Brain (2002), 125, 2681±2690
Clinical and genetic characterization of families
with triple A (Allgrove) syndrome
Henry Houlden,1 Stephen Smith,2 Mamede de Carvalho,5 Julian Blake,3 Christopher Mathias,4
Nicholas W. Wood1 and Mary M. Reilly1
Departments of 1Clinical Neurology, 2Neuroophthalmology, 3Neurophysiology and 4Autonomic
Research Unit, Institute of Neurology, London, UK and
5Department of Neurology, EMG Laboratory, Hospital de
Santa Maria, Lisboa, Portugal
Summary
Triple A (Allgrove) syndrome is characterized by
achalasia, alacrima, adrenal abnormalities and a progressive neurological syndrome. Affected individuals
have between two and four of these relatively common
clinical problems; hence the diagnosis is often dif®cult
in all but the classical presentation. The inheritance is
autosomal recessive, and most cases of triple A have no
family history. Using genetic linkage analysis in a small
number of families, a locus on chromosome 12q13 was
identi®ed. The triple A gene was identi®ed recently at
this locus and called ALADIN (alacrima, achalasia,
adrenal insuf®ciency neurologic disorder). Mutations in
this gene were reported in families from North Africa
and Europe. The majority of mutations were homozygous. We have identi®ed 20 families with between two
and four of the clinical features associated with the
triple A syndrome. Sequencing of the triple A gene
revealed ®ve families that had a total of nine compound
heterozygous mutations, and one Portuguese family
(previously published) had two homozygous mutations;
Correspondence to: H. Houlden, Department of Clinical
Neurology, Institute of Neurology, Queen Square, London
WC1N 3BG, UK
E-mail: [email protected]
these changes were spread throughout the triple A gene
in exons 1, 2, 7, 8, 10, 11, 12, 13 and 16, and the
poly(A) tract. Those bearing mutations had the classical
triple A syndrome of achalasia, alacrima, adrenal
abnormalities and a progressive neurological syndrome.
We identi®ed a spectrum of associated neurological
abnormalities in these cases, including pupil and cranial
nerve abnormalities, frequent optic atrophy, autonomic
neuropathy and upper and lower motor neurone signs
including distal motor neuropathy and amyotrophy
with severe selective ulnar nerve involvement. In these
families, we have made genotype±phenotype correlations. Mutations in the triple A gene are thus an
important cause of this clinically heterogeneous
syndrome, and sequencing represents an important
diagnostic investigation. Identifying further mutations
and de®ning their phenotype along with functional
protein analysis will help to characterize this neuroendocrine gene.
Keywords: triple A syndrome; Allgrove syndrome; genetic mutation
Abbreviations: ALADIN = alacrima, achalsia, adrenal insuf®ciency neurologic disorder
Introduction
The triple A (Allgrove) syndrome was ®rst described in two
pairs of siblings by Allgrove and colleagues in 1978 [Online
Mendelian Inheritance in Man (OMIM) database accession
number: 231550; Allgrove et al., 1978]. The disorder is
characterized by achalasia, alacrima and adrenocorticotrophic hormone (ACTH)-resistant adrenal failure. A number
of associated features have been described including progressive central, peripheral and autonomic nervous system
abnormalities, palmo-plantar and punctate hyperkeratosis,
short stature, osteoporosis, xerostomia, angular cheilitis,
glossitis and ®ssured tongue, and microcephaly (Allgrove
ã Guarantors of Brain 2002
et al., 1978; Geffner et al., 1983; Stuckey et al., 1987; Dumic
et al., 1991; Moore et al., 1991; Grant et al., 1992, 1993;
Gazarian et al., 1995; Heinrichs et al., 1995; HuÈbschmann,
1995; Chu et al., 1996; Clark and Weber, 1998; Huebner
et al., 1999; Zeharia et al., 1999; Bentes et al., 2001).
Necropsy has been carried out in one of the pairs of original
siblings published (Allgrove et al., 1978). This revealed
atrophy of the adrenal cortical zona fasciculata and zona
reticularis and absence of ganglion cells and nerve ®bres in
the lower oesophagus. No comment was made regarding the
rest of the pathological examination.
2682
H. Houlden et al.
A number of families have been reported in the literature.
All display an autosomal recessive pattern of inheritance and
most have congsanguineous parents (Allgrove et al., 1978;
Geffner et al., 1983; Ehrich et al., 1987; Stuckey et al., 1987;
Dumic et al., 1991; Moore et al., 1991; Grant et al., 1992,
1993; Gazarian et al., 1995; Heinrichs et al., 1995;
HuÈbschmann, 1995; Zeharia et al., 1999; Bentes et al.,
2001). Combining a number of these small families, a locus
on chromosome 12q13 (Weber et al., 1996; Huebner et al.,
2000) was identi®ed. This was replicated in two other
population groups (Stratakis et al., 1997; Hadj-Rabia et al.,
2000; Lee et al., 2000). Using positional cloning, the product
of the triple A gene, designated ALADIN for alacrima,
achalasia, adrenal insuf®ciency neurologic disorder, was
identi®ed recently as a WD-repeat protein (Tullio-Pelet et al.,
2000; Handschug et al., 2001). This functionally diverse type
of protein is coded for frequently in the genome consistent
with triple A syndrome, which is a multi-tissue disorder with
ubiquitous gene expression; notably a particularly high
expression is seen in the adrenal gland, gastrointestinal tract
and brain. Mutations in this gene have also been identi®ed in
consanguineous North African and European families
(Tullio-Pelet et al., 2000; Handschug et al., 2001). Two
families with compound heterozygous mutations were also
identi®ed in kindreds where the parents were unrelated
(Handschug et al., 2001). An additional study of triple A
families from Puerto Rico identi®ed further families with the
homozygous IVS14 + 1G®A mutation and one compound
heterozygous mutation, and a Canadian family homozygous
for an exon 1 point mutation (Sandrini et al., 2001).
There is signi®cant heterogeneity in the clinical features
and the types of mutation reported in families with suspected
triple A syndrome. No study has looked at both the clinical
manifestations and the genetic characterization in a large
group of families. We therefore sequenced the triple A gene
in a group of 20 families with progressive neurological
problems and at least one of the triad of achalasia, alacrima
and adrenal abnormalities. To look for and to characterize
fully genotype±phenotype correlations in our families where
mutations were identi®ed, we re-examined affected and
unaffected individuals from these kindreds in detail carrying
out repeat neurophysiology, autonomics and pupillography.
Methods
Patients
Ethics approval was obtained from the joint medical and
ethics committee at The National Hospital for Neurology and
Neurosurgery to carry out this genetic study. We have
collected blood samples from 20 families where the proband
had progressive neurological problems and at least one of the
triad of achalasia, alacrima and adrenal abnormalities.
Members of eight out of these 20 families had all the
characteristics of triple A syndrome (see Tables 2±5 for
details). A blood sample and clinical details from one of these
eight families (family 6) were obtained from Portugal (Bentes
et al., 2001). Individuals in the other 12 families had
progressive neurological problems, consisting of predominantly motor neuropathy; 10 out of the 12 also had optic
atrophy while two of 12 had retinopathy. Members of six of
the 12 families had mild spasticity in the upper and lower
limbs. Ten out of 12 of the families had gut abnormalities,
mainly achalasia, and the other two families had alacrima.
Nineteen families were patients of The National Hospital for
Neurology, and one family (family 6) was from the Hospital
de Santa Maria in Lisbon, Portugal (Bentes et al., 2001). All
had had detailed clinical examination and investigations,
including routine and specialized blood tests, thyroid function
tests, CSF for cells, protein glucose and bands. MRI brain and
genetic testing had been carried out to exclude the common
mitochondrial mutations. Clinical neurophysiology and
autonomic function tests were carried out in all families. In
families with triple A mutations, further examination (by
H.H. in families 1±5) of the proband and relatives was carried
out. Investigations including repeat neurophysiology and
pupillography were also done. All families were of European
origin, the majority English; there was no history of
consanguinity (Fig. 1).
Genetic sequencing
DNA was extracted from blood samples obtained from
affected and unaffected individuals from families. We
sequenced the entire coding region and ¯anking introns of
the triple A gene in an affected individual from each family.
The 16 exons and ¯anking intronic regions (see Table 1 for
primer sequences) of the triple A gene were ampli®ed by
PCR. The intronic regions between exons 4 and 5, 10 and 11,
12 and 13, and 14 and 15 were small enough to allow these
exons to be ampli®ed together. PCR was carried out in a total
volume of 50 ml, which contained 20 ng of DNA, 0.2 mM
dNTPs, 1 U of TaqGold polymerase, 1.5 mM MgCl2, 75 mM
Tris±HCl pH 9.0, 20 mM (NH4)2SO4, 0.01% Tween-20 and
50 pmol of each primer. PCR was performed on a PerkinElmer 9700 thermal cycler (Perkin Elmer, Applied
Biosystems, Foster City, California, USA). The cycling
consisted of denaturation at 94°C for 15 min, followed by 25
cycles of 94°C for 30 s, 60°C to 50°C touching down protocol
for 30 s, and 72°C for 30 s. After that, 12 cycles of constant
annealing temperature at 50°C and a ®nal ampli®cation at
72°C for 10 min was carried out. PCR fragments were
checked on a 1% agarose gel. The PCR products were
puri®ed by using a Qiaquick puri®cation kit (Qiagen, Hilden,
Germany) and resuspended in 50 ml of deionized water. For
each exon, 100 ng of ampli®ed product was sequenced using
forward and reverse primers and the BigDye Terminator
cycle sequencing kit (Perkin Elmer). Sequencing was
performed on an ABI377 automated sequencer; alignment
and analysis were carried out with Sequence Navigator
(Perkin Elmer).
Genetics of triple A syndrome
2683
Fig. 1 Family trees of all kindreds with mutations in the triple A gene. Circle = female; square = male; ®lled = clinically affected. Arrow
indicates the proband. +/+ = two mutations; +/± = normal carrier of a mutation; ±/± = no mutation; del = deletion; ins = insertion; n/a = not
tested genetically. The mutation details are given beneath the affected individual. In the text and subsequent tables when referring to the
family number this means the affected individual in the family. In family 3, 3.1 is the proband and 3.2 the other affected sibling, family 5,
5.1 is the proband and 5.2 the other affected sibling. Family members that are heterozygous are indicated with the exonic location of the
mutation. All mutations segregate with the disease in an autosomal recessive pattern. AMB = stop codon TAG; OPA = stop codon TGA.
Table 1 Triple A gene primer sequences
Exon
Forward (5¢±3¢)
Reverse (5¢±3¢)
1
2
3
4 and 5
6
7
8
9
10 and 11
12 and 13
14 and 15
16
GGAGTTTGCCGACTGCAGAC
GCATTTGAGTTCTATAATAAGGAC
CACTCTGGACACCCACTC
AGTAGGAGTCTTTGCCTTCTC
TCAGGTTCAAGAACTACAGGAC
CTCCAGATTAGAGTATTCTCAGC
AGACCTTGCAGATTACCTTC
ATTAGAGAGGCCAGCCCACTG
AAAGGCACTTAGCTCCTGGAAG
TTAGGAGATTTCGAGGTGTTGATG
GAGTTCTCCTCTGCCCATGTC
AGTTGGATGGAGAAGCTGAGG
CCTGTCACACTGCCTCCTTTC
CTCTGGAATCTCTTATACTTAGC
CATAGTTGGCACTCATTAATTG
AGAGTGTGGTGTTGAGAGCAC
TTACTGGAAATGAATGTGAGC
TCCTTAACTGCACTCTGGTC
TCTGGGTAAGTTTAAGACTG
AAGTTGGACCTACCTCCCTTGAC
TCTATATTTCCCTTTATCCCTCAGAG
GGCACGGCCTCATTAGATTAAC
AGAGCCATACAGCAGCCAAG
GCCTTAACCCAAAGTCCATG
Mutation analysis
DNA from parents was available in families 2, 4 and 5 and
also from grandparents in family 4. Mutation analysis was
carried out and the inheritance pattern was consistent with
autosomal recessive with mutations segregating with disease;
parents and one grandparent were clinically normal heterozygous carriers in these three compound heterozygous
families (Fig. 1). Siblings were examined in families 1
(normal sister), 2 (normal brother and sister) and 4 (normal
brother and sister). The affected individual in family 6 had
normal parents. Mutations segregated with the disease in all
families and were not present in 75 controls. One silent
polymorphism was identi®ed in the triple A gene; this change
was present in controls.
Autonomic function tests and neurophysiology
Detailed autonomic screening function tests, with a particular
view to determining sympathetic vasoconstrictor and cardiac
parasympathetic function, were performed. These included
the responses to head-up tilt, to a series of stimuli that should
raise blood pressure, to a variety of manoeuvres which would
2684
H. Houlden et al.
Table 2 Triple A families mutation details
Family
Affected
Consanguinity
Ethnic origin
Mutation type
Mutation (allele 1)
Mutation (allele 2)
1
1
No
English
Exon 1, Gln 15 Lys
2
1
No
English
Exon 8, Ser 263 Pro
Exon 11, bases 1148 and
1149 Del CT
Exon 13, base 1273 Ins A
3
2
No
English/German
Exon 1, Gln 15 Lys
Exon 12, base 1186 Ins C
4
1
No
English/Scottish
2
No
Welsh/English
Exon 12, bases 1226±1229
Del TCTG
Exon 2, base 292 Del C
Exon 16, Trp 474 AMB
5
6
1
No
Portuguese
Compound
heterozygote
Compound
heterozygote
Compound
heterozygote
Compound
heterozygote
Compound
heterozygote
Homozygote
Exon 7, Arg 230 OPA
Exon 10, Val 313 Ala
Ins = insertion; Del = deletion; AMB = stop codon (TAG); OPA = stop codon (TGA).
Table 3 Triple A families non-neurological clinical details
Patient
Age of onset
(years)
Age at examination
(years)
Presenting clinical feature
Achalasia
Adrenal dysfunction
Alacrima
1
2
3.1
3.2
4
5.1
5.2
6
16
12
6
3
8
4
3
25
48
37
36
39
18
32
30
35
Lower limb weakness
Lower limb weakness
Alacrima
Alacrima
Hypoglycaemic seizures
Hypoglycaemic seizures
Alacrima
Achalasia + lower limb weakness
Mild
Mild
Mild
Severe
Moderate
Severe
Moderate
Mild
Mild
Borderline abnormal
Mild
Mild
Moderate
Moderate
Moderate
Borderline abnormal
Mild
Mild
Severe
Moderate
Moderate
Mild
Mild
Moderate
Achalasia criteria: mild = symptoms but no surgery; moderate = symptoms requiring one or two oesphageal dilatation procedures;
severe = >3 oesphageal dilatation procedures. Adrenal criteria: borderline abnormal = normal cortisol, raised adrenocorticotrophic
hormone, no replacement; mild = required corticosteroid replacement in the past or continued low dose; moderate = continuous need for
corticosteroid replacement, required mineralocorticoid in the past. Alacrima criteria: mild = abnormal Schirmer test and symptoms, eye
drops as required; moderate = abnormal Schirmer test and symptoms with daily regular treatment; severe = abnormal Schirmer test, severe
symptoms, frequent daily treatment, eye/ski goggles.
change heart rate, and to standing upright. The protocols
followed were as described previously (Mathias and
Bannister, 1999)
Clinical neurophysiology
Nerve conduction studies detailed in Table 5 were performed
by standard methods (De Lisa et al., 1994) using a Nicolet
Viking II EMG machine. Supramaximal nerve stimuli were
used throughout, with a stimulus duration of 0.1 ms.
Sensory conduction in the upper limbs was recorded
orthodromically with ring stimulating electrodes on the
®ngers and recording the evoked responses with surface
electrodes over the median or ulnar nerves at the wrist. The
antidromic recording technique was used in the lower limbs,
stimulating the super®cial peroneal and sural nerves above
the ankle and recording over the nerves at the ankle.
Motor conduction was studied by recording compound
muscle action potentials with surface electrodes over
abductor policis brevis for the median nerve and adductor
digiti minimi for the ulnar nerve, and stimulating at both the
wrist and the elbow in each case. Peroneal motor conduction
was measured recording from extensor digitorum brevis and
stimulating the peroneal nerve at the ankle and at the ®bular
head.
Pupillography
This was carried out in the pupil laboratory (by S.S.) on
families 1±5. Pupil diameters and their responses to light and
accommodation (near) were recorded with a Whittaker/
Applied Science Laboratories infrared television pupillometer as previously described (Smith et al., 1978). Pupil
diameters were recorded in darkness, and near responses
under dim ambient illumination. Light re¯exes were induced
with xenon arc or light-emitting diode white illumination, 1 s
duration at 10 s intervals, at an intensity suf®cient to produce
the largest possible re¯ex for each subject. Darkness diameters, the magnitude of light re¯ex responses and the rate of
redilatation were compared with those obtained in healthy
Genetics of triple A syndrome
2685
Table 4 Triple A families neurological manifestations
Patient
Cranial nerve
examination
ANS
Distal weakness and wasting
Re¯exes
Sensory
examination
Pupil
diameter
Pupil
reactivity
1
Mild
Loss of
vibration
sense
Loss of
vibration
sense
Normal
Normal
Pupillotonia,
LND
Normal
Redilatation
lag
Normal
Pupillotonia,
LND
Brisk, absent
AJ, ExP
Normal
Anisocoria
Pupillotonia,
LND
Mild
Brisk, absent
AJ, FP
Normal
Normal
Normal
5.1
Mild bilateral OA
Moderate
Slightly brisk,
absent AJ, FP
Normal
Miosis
Pupillotonia,
no LND
5.2
Mild bilateral OA
Mild
Slightly brisk,
absent AJ, FP
Normal
Normal
Pupillotonia,
no LND
6
Unilateral IX and X,
bilateral XI and
XII, dysarthria
Moderate
Severe weakness, mild distal
upper and lower limb
amyotrophy
Mild weakness, hypothenar
prominent wasting, lower
limb amyotrophy
Mild weakness, hypothenar
prominent wasting, lower
limb amyotrophy
Mild weakness, hypothenar
prominent wasting, lower
limb amyotrophy
Mild weakness, hypothenar
prominent wasting, lower
limb amyotrophy
Moderate weakness, mild
distal upper and lower limb
amyotrophy
Mild weakness, hypothenar
prominent wasting, lower
limb amyotrophy
Moderate weakness,
symmetrical
Brisk, absent
AJ, ExP
4
Bilateral OA,
bilateral XII,
dysarthria
Bilateral OA,
R Horner's,
RP
Bilateral OA,
bilateral XII,
dysarthria
Bilateral OA,
bilateral XII,
dysarthria
Bilateral OA
Brisk, absent
AJ, FP
Normal
Normal
Normal
2
3.1
3.2
Mild
Mild
Mild
Brisk, absent
AJ, FP
Brisk, absent
AJ, EqP
OA = optic atrophy; RP = retinopathy; ANS = autonomic nervous system abnormalities; AJ = ankle jerks; ExP = extensor plantar re¯ex;
EqP = equivocal plantar re¯ex; FP = ¯exor plantar re¯ex; LND = light-near dissociation; IX, X, XI and XII = 9th, 10th, 11th and 12th
cranial nerves, respectively. Families 1±5 had formal pupillography carried out, family 6 was clinically examined.
subjects (Smith and Smith, 1999). Pupillotonia was sought by
inspection of the rate of constriction in response to continuous
bright illumination.
Nerve biopsy
Nerve biopsy was only carried out in family 6, from Portugal.
Sural nerve fascicular biopsies were obtained from a standard
retromalleolar site. Following routine aldehyde ®xation, postosmication, dehydration through an ascending alcohol series
and resin embedding, semi-thin sections were stained with
thionin and acridine orange (Sievers, 1971).
Results
Out of the 20 families analysed, ®ve were found to have
compound heterozygous mutations and one two homozygous
changes in the triple A gene. A total of 11 different mutations
were found in these kindreds, of which nine were novel
(Table 2). The mutations were spread throughout the triple A
gene in exons 1, 2, 7, 8, 10, 11, 12, 13 and 16. The majority of
mutations were insertions or deletions and created a frameshift change; two were nonsense and two were missense. A
sixth family from Lisbon in Portugal (Bentes et al., 2001) had
two previously unidenti®ed homozygous mutations, one that
caused the amino acid change valine to alanine at codon 313
and another that causes a deletion of ATAA in the poly(A)
tract (bp 1730±1733). The exonic mutation is probably the
pathogenic mutation in this family as the deletion is after the
stop codon of the triple A gene, although mutations that
disrupt the poly(A) tract can be pathogenic by affecting
mRNA stability or splicing (Yuregir et al., 1992; Beaudoing
et al., 2000). The family trees and segregation are shown in
Fig. 1. Family mutation details are given in Table 2. A further
sequence change was identi®ed in exon 5 at codon 138
(GAT®GAC); this was silent and did not change an amino
acid, and was also present in controls.
All individuals with mutations had progressive neurological abnormalities, achalasia, alacrima and adrenal
dysfunction with varying degrees of severity. The affected
individual of families 1 and 2 had less severe manifestations.
The non-neurological clinical features of affected individuals
from families with triple A mutations are presented in Table 3.
The clinical details of the Portuguese family (family 6) have
already been published (Bentes et al., 2001). However, in
brief, the affected individual had late-onset disease starting at
25 years of age with achalasia, followed 5 years later by
slowly progressive lower limb weakness. On examination,
upper and motor neurone signs were detected in the upper and
lower limbs; the tongue was atrophic and with fasciculations.
2686
Patient
Median nerve
Ulnar nerve
SAP
(mV)
SNCV
(ms)
CMAP
(mV)
DML
(ms)
MNCV
(ms)
F wave
(m/s)
SAP
(mV)
SNCV
(ms)
CMAP wrist
(mV)
DML
(ms)
MNCV
(ms)
1
2
3.1
3.2
4
6*
Absent
9
6
22.7
16
21.0
40
55.5
50
n/a
58.5
42.5
2.5
9.4
6.5
5.3
5.3
3.7
5
4
3.4
4.5
4.4
3.6
38
54
47
52
52
45.8
41.3
26±37
32.3
29.6
29.9
30
Absent
6.5
4
12.2
7.5
22.4
Absent
60
47.5
n/a
61.5
39.3
Absent
0.7
0.5
4.9
1.8
2.1
Absent
3.9
3.9
5.2
4.6
3.7
Absent
39
36
n/a
n/a
41.3
Patient
Sural nerve
EMG
CMCT
Common peroneal
nerve
SAP (mV)
SNCV (ms)
CMAP (mV)
DML (ms)
F wave (ms)
Tibialis anterior
1
4
47
0.013
4.5
n/a
2 (post tibial)
7
37
1.2
5.6
Absent
3.1
2.5
37
3.5
3.9
Absent
3.2 (post tibial)
4
6.2
17
n/a
38
0.4
1.3
6.5
4.7
54.8
52.9
5.1 (age 10 years)
26
56
5.1 (age 17 years)
15
5.1 (age 18 years)
7
5.2 (age 13 years)
3
5.2 (age 16 years)
3
6*
5.0
4.0 abductor
hallucis
1.3 abductor
hallucis
1.3 abductor
hallucis
0.3 abductor
hallucis
0.1 abductor
hallucis
0.3
Chronic partial
denervation
Chronic partial
denervation
Chronic partial
denervation
n/a
Chronic partial
denervation
3.9
Absent
41.3
Chronic partial
denervation
Prolonged
Prolonged
Normal
Normal
n/a
Prolonged
SAP = sensory nerve action potential; SNCV = sensory nerve conduction velocity; CMAP = compound muscle action potential amplitude; DML = distal motor latency;
MNCV = motor nerve conduction velocity; CMCT = central motor conduction time, n/a = not available; *this patient was investigated in another laboratory and sural nerve sensory
potential was studied instead of the peroneal nerve sensory potential.
H. Houlden et al.
Table 5 Triple A families neurophysiology details
Genetics of triple A syndrome
Symptoms of autonomic dysfunction (sexual impotence,
orthostatic hypotension and alacrima) appeared 11 years after
the ®rst symptoms. Adrenal impairment was recognized by
laboratory testing. The mutation type in this family is
different from that of the other families (Table 2). Families
3±5 presented in childhood; families 4 and 5 with
hypoglycaemic seizures after diarrhoea in childhood and
family 3 with alacrima. The affected individuals in families 1
and 2 presented in their teens with lower limb weakness.
Details of the neurological examination are presented in
Table 4. Manifestations included bilateral optic atrophy in all
families apart from family 6; the affected individuals from
family 2 also had right Horner's syndrome as well as bilateral
retinopathy, and those from families 1, 3 and 6 had an
atrophic tongue and spastic dysarthria. The affected individual in family 6 had a left palatal paresis, bilateral trapezius
and sternocleidomastoid weakness. The two individuals from
family 5 had mild mental retardation and parkinsonism;
between 10 year periods of examination this had not
deteriorated (Grant et al., 1992). All but one of the families
(family 4) had abnormal pupils, in each instance the
abnormality being bilateral and approximately symmetrical.
There was pes cavus and predominantly distal weakness and
wasting in the upper and lower limbs in all families. In the
upper limbs, this was patchy with marked weakness in an
ulnar distribution and wasting of the hypothenar eminence.
The tone was generally increased in the upper and lower
limbs, re¯exes were brisk in the upper limbs but the ankle
jerks were lost; plantar re¯exes were extensor in one family
and equivocal in another. Sensory examination showed loss
of vibration sense only in families 1 and 2.
Neurophysiology is presented in Table 5 from the families
where testing was carried out at The National Hospital for
Neurology and from family 6 which was analysed in Portugal.
Nerve conduction studies showed a predominant motor
neuropathy in all families, predominantly axonal with mild
sensory abnormalities. EMG showed chronic partial denervation. In all families analysed, there was selective involvement
of ulnar nerves in the upper limbs, with the compound muscle
action potential being severely affected compared with the
median. Nerve conduction studies were carried out in family
5 in detail at age 10 years as well as the data given in Table 5.
The sural nerve sensory action potentials were normal at age
10 years in case 5.01, but by the age of 17 years these had
deteriorated and were reduced further at age 18 years. Repeat
neurophysiology was carried out over between 3 and 6 years
in families 1, 2 and 3; these were abnormal but there was little
deterioration over this time. Conduction block and ulnar
entrapment at the elbow was excluded. Central motor
conduction time was prolonged in families 1, 2 and 6 but
normal in family 3.
Somatosensory and visual potential examinations were
also carried out in family 6 and were normal. MRI of the head
and spinal cord was carried out in families 1±4 and 6 and was
normal. Sural nerve biopsy was carried out in family 6, and
this was normal under light microscopy.
2687
In families 1±5, cardiovascular autonomic testing indicated
that orthostatic hypotension, as de®ned by present criteria
(20 mmHg systolic or 10 mmHg diastolic blood pressure fall),
was present in families 5 and 3 (patients 5.01 and 3.02). The
two were studied again at intervals of 2 and 3 years,
respectively, and over this period progression was seen in
relation to the degree of orthostatic hypotension, impairment
of responses to pressor stimuli and parasympathetic cardiac
impairment. One subject was studied at age 16 and 18 years
(patient 5.01), and the other at age 33 and 39 years (patient
3.02). Patient 5.01 was on replacement hydrocortisone, and
although this might have in¯uenced the responses to some
extent, it would not explain the abnormal tests of autonomic
neural function. In other subjects, there was no clear evidence
of autonomic dysfunction, but there were small falls in
systolic blood pressure during head-up tilt in two families,
suggesting partial defects. Family 6 had cardiovascular
autonomic and sudomotor testing carried out in Portugal,
and these were abnormal.
Two families with the full triple A syndrome phenotype did
not have mutations in the gene. Only one individual was
affected in each non-consanguineous family, but both had
features of achalasia, alacrima and borderline normal adrenal
problems with progressive severe neurological abnormalities.
Both presented in their teenage years with lower limb
weakness. The ®rst had peripheral motor and sensory
neuropathy with marked amyotrophy and cerebellar ataxia;
the second had bilateral optic atrophy with distal motor
neuropathy. Parents were not available for examination but
were understood to have been normal.
Discussion
We collected DNA from 20 families with either a phenotype
consistent with the clinical triad of triple A syndrome or a
progressive neurological syndrome with one of the three
features of the triple A syndrome. Eight of these 20 families
had all the characteristic features of this disorder. Sequencing
the triple A gene in our families revealed ®ve families with
compound heterozygous mutations and one family with
homozygous mutations present (Table 2). Mutations in the
triple A gene were only present in six families with a classical
phenotype of achalasia, alacrima, adrenal abnormalities and a
progressive neurological syndrome. These mutations are
pathogenic; they segregate with the disease and they are not
present in control individuals. Seven of the mutations are
structurally severe frameshift or nonsense mutants and the
other two cause a substitution of a conserved amino acid. In
all the families with compound heterozygous mutations, one
mutation was a frameshift and the other was a missense or a
nonsense mutation; these previously unreported changes are
predicted to cause a truncated, non-functional protein. Two
missense mutations were also identi®ed; these two mutations
have been reported previously (Handschug et al., 2001). The
exon 8 (Ser263®Pro) mutation is likely to disrupt the bpropeller structure of the gene; the exon 1 (Gln15®Lys)
2688
H. Houlden et al.
mutant was identi®ed in two families and is conserved in
species, segregates with the disease and may be in an area
involved in ALADIN protein b-strand stability.
Achalasia, adrenal dysfunction and alacrima were present
in all six families (Table 3). Families 2 and 6 had only
borderline adrenal function abnormality. There was no
signi®cant correlation between each of these manifestations.
Families 1 and 2 had less severe manifestations, which is also
true for their neurological features.
In our series of triple A families, we have carried out
detailed neurological examination (Table 4). Cranial nerve
manifestations were present in all families with varying
degrees of severity. Pupil abnormalities are a common
component of the triple A syndrome. At what stage these
abnormalities occur is not known, but it may be pertinent that
our patient with normal pupils (proband of family 4) was also
the youngest by some years (17 versus 32 years or more). It is
also of note that case 5.01 had normal peroneal sensory nerve
action potential and compound muscle action potential at age
10 years, but this had deteriorated by age 17 years. Thus, this
suggests that symptoms progress and develop in teenage and
early adult years. The most frequent ®nding, pupillotonia,
occurred in ®ve cases. Absent, or reduced and very slow, light
re¯exes have already been reported in this condition (Stuckey
et al., 1987; Hammami et al., 1989; Gazarian et al., 1995),
observations which are compatible with our ®ndings.
Pupillotonia is considered to be a consequence of long-term
parasympathetic nerve damage (Loewenfeld and Thompson,
1967). More widespread parasympathetic abnormalities
involving the cardiovascular system have been reported in
an earlier follow-up study of both patients in family 5 (Grant
et al., 1992), and in two other cases (Chu et al., 1996).
Involvement of the sympathetic system is less certain. In the
pupil, detection of a sympathetic de®cit is problematic if there
is coincident parasympathetic damage, because the physiological signs are masked. Furthermore, we consider that
topical drug studies may be unreliable in the presence of
severe alacrima, because drug penetration into the eye is
likely to be greatly enhanced. We found bilateral pupillary
redilatation lag without pupillotonia in one patient (proband
of family 2), from which we deduce the presence of a
sympathetic de®cit. Abnormal cardiovascular sympathetic
function tests as well as peripheral nerve autonomic abnormalities with cold and dry skin were also seen in our patients
(Grant et al., 1992), and postural hypotension in two other
cases (Chu et al., 1996). It therefore appears that autonomic
neuropathy in this condition is not con®ned to the lacrymal
glands and the upper part of the gastrointestinal tract.
Repeat autonomic function measurements previously had
not been made, unlike in two of our patients, who were
studied at intervals of 2 and 3 years, respectively. In both
patients (5.01 and 3.02), repeat studies indicated a greater
degree of orthostatic hypotension, and further blunting of
sympathetic vasoconstrictor responses, indicating progression. This would be consistent with the pattern observed in
relation to other clinical features, which progress. The case
for longitudinal study in these subjects therefore is strong.
The lesion site resulting in sympathetic dysfunction is not
known, as there are limited neuropathological studies. There
may be central or peripheral involvement; if the latter, there
may be involvement of autonomic ganglia and the nicotinic
acetylcholine receptor, thus affecting both sympathetic and
parasympathetic function. It could be said that the abnormal
sudomotor tests observed in family 6 indicate a postganglionic sympathetic dysfunction. The lesion may have
been to post-ganglionic pathways at either pre- or postsynaptic sites. Further studies, using appropriate physiological and pharmacological tests, should determine the site
or sites involved.
Upper and lower motor neurone abnormalities were
present to a variable degree of severity in all families. This
was also the case for motor neuropathy which was present in
the upper and lower limbs on clinical examination and nerve
conduction studies. In the upper limbs, there was marked
patchiness with selective involvement of the ulnar nerve. In
the Portuguese case where the triple A mutation is reported
here, there was similar motor neuropathy, most evident on
nerve conduction studies (Table 5). In this case, sural nerve
biopsy was carried out and nerve pathology on light
microscopy was normal. The marked selective involvement
of the ulnar nerve is interesting and not seen in hereditary
motor and sensory neuropathies. The cause of this patchiness
is unknown. As the disease progresses, other nerve involvement becomes apparent, as seen in family 1 where there is
both severe ulnar nerve and moderate/severe median nerve
motor neuropathy. Nerve conduction studies in our families
suggest that the peripheral nerve pathology involves axonal
loss. The increased tone, re¯exes and abnormal central motor
conduction time suggests that there is also involvement of the
corticospinal tracts, although MRI of the head and spinal cord
was normal. Therefore, sequencing of the triple A gene
should be considered in those patients with a complicated
axonal neuropathy.
The clinical features of the families with triple A mutations
are diverse, involving many systems. This re¯ects the
function of the ALADIN gene which belongs to the WD
repeat protein family which exhibits wide functional diversity
that includes protein±protein interactions, signal transduction, RNA processing, vesicular traf®cking, cytoskeleton
assembly and cell division control (Neer et al., 1994; Smith
et al., 1999). The families show considerable variability in
disease severity that could be due to different mutation types
and position. Looking for genotype±phenotype correlations is
dif®cult. Although we have the largest reported series of
families with compound heterozygous mutations, no two
families have the same two mutations. However, we can look
at two broad groups of families. Families 1 and 2 have a later
age of onset and less severe clinical phenotype than the other
families analysed. They are both compound heterozygotes
with a frameshift mutation on one allele and an exon 1
missense mutation on the other. Families 3±5 have a more
severe phenotype with a frameshift mutation on one allele and
Genetics of triple A syndrome
a mutation creating a stop codon change or a mutation in an
important functional position on the gene (Ser263®Thr). The
mutations in families 3±5 are more likely to lead to a greater
loss of function in the triple A gene and hence greater
phenotype severity, whereas in families 1 and 2 there is likely
to be only partial loss of function. The issue of genotype±
phenotype correlations along with the structure and function
of the triple A gene need to be analysed further with screening
for mutations in families and subsequent protein analysis and
disease modelling.
Five families had a total of nine compound heterozygous
mutations, and one family (family 6) had two homozygous
mutations, although the deletion in the poly(A) tract may not
be pathogenic; these changes were spread throughout the
triple A gene. Two families had the classical features of triple
A syndrome, but no mutations were detected, suggesting
genetic heterogeneity. In these two cases, no other family
members were available for us to carry out haplotype analysis
using markers on chromosome 12q13. Families with mutations had the classical triple A phenotype of achalasia,
alacrima, adrenal abnormalities and progressive neurological
abnormalities. A number of associated neurological features
are described in detail in these families, such as preferential
ulnar nerve involvement with distal amyotrophy, widespread
autonomic nervous system abnormalities, corticospinal tract
damage and cranial nerve palsies. Genotype±phenotype
analysis based on these families revealed that frameshift,
stop codon and functionally signi®cant mutations are likely to
lead to a more severe phenotype, most probably occurring by
a loss-of-function effect on the protein. Further mutation
screening and functional analysis will lead to a greater
understanding of this ubiquitous developmental gene.
Acknowledgements
We are grateful to the families for their help. This research
was supported by The Wellcome Trust.
References
Allgrove J, Clayden GS, Grant DB, Macaulay JC. Familial
glucocorticoid de®ciency with achalasia of the cardia and
de®cient tear production. Lancet 1978; 1: 1284±6.
Beaudoing E, Freier S, Wyatt JR, Claverie J-M, Gautheret D.
Patterns of variant polyadenylation signal usage in human genes.
Genome Res 2000; 10: 1001±10.
Bentes C, Santos-Bento M, de Sa J, de Lurdes Sales Luis M, de
Carvalho M. Allgrove syndrome in adulthood. Muscle Nerve 2001;
24: 292±6.
Chu ML, Berlin D, Axelrod FB. Allgrove syndrome: documenting
cholinergic dysfunction by autonomic tests. J Pediatr 1996; 129:
156±9.
Clark AJ, Weber A. Adrenocorticotropin insensitivity syndromes.
[Review]. Endocr Rev 1998; 19: 828±43.
2689
De Lisa JA, Lee HJ, Baran EM, Lai K-A, Spielholtz N,
MacKenzie K. Manual of nerve conduction velocity and clinical
neurophysiology, 3rd edn. New York: Raven Press; 1994.
Dumic M, Radica A, Sabol Z, Plavsic V, Brkljacic L, Sarnavka V,
et al. Adrenocorticotropic hormone insensitivity associated with
autonomic nervous system disorders. Eur J Pediatr 1991; 150: 696±9.
Ehrich E, Aranoff G, Johnson WG. Familial achalasia associated
with adrenocortical insuf®ciency, alacrima, and neurological
abnormalities. Am J Med Genet 1987; 26: 637±44.
Gazarian M, Cowell CT, Bonney M, Grigor WG. The `4A'
syndrome: adrenocortical insuf®ciency associated with achalasia,
alacrima, autonomic and other neurological abnormalities. Eur J
Pediatr 1995; 154: 18±23.
Geffner ME, Lippe BM, Kaplan SA, Berquist WE, Bateman JB,
Paterno VI, et al. Selective ACTH insensitivity, achalasia, and
alacrima: a multisystem disorder presenting in childhood. Pediatr
Res 1983; 17: 532±6.
Grant DB, Dunger DB, Smith I, Hyland K. Familial glucocorticoid
de®ciency with achalasia of the cardia associated with mixed
neuropathy, long-tract degeneration and mild dementia. Eur J
Pediatr 1992; 151: 85±9.
Grant DB, Barnes ND, Dumic M, Ginalska-Malinowska M, Milla
PJ, von Petrykowski W, et al. Neurological and adrenal dysfunction
in the adrenal insuf®ciency/alacrima/achalasia (3A) syndrome.
Arch Dis Child 1993; 68: 779±82.
Hadj-Rabia S, Salomon R, Pelet A, Penet C, Rotschild A, de Laet
MH, et al. Linkage disequilibrium in inbred North African families
allows ®ne genetic and physical mapping of triple A syndrome. Eur
J Hum Genet 2000; 8: 613±20.
Hammami A, Trabelsi M, Bennaceur B, Boukhris R. Association
d'une maladie d'Addison, d'une achalasie du cardia et d'une
alacrymation. A propos de deux observations. [Review]. Ann
Pediatr (Paris) 1989; 36: 279±82.
Handschug K, Sperling S, Yoon SJ, Hennig S, Clark AJ, Huebner
A. Triple A syndrome is caused by mutations in AAAS, a new WDrepeat protein gene. Hum Mol Genet 2001; 10: 283±90.
Heinrichs C, Tsigos C, Deschepper J, Drews R, Collu R, Dugardeyn
C, et al. Familial adrenocorticotropin unresponsiveness associated
with alacrima and achalasia: biochemical and molecular studies in
two siblings with clinical heterogeneity. Eur J Pediatr 1995; 154:
191±6.
HuÈbschmann K. Achalasie, Alakrimie und CortisolmangelÐdas
Allgrove Syndrom. [Review]. Klin Padiatr 1995; 207: 126±9.
Huebner A, Elias LL, Clark AJ. ACTH resistance syndromes.
[Review]. J Pediatr Endocrinol Metab 1999; 12 Suppl 1: 277±93.
È zkinay F, Hilscher C, Lee H, Clark AJ, et al.
Huebner A, Yoon SJ, O
Triple A syndromeÐclinical aspects and molecular genetics.
Endocr Res 2000; 26: 751±9.
Lee H, Choi E, Seomun Y, Montgomery K, Huebner A, Lee E, et al.
High-resolution transcript map of the region spanning D12S1629
and D12S312 at chromosome 12q13: triple A syndrome-linked
region. Genome Res 2000; 10: 1561±7.
2690
H. Houlden et al.
Loewenfeld IE, Thompson HS. The tonic pupil: a re-evaluation. Am
J Ophthalmol 1967; 63: 46±87.
Mathias CJ, Bannister R. Investigation of autonomic disorders. In:
Mathias CJ, Bannister R, editors. Autonomic failure: a textbook of
clinical disorders of the autonomic nervous system. 4th edn.
Oxford: Oxford University Press; 1999. p. 169±95.
Moore PS, Couch RM, Perry YS, Shuckett EP, Winter JS. Allgrove
syndrome: an autosomal recessive syndrome of ACTH insensitivity,
achalasia and alacrima. Clin Endocrinol (Oxf) 1991; 34: 107±14.
Neer EJ, Schmidt CJ, Nambudripad R, Smith TF. The ancient
regulatory-protein family of WD-repeat proteins. [Review]. Nature
1994; 371: 297±300.
Sandrini F, Farmakidis C, Kirschner LS, Wu SM, Tullio-Pelet A,
Lyonnet S, et al. Spectrum of mutations of the AAAS gene in
Allgrove syndrome: lack of mutations in six kindreds with isolated
resistance to corticotropin. J Clin Endocrinol Metab 2001; 86:
5433±7.
Sievers J. Basic two-dye stains for epoxy-embedded 0.3±1m
sections. Stain Technol 1971; 46: 195±9.
Stratakis CA, Lin JP, Pras E, Rennert OM, Bourdony CJ, Chan WY.
Segregation of Allgrove (triple-A) syndrome in Puerto Rican
kindreds with chromosome 12 (12q13) polymorphic markers. Proc
Assoc Am Physicians 1997; 109: 478±82.
Stuckey BG, Mastaglia FL, Reed WD, Pullan PT. Glucocorticoid
insuf®ciency, achalasia, alacrima with autonomic motor
neuropathy. Ann Intern Med 1987; 106: 61±34.
Tullio-Pelet A, Salomon R, Hadj-Rabia S, Mugnier C, de Laet MH,
Chaouachi B, et al. Mutant WD-repeat protein in triple-A
syndrome. Nature Genet 2000; 26: 332±5.
Weber A, Wienker TF, Jung M, Easton D, Dean HJ, Heinrichs C,
et al. Linkage of the gene for the triple A syndrome to chromosome
12q13 near the type II keratin gene cluster. Hum Mol Genet 1996;
5: 2061±6.
Yuregir GT, Aksoy K, Curuk MA, Dikmen N, Fei YJ, Baysal E,
et al. Hb H disease in a Turkish family resulting from the interaction
of a deletional alpha-thalassaemia-1 and a newly discovered poly A
mutation. Br J Haematol 1992; 80: 527±32.
Smith SE, Smith SA, Brown PM, Fox C, Sonksen PH. Pupillary
signs in diabetic autonomic neuropathy. Br Med J 1978; 2: 924±7.
Zeharia A, Shuper A, Mimouni M, Kornreich L, Rachmel A,
Lerman-Sagie T. Periventricular brain heterotopias in a child with
adrenocortical insuf®ciency, achalasia, alacrima, and neurologic
abnormalities (Allgrove syndrome). J Child Neurol 1999; 14:
331±4.
Smith TF, Gaitatzes C, Saxena K, Neer EJ. The WD repeat: a
common architecture for diverse functions. [Review]. Trends
Biochem Sci 1999; 24: 181±5.
Received February 26, 2002. Revised June 13, 2002.
Accepted June 22, 2002
Smith SA, Smith SE. Bilateral Horner's syndrome: detection and
occurrence. J Neurol Neurosurg Psychiatry 1999; 66: 48±51.