|Home | About | Journals | Submit | Contact Us | Français|
Hereditary neuralgic amyotrophy (HNA) is an autosomal dominant disorder that manifests as recurrent, episodic, painful brachial neuropathies. A gene for HNA maps to chromosome 17q25.3 where mutations in SEPT9, encoding the septin-9 protein, have been identified.
To determine the frequency and type of mutations in the SEPT9 gene in a new cohort of 42 unrelated HNA pedigrees.
DNA sequencing of all exons and intron-exon boundaries for SEPT9 was carried out in an affected individual in each pedigree from our HNA cohort. Genotyping using microsatellite markers spanning the SEPT9 gene was also used to identify pedigrees with a previously reported founder haplotype.
Two missense mutations were found: c.262C>T (p.Arg88Trp) in seven HNA pedigrees and c.278C>T (p.Ser93Phe) in one HNA pedigree. Sequencing of other known exons in SEPT9 detected no additional disease-associated mutations. A founder haplotype, without defined mutations in SEPT9, was present in seven pedigrees.
We provide further evidence that mutation of the SEPT9 gene is the molecular basis of some cases of hereditary neuralgic amyotrophy (HNA). DNA sequencing of SEPT9 demonstrates a restricted set of mutations in this cohort of HNA pedigrees. Nonetheless, sequence analysis will have an important role in mutation detection in HNA. Additional techniques will be required to find SEPT9 mutations in an HNA founder haplotype and other pedigrees.
Hereditary neuralgic amyotrophy (HNA) is an autosomal dominant disorder characterized by painful, episodic, focal motor and sensory attacks, primarily affecting the nerves of the brachial plexus.1–3 Associated findings in some individuals with HNA include relative hypotelorism, occasional cleft palate, and skin folds or creases on the neck or forearm.1
A gene for HNA was localized to human chromosome 17q25.3 in several pedigrees.4–6 In probands of six unrelated HNA pedigrees linked to chromosome 17q25.3, we identified three different point mutations in the alternatively spliced SEPT9 gene that encodes septin 9 proteins.7 One mutation was a c.-131G>C transversion in the 5′ untranslated region (UTR) found in a Turkish family (nucleotide is based on SEPT9 transcript variant 3, RefSeq NM_006640.4 in GenBank).8–10 Additionally, two mutations causing missense changes in septin-9 protein isoform c, c.262C>T (p.Arg88Trp) and c.278C>T (p.Ser93Phe), were detected in Europeans and North American descendants of Europeans (RefSeq NP_006631.2).8,9 Additional HNA pedigrees from Europe and North America have the c.262C>T mutation.11,12
In this study, we undertook sequence analysis of the SEPT9 gene in a large cohort of 42 pedigrees with HNA that were not included in our HNA gene identification report.7 This analysis was performed to confirm that SEPT9 is the critical gene for HNA, to search for additional mutations, and to estimate the likelihood of detecting mutations in this gene given a clinical phenotype consistent with HNA.
A diagnosis of HNA was established based on published criteria, including the presence of autosomal dominant inheritance.3,13 These clinical features typically included a history of one or more sudden onset, painful attacks in the neck, shoulder, or arm, followed by focal paresis, sensory disturbances, and muscle atrophy. Typically, the attacks involved the brachial plexus; however, other peripheral nerves (e.g., lumbosacral plexus, phrenic nerve) were affected in some cases. Probands and other affected and at risk subjects were evaluated by P.F.C., T.D.B., or referring physicians. Written consent was obtained from the subjects. The Human Subjects Division at the University of Washington approved this study. The clinical features for eight pedigrees included in this cohort of 42, K4001, K4004, K4006, K4012 (SAL819), K4026 (S Family), K4027 (G Family), K4028, and K4043 (classic type), have been described previously.1,14–17 We excluded pedigrees in our cohort that do not link to chromosome 17q25.3: K4008, K4016, and K4044 (chronic undulating type).17,18
Genomic DNA was isolated from peripheral blood or lymphoblastoid cell lines of subjects by standard methods using a Gentra Puregene Blood Kit (Qiagen, Valencia, CA). DNA from an affected individual from each pedigree was subjected to SEPT9 mutational analysis by sequencing known exons and adjacent intron-exon boundaries. Typically, the proband was chosen to be sequenced because we often had the best clinical evidence that they had features of HNA. For larger pedigrees, both clinical and genetic criteria were used to choose an affected individual who had an affected parent and, if possible, an affected offspring. The regions surrounding exons of the SEPT9 gene were amplified by thermal cycling and subjected to bidirectional sequencing according to previously published methods.7 Thermal cycling primers are provided in table e-1 on the Neurology® Web site at www.neurology.org. In each proband where nucleotide changes that varied from the human genome reference sequence were found, the sequence was examined for known single nucleotide polymorphisms (SNPs) and segregation with disease in the pedigree. At least 100 individuals in a control population, unaffected by HNA, were also examined by direct sequencing or restriction fragment length polymorphism analysis to determine if the same SEPT9 variant was present in a healthy, unrelated population (table).
Short tandem repeat (STR) markers, which were used previously to define the conserved shared block in the original five HNA founder pedigrees, were used to determine if pedigrees in the present cohort also possessed the founder haplotype.7,19 Following amplification with the STR primers, products were sized on an ABI 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA). Segregation patterns of the alleles were determined, and haplotypes constructed in a manner to minimize crossovers.
We screened a total of 42 pedigrees with HNA. SEPT9 gene mutations are summarized in the table. The locations of the genomic changes are shown, along with the position in transcript variant SEPT9_v3 and the corresponding protein isoform SEPT9c. Two of the previously reported SEPT9 mutations were also identified among these 42 HNA pedigrees, the c.262C>T (p.Arg88Trp) and the c.278C>T (p.Ser93Phe) mutation.7,11,12 In seven pedigrees, K4009, K4020, K4026, K4027, K4033, K4052, and K4059, the c.262C>T mutation was found in affected individuals. In one pedigree, K4012, from France, the c.278C>T mutation was identified. Segregation of these mutations within HNA pedigrees is shown in the figure.
No other causal mutations were found in the other 34 pedigrees, although one coding, nonsynonymous single nucleotide polymorphism (SNP rs34587622 in dbSNP) was found to segregate with HNA in a subset of seven.20 The rs34587622 SNP, c.380C>T, changes the proline at amino acid position 128 to leucine in SEPT9 isoform c (p.Pro127Leu in NP_006631.2). This SNP was found initially in a single proband, who also possessed the c.262C>T (p.Arg88Trp) mutation, and in controls during candidate gene sequencing in HNA pedigrees.21 The c.380T allele is present within five founder haplotype HNA pedigrees previously identified.7,19 This allele also segregates with HNA in 7 of 42 pedigrees in this cohort that all bear the common North American founder haplotype (data not shown). However, because the c.380C>T allele is present in 28% (29/104) of control individuals of Caucasian descent, and the dbSNP database finds it in 15.4% (4/26) of North Americans and 5% (6/120) of Europeans, we do not believe it is a causal mutation.20
To date, three point mutations (c.-131G>C, c.262C>T, and c.278C>T) and a genetic founder haplotype have been identified in HNA pedigrees supporting a critical role for the SEPT9 gene.7,11,12 Combining the results from the 42 pedigrees in this report with the 7 pedigrees previously reported from our laboratory, we detected SEPT9 mutations in 22.4% (11/49) of our total cohort.7,11 A further 24.5% of these HNA pedigrees possess the founder haplotype defined by the markers previously reported.7,19 The moderate mutation detection rate may be due to several factors. First, our present methods may be unable to detect small SEPT9 insertions, deletions, inversions, or other gene rearrangements. For example, it will likely be necessary to undertake complete genomic sequencing of SEPT9 in a subject having the founder haplotype in order to detect a causal mutation in this subset of HNA pedigrees. Second, more subtle regulatory mutations may occur outside of the region of exons and intron-exon boundaries that could, in turn, influence the relative abundance of specific SEPT9 alternative 5′ exon isoforms. Third, genetic heterogeneity does exist for HNA. To date, the gene for HNA in five pedigrees has been shown to be unlinked to markers in the 17q25.3 region.17,18,22 In some cases, HNA pedigrees are not large enough to confirm or exclude chromosome 17q25.3 linkage. Fourth, the potential for phenocopies may exist, with familial clustering of similar, but distinct multifocal, painful neuropathies that may be caused by peripheral nerve vasculitis, an example of which can be seen in proximal diabetic neuropathy or diabetic amyotrophy.23,24
HNA is the first Mendelian disorder shown to be due to mutations in a member of the septin gene family of related proteins. The septin gene family includes at least 14 members in humans.25,26 Septins participate in cytokinesis and cellular trafficking and have been studied for their relationship to neoplasia.27 Transcriptional and potentially translational regulation of the SEPT9 gene is complex.28 SEPT9 appears to be expressed ubiquitously, but information regarding the distribution and abundance of the various septin-9 protein isoforms in normal tissues is limited.29 The longer isoforms of septin-9 have unique short N-terminal polypeptides and share a proline-rich domain that is found only in the septin-4 and septin-8 proteins.25 Septin-9 has been shown to colocalize with other septins and with septin intermediate filaments that associate with actin microfilaments and microtubules.30,31
At the transcriptional level, the SEPT9 gene shows a remarkable number of alternative first exons. These multiple 5′ alternative transcription start sites, as well as possible alternative splicing within the final exon, potentially create dozens of protein isoforms.28 The two mutations reported here, c.262C>T at g.Chr17:72,909,975C>T and c.278C>T at g.Chr17:72,909,993C>T, can lead to missense mutations in the translated product of the SEPT9_v1, SEPT9_v2, and SEPT9_v3 transcripts, but are in the 5′ untranslated region of SEPT9_v4 and SEPT9_v4* transcripts.28,32 This diversity of transcripts has made analysis of the functional consequences of SEPT9 gene mutations difficult.32 Under hypoxic conditions, the c.262C>T mutation reportedly affects the level of translation of the SEPT9_v4 transcript, encoding SEPT9 protein isoform e (GenBank NM_001113494.1 and NP_001106966.1).32
The HNA-associated mutations in SEPT9_v3, p.Arg88Trp and p.Ser93Phe, altered the colocalization of mutant septin-9 protein isoform c with septin-4 and septin-11 in mesenchymal and epithelial murine mammary gland NMuMG cells.33 These mutations also perturb the regulation of septin-9-containing filaments by Rho/Rhotekin signaling.33 Thus, this N-terminal proline-rich domain, in which these mutations reside, may regulate specific interactions of septin-9 with other septins or other cellular proteins as proposed by Nagata et al.31,34 These findings are only the first steps in determining how mutations in SEPT9 lead to the distinctive phenotype seen in HNA. The mechanism of how SEPT9 mutations lead to a brachial neuropathy and other features of HNA remains unknown.
In this study, we provided further evidence that mutation of the SEPT9 gene is the molecular basis of some cases of HNA. We also found that the detection rate for SEPT9 mutations in pedigrees consistent with this disorder is low. This is due to few identifiable mutations found by our present mutation analysis strategy of sequencing SEPT9 exons. This strategy has not found a mutation in the significant fraction of HNA pedigrees that harbor a genetic founder chromosome. Clearly, identification of the specific SEPT9 mutation associated with the genetic founder will greatly improve the clinical usefulness of SEPT9 analysis as a diagnostic and predictive tool for HNA.
The authors thank Jonathan Adkins and Melissa Eckert for DNA isolation and lymphoblastoid cell line establishment and all of the clinical contributors, patients, and families for facilitating pedigree identification and obtaining research materials.
Address correspondence and reprint requests to Dr. Mark C. Hannibal, Neurogenetics Laboratory, 1959 NE Pacific St., Health Sciences RR236, Box 356320, Division of Genetics and Developmental Medicine, Department of Pediatrics, University of Washington School of Medicine, Seattle, WA 98195-6320 ude.notgnihsaw.u@innahm
Supplemental data at www.neurology.org
Supported by funds from the NIH (National Institute of Neurological Disorders and Stroke), NS38181 (P.F.C. and M.C.H.); The Neuropathy Association, New York, NY (M.C.H. and P.F.C.); and the Allan and Phyllis Treuer Endowed Chair for Genetics and Development (P.F.C.).
Disclosure: Dr. Chance has received Speaker’s Bureau honoraria from Athena Diagnostics, Inc. Dr. Bird has received licensing fees from Athena Diagnostics, Inc.
Medical Device: Gentra Puregene Blood Kit (Qiagen, Valencia, CA).
Received August 11, 2008. Accepted in final form February 13, 2009.