Tooth disease in Tunisia. Acta
Myol 2001; 20: 25±8.
Ionasescu VV, Ionasescu R, Searby Ch, Neahring R. Dejerine-
Sottas disease with de novo dominant point mutation of the PMP22
gene. Neurology 1995; 45: 1766±7.
Kaku DA, Parry GJ, Malamut R, Lupski JR, GarcõÂa CA. Uniform
slowing of conduction velocities in Charcot±Marie±Tooth
polyneuropathy type 1. Neurology 1993; 43: 2664±7.
Kalaydjieva L, Hallmayer J, Chandler D, Savov A, Nikolova A,
Angelicheva D, et al. Gene mapping in Gypsies identi®es a novel
demyelinating neuropathy on chromosome 8q24. Nature Genet
1996; 14: 214±7.
Krajewski KM, Lewis RA, Fuerst DR, Turansky C, Hinderer SR,
Garbern J, et al. Neurological dysfunction and axonal degeneration
in Charcot±Marie±Tooth disease type 1A. Brain 2000; 123: 1516±
27.
LeGuern E, Guilbot A, Kessal M, Ravise N, Tassin J, Maisonobe T,
et al. Homozygosity mapping of an autosomal recessive form of
demyelinating Charcot±Marie±Tooth disease to chromosome 5q23-
q33. Hum Mol Genet 1996; 5: 1685±1688.
Misu K, Yoshihara T, Shikama Y, Awaki E, Yamamoto M, Hattori
N, et al. An axonal form of Charcot±Marie±Tooth disease showing

distinctive features in association with mutations in the peripheral
myelin protein zero gene (Thr124Met or Asp75Val). J Neurol
Neurosurg Psychiatry 2000; 69: 806±11.
Nelis E, Erdem S, Van den Bergh PYK, Belpaire-Dethiou M-C,
Ceuterick C, Van Gerwen V, et al. Mutations in GDAP1: autosomal
recessive CMT with demyelination and axonopapthy. Neurology
2002; 59: 1865±72.
Nicholson G, Nash J. Intermediate nerve conduction velocities
de®ne X-linked Charcot±Marie±Tooth neuropathy families.
Neurology 1993; 43: 2558±64.
Ouvrier RA, McLeod JM, Morgan GJ, Wise GA, Conchin TE.
Hereditary motor and sensory neuropathy of neuronal type with
onset in early childhood. J Neurol Sci 1981; 51: 181±97.
Pareyson D, Taroni F, Botti S, Morbin M, Baratta S, Lauria G, et al.
Cranial nerve involvement in Charcot±Marie±Tooth disease type 1
due to early growth response 2 gene mutation. Neurology 2000; 54:
1696±8.
Reilly MM. Classi®cation of the hereditary motor and sensory
neuropathies. Curr Opin Neurol 2000; 13: 561±4.
Rozear MP, Pericak-Vance MA, Fischbeck K, Stajich JM, Gaskell
PC, Krendel DA, et al. Hereditary motor and sensory neuropathy,
X-linked: a half century follow-up. Neurology 1987; 37: 1460±5.
Sahenk Z, Chen L, Mendell JR. Effects of PMP22 duplication and
deletions on the axonal cytoskeleton. Ann Neurol 1999; 45: 16±24.
Senderek J,Bergmann C, Ramaekers VT, Nelis E, Bernert G,
Makowski A, et al. Mutations in the ganglioside-induced
differentiation-associated protein-1 (GDAP1) gene in intermediate
type autosomal recessive Charcot±Marie±Tooth neuropathy. Brain
2003; 126: 642±9.
GADP1 neuropathy and vocal cord palsy Page 11 of 11
Shy ME, Garbern JY, Kamholz J. Hereditary motor and sensory
neuropathies: a biological perspective. Lancet Neurol 2002; 1: 110±
8.
Thomas PK, Marques W Jr, Davis MB, Sweeney MG, King RHM,
Bradley JL, et al. The phenotypic manifestations of chromosome
17p11.2 duplication. Brain 1997; 120: 465±78.
Thomas PK, Kalaydjieva L, Youl B, Rogers T, Angelicheva D,
King RHM, et al. Hereditary motor and sensory neuropathy-russe:
new autosomal recessive neuropathy in Balkan gypsies. Ann Neurol
2001; 50: 452±7.
Timmerman V, De Jonghe P, Ceuterick C, De Vriendt E, Lofgren
A, Nelis E, et al. Novel missense mutation in the early growth
response 2 gene associated with a Dejerine-Sottas syndrome
phenotype. Neurology 1999; 52: 1827±32.
Yoshioka R, Dyck PJ, Chance PF. Genetic heterogeneity in
Downloaded from http://brain.oxfordjournals.org/ by guest on February 4, 2016
Charcot±Marie±Tooth neuropathy type 2. Neurology 1996; 46:
569±71.
Young ID, Harper PS. Hereditary distal spinal muscular atrophy
with vocal cord paralysis. J Neurol Neurosurg Psychiatry 1980; 43:
413±8.
Young P, Grote K, KuhlenbaÈumer G, Debus O, Kurlemann H,
Halftar H, et al. Mutation analysis in Charcot±Marie±Tooth disease
type 1: point mutations in the MPZ gene and the GJB1 gene cause
comparable phenotypic heterogeneity. J Neurol 2001; 248: 410±5.
Received January 22, 2003. Revised March 29, 2003.
Accepted April 16, 2003