Genetic and phenotypic characterization of complex hereditary spastic paraplegia

Abstract

The hereditary spastic paraplegias are a heterogeneous group of degenerative disorders that are clinically classified as either pure with predominant lower limb spasticity, or complex where spastic paraplegia is complicated with additional neurological features, and are inherited in autosomal dominant, autosomal recessive or X-linked patterns. Genetic defects have been identified in over 40 different genes, with more than 70 loci in total. Complex recessive spastic paraplegias have in the past been frequently associated with mutations in SPG11 (spatacsin), ZFYVE26/SPG15, SPG7 (paraplegin) and a handful of other rare genes, but many cases remain genetically undefined. The overlap with other neurodegenerative disorders has been implied in a small number of reports, but not in larger disease series. This deficiency has been largely due to the lack of suitable high throughput techniques to investigate the genetic basis of disease, but the recent availability of next generation sequencing can facilitate the identification of disease-causing mutations even in extremely heterogeneous disorders. We investigated a series of 97 index cases with complex spastic paraplegia referred to a tertiary referral neurology centre in London for diagnosis or management. The mean age of onset was 16 years (range 3 to 39). The SPG11 gene was first analysed, revealing homozygous or compound heterozygous mutations in 30/97 (30.9%) of probands, the largest SPG11 series reported to date, and by far the most common cause of complex spastic paraplegia in the UK, with severe and progressive clinical features and other neurological manifestations, linked with magnetic resonance imaging defects. Given the high frequency of SPG11 mutations, we studied the autophagic response to starvation in eight affected SPG11 cases and control fibroblast cell lines, but in our restricted study we did not observe correlations between disease status and autophagic or lysosomal markers. In the remaining cases, next generation sequencing was carried out revealing variants in a number of other known complex spastic paraplegia genes, including five in SPG7 (5/97), four in FA2H (also known as SPG35) (4/97) and two in ZFYVE26/SPG15 Variants were identified in genes usually associated with pure spastic paraplegia and also in the Parkinson's disease-associated gene ATP13A2, neuronal ceroid lipofuscinosis gene TPP1 and the hereditary motor and sensory neuropathy DNMT1 gene, highlighting the genetic heterogeneity of spastic paraplegia. No plausible genetic cause was identified in 51% of probands, likely indicating the existence of as yet unidentified genes.

Keywords: Parkinson’s disease; SPG11; gene; hereditary spastic paraplegia; mutation.

High-throughput next-generation sequencing can identify disease-causing mutations in extremely heterogeneous disorders. Kara et al . investigate a series of 97 index cases with complex hereditary spastic paraplegia (HSP). They identify SPG11 defects in 30 families, as well as mutations in other HSP genes and genes associated with disorders including Parkinson’s disease.

Figure 1

Overview of mutations identified in spastic paraplegia genes. ( A ) Pie chart showing the frequency of mutations in spastic paraplegia genes identified. The figures in brackets represent the number and percentage of cases respectively. ( B ) Frequencies of clinical features that were present in addition to the spastic paraplegia in SPG11 probands. ( C ) Frequencies of clinical features that were present in addition to spastic paraplegia in the spastic paraplegia patients with other mutations. In B and C the figures in brackets refer to the number of cases. ( D ) Diagram of the SPG11 gene with mutations identified in the present study. Grey arrows indicate novel and black arrows indicate previously reported mutations.

Figure 2

MRI features in patients with complex HSP. ( A ) Sagittal MRI of SPG11 patients showing progressive thinning of the corpus callosum and cerebral atrophy, which correlates with the progression of clinical features. ( i ) MRI from a healthy individual with labelling of the different parts of the corpus callosum. ( ii ) Case 5 at age 18 (mild disease). ( iii ) Case 8 at age 22 (mild-moderate disease). ( iv ) Case 17 at age 23 (moderate disease). ( v ) Case 10 at age 30 (severe disease). ( vi ) Case 16 age 32 (severe disease). ( vii ) Case 9 age 39 (severe disease). See Table 1 for details of case numbers. ( B ) MRI of complex HSP cases. ( i ) Case 37 (age 24 years), SPG7 showing an axial MRI with high signal in the cerebral peduncles (arrow) and on coronal imaging ( ii ) and sagittal imaging (arrow) ( iii ) with thinning of the body of the corpus callosum (arrow). Case 33 (age 34 years) ( iv to vi ), SPG15 with thinning of the corpus callosum ( iv ) (arrow) and generalized atrophy with periventicular white matter abnormalities (arrow). ( C ) Sagittal MRI of FA2H patients. ( i–iii ) Case 32, age 32 years, slowly progressive with thinning of the corpus callosum, cerebellar and cortical atrophy. ( iv–vi ) Case 52, age 37 years, more rapid and severely affected: shows severe corpus callosum thinning, cerebellar and cerebral atrophy, but preserved white matter, similar to the Case 32.

Figure 3

Photographs of the hands of Case 33 with ZFYVE26/SPG15 mutation (homozygous p.R1378* mutation). Showing the adducted thumbs that are similar to those seen in MASA (mental retardation, aphasia, shuffling gait and adducted thumbs) syndrome. Age of patient: 34 years.

Figure 4

Morphological findings of one nerve and four muscle biopsies in five genetically characterized patients with HSP. Patient with ZFYVE26/SPG15 mutation ( A and C ) and control individual ( B and D ). SPG11 mutation ( E ), SPG7 mutation ( F and G ). Semi-thin resin preparations stained with methylene blue azure—basic fuchsin ( A and B ) show a reduction of large myelinated fibre density in the patient’s biopsy, compared with a biopsy from age-matched control. Large fibre loss is further confirmed by electron microscopy ( C and D ), while unmyelinated fibres are better preserved. There is no evidence of active axonal degeneration, regeneration or demyelination and the overall picture is that of chronic axonal neuropathy. The muscle biopsies from three patients investigated for signs of denervation show varied appearances. In the biopsy from the patient with a known SPG11 mutation ( E ) there is predominance of type 1 fibres (yellow arrow, ATP pH4.3). In one patient with SPG7 mutation ( F ) the most significant finding in the muscle biopsy is that of several COX-deficient fibres (red arrows) seen on combined COX-SDH preparation. In another patient with SPG7 mutation ( G ) the biopsy confirms neurogenic change with evidence of previous denervation with re-innervation (blue arrow indicates a group of type 1 fibres, ATP pH4.3). Scale bars: A , B and E–H = 40 μm; C and D = 5 μm.

Figure 5

Reverse transcription PCR to assess SPG11 mRNA expression in fibroblasts. Wide expression is seen across SPG11 affected and control fibroblasts. Numbers in red are SPG11 affected cases from Table 1 and Supplementary Table 1 . Letters in blue are controls from Supplementary Table 1 . GAPDH = housekeeping gene glyceraldehyde 3-phosphate dehydrogenase.

Figure 6

Analysis of markers for autophagy and lysosomal function in human fibroblast cells. ( A ) Fibroblast protein expression levels of LAMP1 and LC3 as compared to beta actin in SPG11 affected and control fibroblasts after starvation induced autophagy (overnight serum starvation, followed by 2.5 h amino acid starvation in low glucose). ( B ) Fibroblast protein expression levels of p62 and HSP70 as compared to beta actin in SPG11 affected and control fibroblasts after starvation induced autophagy (overnight serum starvation, followed by 2.5 h amino acid starvation in low glucose).