Abstract
Background: Charcot-Marie-Tooth disease (CMT) is the most common inherited neuromuscular disorder and is characterized by significant clinical and genetic heterogeneity. Recently, mutations in both the small heat shock protein 27 (HSP27 or HSPB1) and 22 (HSP22 or HSPB8) genes have been reported to cause autosomal dominant CMT with minimal sensory involvement (CMT 2F/CMT2L) and autosomal dominant distal hereditary motor neuropathy type II (dHMN II).
Methods: We analyzed the HSPB1 and HSPB8 genes in a large clinically well-characterized series of dHMN and CMT type 2 (CMT2) cases and families using linkage analysis and direct sequencing of these genes.
Results: We identified a novel homozygous mutation in the α-crystallin domain of HSPB1 segregating in an autosomal recessive fashion in a family with distal HMN/CMT2. A further four heterozygous HSPB1 mutations were identified in four autosomal dominant families dHMN/CMT2, and two sporadic cases were identified with probable de novo mutations. In the autosomal dominant and autosomal recessive families, there were no clinical sensory findings, but reduced sural nerve action potential amplitudes were found in some affected individuals, indicating that long sensory axons are mildly affected in this predominantly motor disorder.
Conclusions: This extends the clinical and electrophysiologic spectrum of HSPB1 mutations and identifies four unreported dominant HSPB1 mutations and the first family where the HSPB1 mutation acts in a recessive way to cause distal HMN.
Glossary
- CMAP=
- compound muscle action potential;
- CMT=
- Charcot-Marie-Tooth disease;
- dHMN=
- distal hereditary motor neuropathy;
- HSP=
- heat shock protein;
- NCS=
- nerve conduction studies.
Charcot-Marie-Tooth disease (CMT) is a common, clinically and genetically heterogeneous group of hereditary neuropathies with an estimated prevalence of 1 in 2,500.1,2 Distal hereditary motor neuropathy (dHMN) resembles CMT but is a pure motor neuron syndrome with no sensory involvement. The dHMNs were previously subdivided into seven types based on clinical features, age at onset, and inheritance, but the identification of disease genes in this group of disorders has revealed greater subtypes and heterogeneity within these groups.3–5
Recently, five missense mutations in the HSPB1 gene in autosomal dominant CMT type 2F (minimal sensory involvement) and distal HMN type II families were identified. Four of the mutations occurred in the highly conserved α-crystallin domain of the protein, the fifth in the C-terminus region.6 Mutations were associated with dysfunction of the axon cytoskeleton, formation of aggresomes, and oxidative cell death. The HSPB1 R127W mutation has also been identified in Chinese CMT2 patients7 and a Pro182Ser mutation has been identified in a dHMN proband from Japan.8
The HSPB1 gene is a member of the HSP superfamily of genes thought to be protective stress proteins.9 HSPB8 is also a member of this family of genes and two missense mutations have also been identified in the α-crystallin domain of this gene in European, Chinese, and Japanese CMT2 and distal HMN families.10,11 A promoter polymorphism has also been identified in HSPB1 that impairs the stress response of this protein and may be a risk factor in motor neuron disease.12 Treatment with arimoclomol, a coinducer of heat shock proteins, delays disease progression in SOD1 mutant ALS mice, implying that these proteins are important to motor neuron survival.13
Here we report the genetic analysis of the HSPB8 and HSPB1 genes in large cohort distal HMN and CMT2 families and sporadic cases. Novel HSPB1 mutations were identified in both autosomal dominant and autosomal recessive families, revealing further insights into the pathogenesis of this group of hereditary disorders.
METHODS
Patients.
Ethics approval was obtained from the joint medical and ethics committee at The National Hospital for Neurology and Neurosurgery to perform this clinical and genetic study.
We identified and characterized 25 families with clinically diagnosed autosomal dominant or autosomal recessive and a further 30 sporadic dHMN cases. These cases had distal lower and upper limb weakness, wasting, and reduced reflexes. These patients had no clinical sensory symptoms or signs. The SMN deletion and chromosome 17 duplication were excluded in these families. We also analyzed the probands from 90 CMT2 families with a typical progressive history of sensory and motor symptoms, signs of weakness and wasting, and an axonal neuropathy identified with nerve conduction studies (NCS). The families had between two and nine affected individuals. Acquired causes of neuropathy were excluded in all cases. There was likely consanguinity in two dHMN and 2 CMT2 families. These families were negative for the chromosome 17p11.2 duplication, PMP22 gene sequencing, MPZ, and Cx32 in all cases; the MFN2 gene had been excluded in 12 cases. The majority of families were seen in the peripheral nerve or the neurogenetics clinic at the National Hospital for Neurology. A small number of cases were sent as DNA from other neurologic centers. DNA was extracted from blood samples obtained with informed consent from affected and unaffected individuals. Linkage analysis was carried out in two families with autosomal dominant dHMN type II (family 1 and family 8). The proband from each dHMN family was sequenced for the HSPB8 and HSPB1 genes, as were the sporadic cases. In the CMT2 group the three HSPB1 exons were analyzed as mutations have been identified in all three regions, and in HSPB8 exon 2 was analyzed as only this region has so far been identified to contain mutations. In the families with HSPB1 mutations other family members were collected and screened for the mutation identified to prove segregation.
Linkage analysis.
Microsatellite markers (Applied Biosystems, v2.5) around the HSPB1 and HSPB8 loci were genotyped to assess linkage in two distal HMN families, family 1 with 10 affected individuals and family 8 with 8 affected. For the HSPB1 gene the microsatellite markers D7S519, D7S502, D7S669, D7S630, and D7S657 were used and for HSPB8, D12S346, D12S78, D12S79, D12S86, and D12S324. PCR was performed according to the manufacturer protocol, and the products were separated on a 3700 DNA analyzer (Applied Biosystems). The program Genotyper was used to analyze the data. Linkage analysis was carried out using MLINK (Lathrop et al., 1984) and SimWalk.14 A disease-allele frequency of 0.0001 was estimated, equal marker allele frequencies were assumed, only affected individuals and unaffected spouses were included in the linkage analysis and lod-score calculations.
Genetic sequencing.
HSPB1 and HSPB8 genes exons and flanking introns were sequenced (primers available on request). PCR was carried out in a total volume of 50 μL, which contained 20 ng of DNA, 0.2 mM dNTPs, 1 unit ABgene Taq polymerase, 1.5 mM MgCl2, 75 mM Tris-HCl pH 9.0, 20 mM (NH4)2SO4, 0.01% Tween 20, and 50 pmoles of each primer. PCR was performed on a Perkin-Elmer 9700 thermal Cycler (Perkin Elmer, Applied Biosystems, Foster City, CA). The cycling consisted of denaturation at 94°C for 15 minutes, followed by 25 cycles of 94°C for 30 seconds, 60°C to 50°C touch down protocol for 30 seconds, and 72°C for 30 seconds. After that, 12 cycles with a constant annealing temperature at 50°C and a final amplification at 72°C for 10 minutes were carried out. PCR fragments were checked by electrophoresis on a 1% agarose gel. The PCR products were purified by using Millipore PCR cleanup plates and resuspended in 50 μL of deionized water. For each exon, 100 ng of amplified product was sequenced using forward and reverse primers and a BigDye Terminator cycle sequencing kit (Perkin-Elmer). Sequence reactions were cleaned up using ABgene dye terminator removal plates. Sequencing was performed on an ABI3730XL automated sequencer and alignment and analysis were carried out with Sequence Navigator (Perkin Elmer).
Identified mutations were validated by repeat sequencing and to confirm segregation in other affected and unaffected family members. None of the mutations were identified in 220 UK and 105 Asian controls that were screened.
RESULTS
Details of mutations are listed in table 1 and given in figures 1 and 2. Five out of 25 familial and 2 out of 30 sporadic dHMN cases had HSPB1 mutations, no HSPB8 mutations were found. No mutations were identified in the CMT2 cases screened.
Table 1 Families and sporadic cases with HSPB1 gene mutations
Figure 1 Pedigrees of the families identified with HSP27 mutations
An arrow indicates the proband, a square a male, a circle a female, a filled symbol indicates affected, half filled indicates possible affectation, and an arrow through a symbol indicates the individual is deceased. The individuals are indicated as affected if they have been examined and found to be so. +/− = Heterozygous, +/+ = homozygous affected, −/− = homozygous normal for the HSPB1 mutation.
Figure 2 HSP27 sequence analysis with electropherograms showing the missense mutations and the sequence variant identified in the distal HMN families with controls
In family 5, the mutation acts in a recessive fashion as the proband is homozygous whereas his unaffected father is heterozygous.
The linkage analysis carried out in family 1 revealed a multipoint lod score of 3.087 at the marker D7S484, flanking markers confirmed linkage to the HSPB1 gene region in this family. This family was subsequently shown to have an HSPB1 mutation (S135F) as discussed below. This mutation segregated with the disease in the eight affected individuals available for sampling and not in the unaffected cases. Linkage to HSPB8 and HSPB1 was excluded in family 8 with lod scores across the markers of this region <−2.0 (figures 3 and 4).
Figure 3 HSP27 sequence analysis showing the clustal multiple protein alignments of HSP27 orthologs
The HSP-α-crystallin domain is indicated at the bottom of the figure in gray. The mutations are highlighted in red and nonconserved amino acids are indicated in blue. All mutations are highly conserved and three are located in the HSP-α-crystallin domain. Four out of five mutations were also conserved in the α-A (HSPB4) and α-B (HSPB5) crystalline genes.
Figure 4 HSP27 (HSPB1) contains three exons shown in the gray boxes
Intron 2 between exons 1 and 2 is 725 bp and intron 3 between exons 2 and 3 is 118 bp. The start ATG (codon 1) is indicated in the figure and all mutations are labeled from this codon up to the stop codon TAA at 206. Reference sequence NM_001540. The reported HSP27 mutations and the mutations identified here are indicated on the figure. The position of these mutations is shown in relation to the α-crystallin and the HSP20 domains, which are drawn to scale below the HSP27 gene figure.
We sequenced the HSPB8 and HSPB1 genes in our cohort of distal HMN/CMT2 cases. Seven HSPB1 mutations were identified (table 1, figure 2,A–C). Four mutations were identified in autosomal dominant families (S135F, R140G, P39L, and G84R) (family trees). Three of these mutations were novel (R140G, P39L, and G84R) and one had previously been reported (family 1, S135F).6 An autosomal recessive distal HMN/CMT2 family (figure 1, family 5) was identified where the proband was homozygous for a novel L99M mutation, his father and an unaffected sister were heterozygous. The father aged 75 years and sister aged 41 years were examined by a neurologist (H.H.) and were asymptomatic and clinically normal. The proband came from a consanguineous Pakistani family where the parents were cousins. The mother died of an unrelated illness in her 50s but had no reported similarities to her son. No other affected members in the family were identified.
Family 2 (figure 1) came from Northern India where the proband’s father, possibly grandfather, and one child were affected. At the age of 6 years he had difficulty with his feet but played squash to a high level in his 20s. In the mid 30s he noticed distal limb weakness after a long walk and since then has progressed with lower limb and later upper limb distal weakness and wasting. The affected father was also heterozygous for the mutation and the clinically unaffected brother was normal homozygous. One of the proband’s children has symptoms but has not been examined. The proband’s father also had the typical clinical features of distal HMN and this was electrically confirmed. His age at onset was around 52 years. In family 3, the proband had an age at onset of 54 years but few details were available on this case apart from his neurologic diagnosis of distal HMN. His mother had walking problems in her 50s and used a stick but she was deceased. There were no children or other family members to clarify the phenotype or look for segregation of the mutation. Family 4 consisted of two affected brothers who had no children; their father who is deceased had walking problems but also had diabetes. The age at onset in the proband’s affected brother was 40.
In two apparently sporadic dHMN cases (figure 1, families 6 and 7) the HSPB1 mutation R140G was also identified. In both patients’ families there were no other affected relatives and the parents were said to be healthy. This mutation is the same as family 2. All three families were originally from India and have the same surname, which suggests they may be genetically related.
One sequence variant in the coding region of HSPB1 was identified as a heterozygous C379A change, causing a silent R127R change. This individual was a sporadic case with typical dHMN with no other affected individuals in the family and unaffected parents (figure 2). This change was not identified in 220 controls and it is not predicted to cause a splice site change and therefore, the pathogenicity remains unknown. No mutations were identified in the cohort of CMT2 families.
The clinical features from the probands of each of the seven families are detailed in table 2. All patients had a remarkably similar slowly progressive disease course. Muscle weakness and atrophy started and predominated in the distal lower limb muscles. Over several years the weakness and wasting progressed to the upper limbs approximately 5–10 years later along with proximal lower limb problems. Sensory disturbances were absent in all patients except the proband from family 4; he had a 20-year history of insulin-dependent diabetes. His sensory symptoms of pain, paraesthesia, and distal limb numbness only came on after several years of diabetes. Tendon reflexes were depressed or absent in all cases. Only one patient (family 6) had muscle fasciculation and tremor.
Table 2 Clinical features of the probands with HSPB1 mutations
Patients with an early onset were more severely affected. The only variable feature was a later age at onset in the cases with HMN mutations out of the α-crystallin domain and the two sporadic cases with the R140G mutation also had a later age at onset.
Electrophysiologic findings.
Neurophysiologic studies were carried out on the probands from each of the seven families as given in table 3. Neurophysiology in affected individuals showed reduced (0.4 to 1.1 mV) or absent lower limb compound muscle action potential (CMAP) amplitudes. In the upper limbs CMAPs were small or in the normal range (median 5.3 to 7.6 mV, ulnar 2.1 to 6.7). Nerve conduction velocities were in the normal range in the upper limbs. EMG showed severe chronic denervation, more pronounced in the distal lower limb muscles. The sensory nerve conduction velocities were normal in the upper limbs. In the lower limbs the sural amplitude was reduced (3 μV) or not recordable in four patients. Sensory nerve conduction study in the upper limbs was within normal limits in all patients. The abnormal sural response in one patient from family 1 may be related to age (proband 74) and to lower limb edema in family 3. The neurophysiologic findings in the proband from family 4 are difficult to interpret since the test was performed at age 70 after a 20-year history of insulin-dependent diabetes. There are no electrophysiologic data prior to the onset of diabetes. In family 5 with presumed recessive inheritance, despite the lack of sensory symptoms or signs, the sural nerve action potentials were reduced in the proband at 3 mV.
Table 3 Electrophysiologic data in the seven probands from the families with HSPB1 mutations
DISCUSSION
In this study we screened 115 familial dHMN/CMT2 cases and 30 sporadic dHMN cases for mutations in the HSPB1 and HSPB8 genes. Seven HSPB1 mutations were identified in four dominant, one recessive, and two sporadic families. No CMT2 mutations were identified, but this should not preclude this group from further screening as the sensory finding in CMT2 can often be minor. The most severe phenotype is observed in family 1 with a S135F mutation. The lack of mutations in the HSPB8 gene confirms that families with defects in this gene are rare. The sporadic cases are possibly due to de novo mutations, although DNA from the parents was not available to prove this. In the one family with a recessive HSPB1 mutation, the proband was homozygous and the clinically unaffected father and sister were heterozygous for the L99M mutation. The proband with the recessive mutation had typical dHMN/CMT2 with an age at onset of 37 years. This mutation was not found in control individuals and the amino acid residue is conserved across species as well as the α-A and α-B crystalline proteins. These data indicate that certain HSPB1 mutations can act in an autosomal recessive fashion and this is the first family to be reported.
Four mutations were identified in dominant dHMN/CMT2 families. In sporadic HMN cases, 2/30 (6%) cases had mutations. The R140G mutation is particularly important and interesting. We identified this change in familial as well as sporadic HMN/CMT2 families that all originated from Northern India. The position of the R140G mutation is important and implies that it is pathogenic given that it corresponds to the R120 mutation in the α-B crystalline (HSPB5) gene which is mutated in desmin related myopathy,15 the R116 mutation in α-A crystalline (HSPB4) gene which is mutated in familial congenital cataracts,16 and its proximity to the K141 mutation in the HSPB8 gene. The R140G mutation is also found in three Indian families and it is likely these cases are genetically related although this is not proven.
All the mutations identified were in cases with a classic dHMN phenotype or CMT2 with minimal sensory involvement. Neurophysiology showed a predominant motor neuropathy with reduced CMAP amplitudes and chronic partial denervation on EMG. Clinically the patients with HSPB1 mutations had a predominant motor neuropathy. There were no sensory symptoms or signs, except in the proband of family 4 who had 20 years of insulin-dependent diabetes starting 12 years after the onset of his dHMN. The SAP amplitudes were borderline or reduced in all families; they were absent in families 1, 3 (likely due to peripheral edema), and 4 (severe diabetes). This suggests that minor sensory abnormalities are not uncommon in HMN with HSPB1 mutations as previously suggested as there are families with sensory NCS abnormalities.6 The sensory data are surprising given that HSPB1 mutations primarily affect the motor neurons.
A number of roles have been identified for HSPB1 in the central and peripheral nervous system including regulation and maintenance of various components of the cytoskeleton,17,18 an interaction with several members of the intermediate filament family, and disruption of the NF assembly.6 The HSPB1 protein has been shown to be protective in many of these studies.19 Mutant HSPB1 also fails to be transported within neurites of cortical neurons, forms intracellular aggregates, and leads to the sequestration in the cytoplasm of selective cellular components, including neurofilament middle chain subunit (NF-M) and p150 dynactin.20 The dominant mutations that we have identified here are likely to act in the same way as previously studied dominant mutations. Mutations of amino acids within the alpha-crystallin domain (S135F,6 also reported here) and outside this region in the C-terminal part of the protein have been investigated.20 It will be interesting to carry out these functional studies on the HSPB1 P39L and G84R mutations that lie toward the N-terminus of the protein and are out of the alpha-crystallin domain.
Functional studies on the autosomal recessive HSPB1 that we report here will be important given that this change is likely to cause a similar disruption as the dominant mutations since it lies within the alpha-crystallin domain. It is likely that a heterozygous L99M mutation acts in a dose dependent fashion and is simply not enough to cause disease; given the mutation type and location, a mechanism of gain of function is most likely. The analysis of further recessive families will be important to confirm the pathogenic nature of HSPB1 mutations that act in this way.
Mutations acting in a dominant and recessive fashion are also seen in other disorders of motor nerve degeneration. An important example is seen in amyotrophic lateral sclerosis cases with the SOD1 D90A mutation, which can cause dominant and recessive disease.21,22 A tightly linked protective factor is thought to modify the toxic effect of mutant SOD1 in recessive families. The effect of modifier genes could have a similar protective effect in the recessive family we report here. Dominant and recessive mutations are seen in other types of neuropathy such as CMT4A (GDAP1 mutations)23–27; where the mutation mechanism is likely to be gain of function, heterozygous carriers of these recessive mutations are also unaffected. Characterization of the recessive and dominant mutations in neuronal cell lines, primary motor neuron cultures, and animal models will be crucial in defining their pathogenesis.
ACKNOWLEDGMENT
The authors thank the families involved in this work and in particular Courtenay Davis who was instrumental in family collection. They also thank the Jennifer Trust for spinal muscular atrophy for help with the authors’ work.
Footnotes
-
Editorial, page 1656
e-Pub ahead of print on October 1, 2008, at www.neurology.org.
The authors acknowledge the Medical Research Council (MRC) for the clinician scientist fellowship to H.H. and the Muscular Dystrophy Campaign and Brain Research Trust (BRT) for funding support. This work was undertaken at University College London Hospitals/University College London, which received a proportion of funding from the Department of Health’s National Institute for Health Research Biomedical Research Centers funding scheme.
Disclosure: The authors report no disclosures.
Received October 24, 2007. Accepted in final form April 24, 2008.
REFERENCES
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hastening the Diagnosis of Amyotrophic Lateral Sclerosis
Dr. Brian Callaghan and Dr. Kellen Quigg
► Watch
Topics Discussed
Alert Me
Recommended articles
Mutations in HSPB8 causing a new phenotype of distal myopathy and motor neuropathy
Dominant GDAP1 mutations cause predominantly mild CMT phenotypes
Phenotypic spectrum and incidence of TRPV4 mutations in patients with inherited axonal neuropathy
Sphingosine 1-phosphate lyase deficiency causes Charcot-Marie-Tooth neuropathy