A SPG3A mutation with a novel foot phenotype of hereditary spastic paraplegia in a Chinese Han family

Hereditary spastic paraplegia (HSP) (MIM#182600) is a group of heterogeneous neurodegenerative disorders, with 35 underlying loci recognized by the HGNC (HUGO Gene Nomenclature Committee; http://www.gene.ucl.ac.uk/nomenclature/) and 10 identified genes (http://www.gene.ucl.ac.uk/cgi-bin/nomenclature/searchgenes.pl plus NIPA1, last search July 2006). The mode of inheritance may be autosomal dominant, autosomal recessive or X-linked. Among these, autosomal dominant spastic paraplegia (AD-HSP) is the most common type, accounting for 70%-80% of all families.1 The disease is characterized by lower limb spasticity, hyperreflexia, progressive spastic gait and an extensor plantar response. Pes cavus is one of the commonly reported foot phenotypes.2 SPG3A (MIM# 606439), the second most common gene responsible for approximately 10% of AD-HSP, is associated with early onset, mild and noncomplicated phenotypes.3 It codes for a GTPase, atlastin, on chromosome 14q22.1. More than ten HSP causing mutations in the gene have been reported (OMIM database). Recently we reported a novel HSP mutation in a Tibetan family,4 however there has been no data regarding SPG3A mutations and associated phenotypes in the Han population. In this paper, we report our results in screening 84 AD-HSP patients from 20 Chinese Han families and 23 sporadic cases for SPG3A mutations and studies of the associated phenotypes. METHODS Subjects Chinese Han patients (n=84) from 20 families and 23 sporadic cases were investigated. All patients were diagnosed by the criteria of Harding.2 The mode of inheritance in the families was autosomal dominant. Individuals aged over 60 attending the neurological clinic who did not have HSP were recruited as controls (n=100). Informed consent was obtained from all subjects and the principles outlined in the Declaration of Helsinki were followed. Genetic analysis DNA was extracted from peripheral blood using QIAamp DNA Blood Maxi Kits (Qiagen, Germany). Mutations were detected by sequencing the PCR products flanking all 14 exons, intron-exon boundaries and the 5’and 3’ flanking regions ofthe SPG3A gene in both directions. The guidelines (http://www.hgvs.org/mutnomen/) for describing sequence variations were followed with +1 corresponding to the A of the ATG translation initiation codon of the GenBank mRNA sequence (NM_181598). Evolutionary comparisons and structural prediction Clustal X1.83 was used to compare human SPG3A cDNA sequence (NM_181598.1) and protein sequence (NP_853629.1) with the predicted orthologues of chimpanzee, cow, dog, chicken, rat, mouse, tetraodon and C. elegans (sequences retrieved from http://www.ensembl.org). Prospect was used to build the structure model, Pfam 14.0 for searching protein domains, and DAS-TMFillter for transmembrane helices. RESULTS Clinical features of theSPG3Apatients The SPG3A mutations were detected in Family 4 (Fig. 1A) and Family 3 (Fig. 1B) from Guangdong and Heilongjiang provinces respectively. The symptoms in the two families were similar. There were 13 affected individuals in the families, 2 had died since. The age at onset ranged from 1 to 6 years. Nine patients noticed the symptoms when they first started walking, one patient at age 4, the other at 6 years. The patients were examined between the ages of 4 and 67 years. Lower limb spasticity and walking difficulties progressed slowly in all the patients. To date, none needs walking aids. Hyper-reflexia and Babinski sign were observed bilaterally with ankle clonus. The upper limbs were normal. No additional neurological signs were detected. There was no evidence of urinary sphincter disturbance or cognitive deficit. The patients were generally well and could manage the activities of daily living. The adult patients were able to perform some farming work. II-6 in Family 4 could ride motorcycles and was practising in a country clinic. The overall clinical picture was of a pure, benign spastic gait disorder.Fig. 1. A:: Pedigree of Family 4 with autosomal dominant hereditary spastic paraplegia. Linkage analysis showed that the maximum Lod Score was 2.87 (θ=0) at the marker D14S75. B: Pedigree of Family 3 with autosomal dominant hereditary spastic paraplegia.Interestingly, instead of pes cavus, flat feet were found in 9 out of 11 patients in these two families (I-1, I-2, II-4, II-6, III-5, III-6 in Family 4, and II-1, II-3, III-3 in Family 3, Fig. 2). No pes cavus or scoliosis was identified in any patients. All the unaffected members in the two families had normal feet.Fig. 2. A:: II-6 of family 4, non weightbearing lateral radiographs. The different angle are: A=132°, B=10°, C=150°, D=18°. B: II-6 of family 4, weightbearing lateral radiographs. The different angle changes are: A=139°, B=8°, C=158°, D=15°. According to the diagnosis criteria of Chinese standards: the normal angle is: A= 113°-130°, or the difference of weightbearing and non weightbearing >7°, B=12°, C=129.97°-150.10°, D=21°. The angle changes in this foot radiograph subject that it is a flatfoot. C: II-4 of family 4, weightbearing anterior posterior radiographs, and D: II-4 of family 4, weightbearing lateral radiographs. The talocalcaneal angle (TC) increasesed (34° in the anterior posterior, 55° in the lateral radiograph), calcaneal pitch is decreased with depression of the medial longitudinal arch, and there is deviation of the talar first metatarsal angle (TMT) which is quite marked (162°) on the lateral film. It accords with the flat feet diagnosis criteria from USA.Electromyographs were obtained in 8 patients (5 in Family 4, 3 in Family 3). Mild neurogenic pattern in the tibial muscle were observed in 7 patients II-4, II-6, III-3, III-4 in Family 4 and II-1, II-3, III-3 in Family 3). Slowed sensory and motor conduction velocity with reduced amplitude in tibial and peroneal nerves were present in 3 patients in Family 3 (II-1, II-3, III-3). Somatosensory evoked potential (II-1, II-3, III-3 in Family 3) and motor evoked potential(II-4, II-6, II-3, III-4, III-5, III-6 in Family 4) showed reduced amplitude and delayed central conduction time in the lower limbs in all patients. Positron emission tomography of the brain was performed on two adult patients (II-4, II-6 in Family 4), which revealed normal images. Comparisons of the clinical features with SPG6 and SPG4 HSP The clinical features of our patients are quite different from the SPG4 and SPG6 AD-HSP reported from Chinese Han patients,5,6 mainly at age of onset and speed of progression of the disease (Table).Table: Comparisons of clinical characteristics of HSP with different SPG in Chinese HanGenetic analysis A c.565C>G mutation was identified in exon 5 in Family 4 (+1 corresponding to the A of the ATG translation initiation codon of the GenBank mRNA sequence NM_181598, Fig. 3A). This would result in the substitution of the histidine residue at amino acid 189 with aspartic acid (NP_853629.1). The mutation was detected in all the affected members of the family and was absent in the unaffected. All the unrelated control subjects were homozygous for C at the nucleotide position.Fig. 3. A:: Sequencing result of the affected members of Family 4 in SPG3A gene. Arrow indicates c.565C>G substitution. B: Sequencing result of the affected members of Family 3 in SPG3A gene. Arrow indicates c.715C>G substitution.A previously described mutation, c.715C>T (originally reported as C>T at SPG3A cDNA nt 884), causing a R239C change was identified in Family 3 (Fig. 3B).7 We did not find SPG3A mutations in any of the sporadic cases. Evolutionary comparisons and structural prediction Evolutionary comparisons showed a conserved C in the cDNA sequence and a conserved His in the protein sequence at the positions in all species examined for the novel mutation in Family 4 (data not shown). The structure of the wild atlastin molecule has been reported by us.4 The side chain of the H/D residues was located at the edge of the binding pocket. Therefore, the p.H189D substitution is predicted to affect guanylate binding/GTPase activation (data not shown). DISCUSSION In screening 20 families and 23 sporadic cases in Chinese Han population, we have identified one novel and one known SPG3A mutations in two families. This is the first report of SPG3A mutations from the Chinese Han population. Atlastin has the strong sequence homology to guanylate binding protein 1 (GBP1), a member of dynamin family of large GTPases.8 Atlastin is involved in guanylate binding/GTPase activation and a wide range of membrane associated functions: formation of clathrin coated pits from plasma membrane, receptor mediated endocytosis and endosome transport to the trans Golgi network.7 The disruption of the guanylate binding has provided evidence for its GTPases activation functions.7 SPG3A responsible HSP is reportedly associated with early onset, slow progression and benign and uncomplicated phenotypes. This is certainly consistent with the findings from our patients, particularly in comparison with the SPG4 and SPG6 caused AD-HSP,5,6 although another SPG3A mutation caused severe phenotypes in a Tibetan family.4 All patients in SPG3A families manage their own lives without any assistance, even in their sixth decade. In contrast, some SPG4 and SPG6 HSP patients needed walking assistance after their teens. Apart from the pyramidal sign, SPG3A AD-HSP showed no other abnormal neurological sign, whereas SPG4 or SPG6 AD-HSP often have complicated and nonpyramidal phenotypes, such as epilepsy, spastic pain of the calf muscles, cognitive deficit, extrapyramidal sign, ataxia and abnormal sensation. Although 10 mutations in the gene have already been reported, the SPG3 AD-HSP phenotypes were reportedly more uniform than those caused by other AD-HSP genes. Abel et al1 analyzed 10 different mutations in 86 SPG3A patients and found that there were no major consistent phenotypic diversities. We have recently reported severe AD-HSP phenotypes in a Tibetan family caused by a mutation predicted to disrupt the gross structure of the transmembrane regions of the Atlastin molecule.4 However, mutations reported in this paper identified in the Han population are only predicted to affect the GTP binding. This may explain the discrepancies in the severity and progression of the disease in the families. Pes cavus is a commonly described change in HSP, whereas flat feet has not been reported.2 However, it was evident in 9/11 patients at a wide range of age in our two families by either the chinese, or American criteria. Because it occurred in both families, in patients with different ages and was absent in any of the unaffected members, it is likely to be a HSP associated phenotype previously unreported. However, its underlying mechanism and significance are unknown. Our electrophysiological analysis showed delayed and blocked central conduction time in somatosensory evoked potential in Family 3 and in motor evoked potential in Family 4. The reported results for electrophysiological examinations in literature have been controversial. Alexandra9 reported increased central conduction time in motor evoked potentials, but not in somatosensory evoked potentials, whereas Scarano8 reported contrary results. Reviewing the available literature, most investigators did not observe any abnormalities in electromyography and nerve conduction velocity. However, recently Scarano and his colleagues8 reported a sensory motor, axonal neuropathy in SPG3A patients with R495W mutation. Consistent with this, patients in Family 3 showed reductions in velocities and in amplitude of nerve conduction velocity in tibial and peroneal nerves. This implies that axonal nerve damages may be involved in SPG3A responsible HSP.