Patients and methods

Patient recruitment

A continuous series of 36 index cases with sporadic or recessive early onset (<25 years) spastic paraplegia was recruited. Patients presented with at least one of the following symptoms associated with SPG11 and SPG15: cognitive impairment, TCC, peripheral neuropathy or cerebellar ataxia.

Standardised examination included the Spastic Paraplegia Rating Scale and the inventory of complicating signs and symptoms.10 The study was approved by the local ethics committee. Written informed consent was obtained from all patients.

Genetic analysis

DNA was extracted from peripheral blood samples following standard protocols. Direct sequencing of the SPG15 gene ZFYVE26 (ENSG00000072121, spastizin) exon 2–42 was performed in all index patients. The SPG11 gene KIAA1840 (ENSG00000104133, spatacsin) was examined by high resolution melting (HRM) curve analysis. Exons 1–40 were amplified with an average fragment size of 300 bp. PCR and HRM were performed in a single run on a LightCycler 480 instrument (Roche Diagnostics, Mannheim, Germany). For each resulting HRM group, at least one sample was directly sequenced. Patient DNAs were spiked with 1/5 wild-type control DNA in order to identify heterozygous and homozygous sequence variants. See supplementary data for detailed experimental procedures and primer sequences (supplementary tables 1 and 2 available online).

Results

Thirty-six index patients were included in the study. In 17 cases, family history was positive and compatible with autosomal recessive inheritance (consanguineous parents in three patients), 18 cases were sporadic and in one case the family history was unknown. Complicating features were present at the following frequencies: cognitive impairment 61%, peripheral neuropathy 44%, cerebellar ataxia 67% and TCC 42%. Additional features included atrophy of intrinsic hand muscles, extrapyramidal signs and cataract (supplementary table 3 available online).

SPG11 mutation screening

SPG11 mutations were identified in a total of five patients (5/36, ∼14%). In four of these, only a single mutation was found. In one patient we identified a genomic deletion of exons 31–34 (described in Bauer and colleagues11). In the remaining three patients, all harbouring a truncating SPG11 mutation, no second disease causing mutation has been identified (see supplementary table 4 available online).

Clinical characteristics of SPG11 patients

Mean age at onset in the five SPG11 patients was 19 years (range 13–25). All patients showed a combination of mental retardation and TCC. Peripheral neuropathy (3/5), mild cerebellar ataxia (2/5) and hand muscle atrophy (3/5) were variably present (see supplementary table 3 available online).

SPG15 mutation analysis

A German patient (P35) with sporadic complicated HSP was found to be compound heterozygous for two novel SPG15 mutations (see supplementary fig available online). One was a nonsense mutation (c.592C>T, p.R198X) and the second a splice mutation due to loss of the splice donor site of intron 38 (c.7128+1G>C). This substitution reduces the predicted splicing efficiency from 0.93 to <0.1.12 Neither mutation was found in 360 control chromosomes.

Additionally, a total of seven unpublished sequence variants of unknown significance were identified, including five non-synonymous coding single nucleotide polymorphisms, one synonymous coding single nucleotide polymorphism and one intronic deletion in intron 12 (table 1).

In S615F variants, in silico analysis of putative phosphorylation sites (NetPhos 2.013) predicts loss of a highly probable serine phosphorylation site (NetPhos score 0.997); according to Polyphen,14 this variation is possibly damaging to protein function (ΔPSIC 1.998).

C1871Y present in two index patients is located within the zinc finger domain of ZFYFE26; this variant is predicted to affect protein function. However, tyrosine is the wild-type amino acid in the rat and mouse zfyve26 protein at the homologous position. S615F and C1871Y were observed with similar frequency in patients and controls.

Case report of patient P35 (c.592C>T, p.R198X/c.7128+1G>C, splice mutation)

This patient had a normal delivery and normal early postnatal milestones apart from a mild delay in speech development. At the age of 3 years he developed slurred articulation and mild stuttering. Clumsy hands and poor coordination became obvious when entering school. Progressive gait disturbance did not start before 16 years of age. He visited a school for mentally handicapped children but finally passed the secondary general school certificate (“Hauptschulabschluss”).

Examination at age 19 years revealed normal eye movements, mild dysarthria, spastic paraplegia with pareses of foot dorsal extensors and hip abductors on both sides (MRC grade 4/5). Lower limb tendon reflexes were brisk and plantar response was extensor. Repetitive hand movements were slightly slowed and clumsy but not ataxic or dysmetric. Gait was spastic and narrow based.

Electrophysiology confirmed that the corticospinal tract was affected with prolonged central motor conduction times to the legs. Additionally, somatosensory potentials revealed that the dorsal columns were affected whereas nerve conductions studies were normal. MRI showed generalised cortical atrophy sparing the cerebellum, white matter changes affecting predominantly the periventricular regions and centrum semiovale, and TCC most pronounced in genu and body (fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

MRI in SPG15, SPG11 and SPG4. (A) SPG15: T2 weighted image of a 19-year-old man with a 16 year history of complicated hereditary spastic paraplegias (HSP) (P35 in supplementary table 3 available online). Thinning of the corpus callosum is emphasised in the anterior parts (genu and body). (B) SPG15: transversal T2 flair image of patient P35 (see (A)) showing mild periventricular white matter lesions. (C) SPG11: T1 weighted image of a 39-year-old female who developed spastic paraplegia at the age of 16 years (P24 in supplementary table 3 available online). The MRI shows thinning of the corpus callosum most pronounced in the anterior parts (genu and body). (D) SPG4: T1 weighted image of a 52-year-old man with pure HSP since his 29th year of age. He carries a nonsense mutation in exon 15 of the spastin gene. Corpus callosum shows a normal configuration in this patient.

Discussion

Molecular diagnostic testing for SPG11 and SPG15 is time consuming and expensive because of the large gene size and involvement of genomic deletions not detectable by conventional sequence analysis, at least in SPG11.11 Careful patient selection is therefore essential and clear criteria for genetic testing are required:

Phenotypic differences between SPG11 and SPG15

TCC, cognitive impairment, peripheral neuropathy, including intrinsic hand muscle atrophy, and mild cerebellar signs are common clinical hallmarks in both SPG11 and SPG15. In SPG11, TCC is present in the vast majority of reported cases (>90%) and is the best single indicator for SPG11 in complex HSP.4 In SPG15, the reported frequency of TCC varies between 25% and 100% of cases.6 7 16 17

No apparent difference in the characteristics of the corpus callosum volume loss existed in our study. In SPG11 as well as in SPG15, volume loss was most pronounced in the genu and body of the corpus callosum that contain prefrontal, premotor, primary motor and primary sensory connections (fig 1).18 Additional MRI abnormalities such as white matter lesions and cortical atrophy have been reported in both, SPG11 and SPG15. No differences were established with regard to other complicating signs and symptoms such as macula pigmentation and peripheral neuropathy.4 5 6 7 16 19 Therefore, the phenotype does not permit discrimination between SPG11 and SPG15 in individual cases at present.

Frequency of SPG11 and SPG15

In order to provide an estimate of the frequency of SPG11 and SPG15 in complicated HSP, we selected a continuous series of early onset complicated HSP patients compatible with published SPG11 and SPG15 phenotypes. In this sample of predominantly German and Turkish descent, recruited in German HSP outpatient clinics, the overall frequency of SPG11 was about 14%. In the subgroup of patients with TCC, the frequency of SPG11 is considerably higher (5/12, ∼42%). Our frequency data are in concordance with published data.4 In the latter study, the consanguinity rate was 36% compared with 9% in our cohort. Interestingly, the lower consanguinity rate in our sample did not seem to influence frequency of SPG11. Similarly, no correlation between family history (autosomal recessive versus sporadic) and SPG11 was noted (p = 0.29).

SPG15 was positive in a single patient only (2.6%). This frequency is substantially lower than frequencies derived from mapping studies in mostly Arab populations with high consanguinity rates (25% in Boukhris and colleagues,7 15% in Elleuch and colleagues8). Larger studies are needed to specify these estimates.

“SPG11/SPG15-like” phenotype in other conditions

The SPG11/SPG15 phenotype is by no means unique to these two HSP subtypes. Supplementary table 5 (available online) gives an overview of conditions combining spastic paraplegia and TCC.

In conclusion, we found SPG15 to be a rare cause of autosomal recessive complicated spastic paraplegia. Best predictor for SPG15 as well as SPG11 is the presence of TCC in complicated HSP with sporadic or autosomal recessive disease. No phenotypic criteria could be identified to predict the genotype and differentiate clinically between SGP11 and SPG15. Based on the substantially higher frequency of SPG11 compared with SPG15, we suggest commencing with SPG11 testing and to restrict analysis of SPG15 to SPG11 negative cases.