Hereditary Spastic Paraplegia With Mental Impairment and Thin Corpus Callosum in Tunisia: Title and subTitle BreakSPG11, SPG15, and Further Genetic Heterogeneity FREE

Hereditary spastic paraplegia (HSP) is a clinically and genetically heterogeneous group of neurodegenerative disorders characterized by slowly progressive spasticity of the lower extremities. In addition to pure forms, complicated forms involving additional neurological features, such as mental retardation, ataxia, peripheral neuropathy, retinopathy, optic atrophy, deafness, and ichthyosis, have also been reported.^{1 - 2} Pathologically, HSP is characterized by axonal degeneration in the long descending and ascending tracts of the spinal cord, especially in their terminal portions.³ The corticospinal tracts are mainly affected, but the Goll and spinocerebellar tracts may also be involved.

Autosomal dominant, autosomal recessive (AR), and X-linked forms of HSP have been identified.⁴ To date, more than 33 HSP loci and 15 spastic paraplegia genes have been identified.^{4 - 7} The corresponding proteins are often involved in axonal trafficking or mitochondrial metabolism.⁸

Hereditary spastic paraplegia with mental impairment and thin corpus callosum (HSP-TCC) is a frequent subtype of complicated HSP, often inherited as an AR trait. A frequency of 35% among recessive forms has been reported in Brazil.⁹ Mental deterioration usually starts during the first or second decade of life, associated with various signs of complicated HSP.¹⁰

The purpose of this study was to establish, through clinical and paraclinical investigations of a large series of patients with AR HSP, the relative frequency of complex forms associating TCC and mental impairment (HSP-TCC) and its molecular bases in Tunisia. We report 8 Tunisian AR HSP-TCC families and confirm the phenotypic and genetic heterogeneity of this particular form of HSP.

During the last 15 years, we assessed 73 affected patients from 33 “apparently” unrelated Tunisian families of Arab origin in the Department of Neurology of the University Hospital of Sfax, the reference center in neurological disorders in South Tunisia. After obtaining informed consent, all available affected and apparently unaffected family members were assessed neurologically. Age at symptom onset was obtained from the parents or the available medical records. Disability was assessed on a 7-point scale in which 1 indicates minimal disability (slight stiffness of the legs); 2, mild disability (unable to run, but full autonomy); 3, moderate disability in walking (reduced perimeter, frequent falls); 4, severe disability (unilateral assistance required to walk); 5, bilateral assistance required to walk; 6, wheelchair bound; and 7, bedridden. Mental impairment was assessed by the Mini-Mental State Examination (MMSE) and was considered as mild (MMSE score, 21-26), moderate (MMSE score, 16-20), or severe (MMSE score, < 15). Thin corpus callosum was assessed by brain magnetic resonance imaging (MRI) (1.5 T) by experienced neurologists by comparison with normal MRI findings after a mean (SD) disease duration of 17.8 (9.1) years (range, 3-39 years).

Only families meeting the following criteria were diagnosed as having AR HSP-TCC: (1) inheritance consistent with an AR trait, (2) slowly progressive spastic paraparesis and mental impairment in at least 1 affected patient, (3) thinning of the corpus callosum as revealed by brain MRI in at least 1 member of the family, and (4) exclusion of other disorders by brain and spinal MRI and other laboratory tests.¹¹

DNA was extracted from blood using a standard protocol. According to the clinical presentation and transmission mode, we tested the selected families for linkage to several loci previously described for AR HSP frequently or rarely associated with mental impairment and TCC using appropriate microsatellite markers (list available on request): SPG7, SPG11, SPG15, SPG21/MAST, SPG32, and HSP-TCC epilepsy.^{12 - 17} Genotypes were determined using standard methods, and linkage analysis was performed using Allegro (deCODE Genetics, Reykjavik, Iceland) assuming a fully penetrant recessive disease with similar male-female recombination frequencies and equal allele frequencies. Gene frequency values of 0.0005 or 0.002 were used and did not affect the logarithm of odds (LOD) score results. Young unaffected individuals (younger than the mean age at onset of the affected patients in the family) were considered to have unknown status for LOD score calculations. Haplotypes segregating at each locus were reconstructed manually by minimizing the number of recombination events. The index patient from each family with putative linkage to SPG11 was screened for mutations in all coding exons and splice junctions of the KIAA1840 gene, as described previously.⁶

Among the 33 families with AR HSP that could be assessed, we identified 8 that fulfilled the criteria for HSP-TCC (24%). Consanguinity was present in 7 families. Two cases were apparently sporadic but both were consanguineous. Onset in 19 patients (10 men, 9 women) from the 8 AR HSP-TCC families occurred at mean (SD) 10.6 (5.1) years (range, 2-16 years). In all patients, the presenting symptoms were gait abnormalities due to the insidious appearance of stiffness and weakness in the lower extremities (Table 1).

After a mean (SD) disease duration of 18.9 (7.8) years (range, 6-33 years), cognitive decline was present in all affected individuals, except one who had only 7 years of evolution, in addition to lower limb hyperreflexia and bilateral Babinski signs. Others signs related to pyramidal tract dysfunction were found in several patients: pseudobulbar dysarthria, upper limb spasticity, and bladder dysfunction. Additional signs were occasionally observed, such as cerebellar ataxia (7 of 19), lower limb distal amyotrophy (5 of 19), cataract (5 of 19), and decreased vibration sense (2 of 19).

Magnetic resonance imaging was performed at the mean (SD) age of 26.5 (5.4) years (range, 20-31 years) and after a mean (SD) disease duration of 16.3 (6.4) years (range, 6-26 years). Consistent with the clinical findings, all of the 9 patients studied by brain MRI had TCC, most prominently in the rostrum, genu, and body, associated with cerebral and cerebellar atrophy (Table 2). White matter changes were found frequently (6 of 9 patients [66%]). In 3 individuals, hyperintense T2-weighted lesions were restricted to occipital and frontal regions, but in the other cases, there were diffuse abnormalities involving periventricular regions (Figure 1). Cortical cerebral atrophy with frontal predominance seemed to be a late feature, since it was absent in patient TUN 17-168, who had the shortest disease duration when examined (6 years). Electroneuromyograms showed electrophysiological signs of predominantly axonal motor peripheral neuropathy in 4 of 8 patients tested (Table 2). Muscle and nerve biopsy specimens of these 4 patients showed signs of axonal degeneration associated with chronic neurogenic changes without specific alterations.

We first evaluated linkage to known genes (SPG7, SPG11, and SPG21) and subsequently looked for mutations in linked families. Further explorations in nonmutated and nonlinked families included loci SPG15 and SPG32 and the locus for HSP-TCC epilepsy.

In 5 families (62.5%), linkage to SPG11 was suggested by identical homozygosity at several markers and positive multipoint LOD score values from 1.1 to 2.5 reaching the maximal expected values in the pedigrees (Table 3). Patients in families TUN 3, TUN 4, and TUN 22 had partially similar haplotypes encompassing the SPG11 gene, suggesting inheritance of a common mutated ancestral chromosome. Similarly, families TUN 9 and TUN 12 shared the same homozygous alleles for the whole SPG11 interval. In accordance with these results, direct sequencing of all exons and splice junctions of the SPG11 gene in the 5 chromosome 15q–linked families revealed 2 recurrent mutations: c.733_734delAT (exon 4)/p.M245VfsX in families TUN 3, TUN 4, and TUN 22 and c.6100C>T (exon 32)/p.R2034X in families TUN 9 and TUN12 (Figure 2). The mutations segregated with the disease in all families but were also present in the homozygous state in individual TUN 3-42, who was still unaffected at 11 years of age. This individual only had brisk reflexes on examination and is still younger than the age at onset of his affected relatives. Because of his young age, no further investigations could be performed.

In all 3 non-SPG11 families, haplotype reconstructions excluded linkage to SPG7, SPG21, SPG32, and the locus for HSP-TCC epilepsy because there were no haplotypes identical by descent in the affected patients of each kindred and because the multipoint LOD scores for the candidate intervals were negative (Table 3).

The exclusion of other known HSP-TCC loci and the cosegregation with the disease of chromosome 14q haplotypes between markers D14S1038 and D14S270 in 2 families (TUN 8 and TUN 17, 25%) suggested their linkage to SPG15 (Figure 3A). Homozygosity was observed in the whole interval in family TUN 17 and between markers D14S1012 and D14S1002 in family TUN 8. In addition, linkage to SPG15 was supported by positive LOD scores of 2.0 and 1.8 in families TUN 8 and TUN 17, respectively. The haplotypes segregating with the disease were different in the 2 families, suggesting the possibility of different allelic mutations.

In the remaining family (TUN 35) (Figure 3B), linkage to the 6 known loci of HSP-TCC was excluded (Table 3), suggesting the existence of another genetic locus for this condition.

The mean age at onset (12.4 vs 14.3 years) and the clinical profiles of patients with SPG11 (n = 11) and SPG15 (n = 3) were similar, but there were a few subtle differences between them. SPG11 patients tended to have a more severe disease than SPG15 patients (mean severity score, 5.18 vs 4.33), but this was correlated with the disease duration (mean disease duration, 16.54 vs 13 years). Muscle atrophy or cerebellar signs were never observed in SPG15 patients. Indeed, electrophysiological analysis of the 2 SPG15 patients studied showed no evidence of peripheral neuropathy, and radiological analysis showed less cerebellar atrophy and slighter white matter lesions that were restricted to the frontal and occipital horns. Not surprisingly, the phenotypes of SPG11 patients with the M245fsX or the R2034X mutations were undistinguishable.

The phenotype of family TUN 35 was clearly different. Although the age at onset was earlier (mean [SD], 4.4 [3.5] years), the cognitive decline was mild, progression was less severe (mean severity score, 3), and patients were still able to walk after a mean disease duration of 27.6 years. In addition, all patients had cataracts and cerebellar ataxia, and pes cavus was frequent (4 of 5). Patient TUN 35-27/07 had facial dyskinesia without other extrapyramidal signs, associated with mild hand tremor. Electroneuromyography findings were normal in a single patient. Cerebellar atrophy was more severe on brain MRI and white matter abnormalities were unremarkable.

We report the clinical and genetic study of 8 Tunisian families with HSP-TCC, including 19 patients, the largest group of North African patients of Arab origin investigated to our knowledge so far. Hereditary spastic paraplegia with TCC and mental impairment was common in our AR HSP families (8 of 33 or 24% of all families) but less than in a Brazilian cohort (35%).⁹

Autosomal recessive HSP-TCC is a subtype of complicated HSP that is clinically and genetically heterogeneous. This condition was originally described in Japanese patients and has been found in many countries.^{6 ,9 ,13 ,18 - 20} Autosomal recessive HSP-TCC is characterized clinically by slowly progressive spastic paraparesis and mental deterioration that begins mainly in the second decade of life.²¹ Cerebellar ataxia and impaired vibration sense are also observed in some patients. Additional manifestations include spasticity and hyperreflexia in the upper limbs, distal amyotrophy, pes cavus, urinary disturbance, dysarthria, nystagmus, congenital cataracts, seizures, and extrapyramidal signs.^{10 - 11 ,21} On brain MRI, progressive thinning of the corpus callosum is the neuroradiologically distinctive feature of this syndrome.⁹ Periventricular white matter changes and late cortical atrophy are additional and frequent features. Cerebral and cerebellar atrophy are slowly progressive, and abnormalities of cerebral white matter with frontal predominance have been observed in long-standing cases.²²

It is clear from molecular genetic analyses that there are several underlying causes of this syndrome,²³ with 5 genetic loci identified (SPG11, SPG15, SPG21, SPG32, and HSP-TCC epilepsy). SPG7 has also been associated with a similar phenotype. The major locus is, however, SPG11 (62% in Tunisia), a frequency similar to Japanese (11 of 13 [77%]), Mediterranean (6 of 10 [60%]), and Italian (5 of 12 [41%]) families.^{11 ,13 ,24 - 25} Results of linkage analysis were confirmed by the subsequent identification of 2 recurrent mutations, R2034X and M245VfsX, in 2 and 3 Tunisian families, respectively, suggesting founder effects.⁶ R2034X-segregating haplotypes were similar in a previously published Moroccan kindred with the same mutation, suggesting inheritance of a common ancestral mutation.⁶ On the contrary, the segregating haplotypes in French and Italian kindreds carrying the M245VfsX mutation were different, indicating independent ancestral events or a very ancient mutation.^{6 ,26} SPG15, responsible for 25% of AR HSP-TCC in our series, was the second most important locus, in agreement with a frequency of 15% reported in another study of families from the Mediterranean Basin.²⁷ The similarity between the neurological and radiological findings in SPG11 and SPG15 patients suggests that the responsible genes may be functionally related. Although a few subtle differences were observed, such as the absence of lower limb/upper limb amyotrophy, peripheral neuropathy, or cerebellar signs in some SPG15 patients, who also had severe lesions of the cerebellum, these features may be present in other SPG15 patients²⁷ and cannot be used to distinguish SPG11- from SPG15-linked families on a clinical basis.

One family in our series was not linked to any of the loci tested, suggesting that an as-yet-unmapped locus is responsible for their disease. Interestingly, this family, which represents a new genetic entity, also had a unique phenotype that differed from that of SPG11 and SPG15 by an earlier onset and a slower progression of spastic gait, as well as the presence of cataracts and cerebellar ataxia.

In conclusion, we describe a group of Arab families from North Africa with AR HSP-TCC. We demonstrate that this clinical radiological syndrome is a frequent form of HSP and that SPG11 and SPG15 are the major loci for this clinical entity. In addition, we show the existence of a novel genetic form of AR HSP-TCC with unique and different clinical features.

Correspondence: Giovanni Stevanin, PhD, INSERM U679, Groupe Hospitalier Pitié-Salpêtrière, 47 boulevard de l’Hôpital, 75013 Paris, France (stevanin@ccr.jussieu.fr).

Accepted for Publication: December 9, 2007.

Author Contributions:Study concept and design: Boukhris, Stevanin, Belal, Mhiri, and Brice. Acquisition of data: Boukhris, Stevanin, Feki, Denis, Elleuch, Miladi, Truchetto, Denora, Belal, and Mhiri. Analysis and interpretation of data: Boukhris, Stevanin, Feki, Elleuch, Truchetto, Mhiri, and Brice. Drafting of the manuscript: Boukhris, Stevanin, Feki, Truchetto, and Mhiri. Critical revision of the manuscript for important intellectual content: Boukhris, Stevanin, Feki, Denis, Elleuch, Miladi, Denora, Belal, Mhiri, and Brice. Statistical analysis: Stevanin and Mhiri. Obtained funding: Stevanin, Mhiri, and Brice. Administrative, technical, and material support: Boukhris, Feki, Mhiri, and Brice. Study supervision: Stevanin, Belal, Mhiri, and Brice.

Financial Disclosure: None reported.

Funding/Support: This work was financially supported by the French Tunisian cooperation project (Drs Brice and Mhiri) led by INSERM (France) and Direction Générale de la Recherche Scientifique et de la Rénovation Technologique (DGRSRT) (Tunisia), the VERUM Foundation (Dr Brice), the GIS (Groupement d’Intérêt Scientifique)–Maladies Rares (Dr Stevanin), and the French Agency for Neuroscience Research (to the SPATAX Network and Dr Stevanin). Dr Boukhris received a fellowship from the French Association Strümpell–Lorrain.

Additional Contributions: Merle Ruberg, PhD, and Cyril Goizet, MD, PhD, critically read the manuscript and Chahnez Triki, MD, and Fatma Kammoun, MD, performed clinical examinations. We thank the family members for their participation. We also thank Sylvie Forlani, PhD, and Sylvain Hanein, PhD, for their kind help.