Abstract

Hereditary spastic paraplegias are a heterogeneous group of neurodegenerative disorders, clinically classified in pure and complex forms. Genetically, more than 70 different forms of spastic paraplegias have been characterized. A subgroup of complicate recessive forms has been distinguished for the presence of thin corpus callosum and white matter lesions at brain imaging. This group includes several genetic entities, but most of the cases are caused by mutations in the KIAA1840 (SPG11) and ZFYVE26 genes (SPG15). We studied a cohort of 61 consecutive patients with complicated spastic paraplegias, presenting at least one of the following features: mental retardation, thin corpus callosum and/or white matter lesions. DNA samples were screened for mutations in the SPG11/KIAA1840, SPG15/ZFYVE26, SPG21/ACP33, SPG35/FA2H, SPG48/AP5Z1 and SPG54/DDHD2 genes by direct sequencing. Sequence variants were found in 30 of 61 cases: 16 patients carried SPG11/KIAA1840 gene variants (26.2%), nine patients carried SPG15/ZFYVE26 variants (14.8%), three patients SPG35/FA2H (5%), and two patients carried SPG48/AP5Z1 gene variants (3%). Mean age at onset was similar in patients with SPG11 and with SPG15 (range 11–36), and the phenotype was mostly indistinguishable. Extrapyramidal signs were observed only in patients with SPG15, and epilepsy in three subjects with SPG11. Motor axonal neuropathy was found in 60% of cases with SPG11 and 70% of cases with SPG15. Subjects with SPG35 had intellectual impairment, spastic paraplegia, thin corpus callosum, white matter hyperintensities, and cerebellar atrophy. Two families had a late-onset presentation, and none had signs of brain iron accumulation. The patients with SPG48 were a 5-year-old child, homozygous for a missense SPG48/AP5Z1 variant, and a 51-year-old female, carrying two different nonsense variants. Both patients had intellectual deficits, thin corpus callosum and white matter lesions. None of the cases in our cohort carried mutations in the SPG21/ACP33 and SPG54/DDH2H genes. Our study confirms that the phenotype of patients with SPG11 and with SPG15 is homogeneous, whereas cases with SPG35 and with SPG48 cases present overlapping features, and a broader clinical spectrum. The large group of non-diagnosed subjects (51%) suggests further genetic heterogeneity. The observation of common clinical features in association with defects in different causative genes, suggest a general vulnerability of the corticospinal tract axons to a wide spectrum of cellular alterations.

Introduction

Hereditary spastic paraplegias (SPGs) are clinically and genetically heterogeneous diseases, classified in ‘pure’ and ‘complicated’ forms. Pure hereditary SPGs present progressive spasticity and weakness of the lower limbs often associated with mild sensory abnormalities, urinary dysfunctions, and pes cavus (Harding, 1983). Complicated forms encompass the previous clinical features associated with a large variety of additional neurological and non-neurological manifestations including cerebellar signs, amyotrophy, peripheral neuropathy, optic atrophy, cognitive decline, deafness, retinopathy or cataract (Harding, 1983; Finsterer et al., 2012; Fink, 2013). The age at onset is extremely variable and both early and late-onset conditions are observed. Genetic classification includes the type of transmission (autosomal dominant, autosomal recessive and X-linked), and the number of locus/gene identified. To date, >70 spastic paraplegia loci have been classified (SPG1–71), and for ∼50 forms the causative genes have also been identified (Schüle and Schöls, 2011; Finsterer et al., 2012; Fink, 2013; Novarino et al., 2014).

Autosomal dominant paraplegias are mainly characterized by pure phenotypes, and ∼40–50% of cases are caused by mutations in the SPAST gene (SPG4). Other common autosomal dominant paraplegias are SPG3A/ATL1 (10% of cases) and SPG31/REEP1 (5%) (Finsterer et al., 2012). The autosomal recessive hereditary SPGs are often associated with complex phenotypes, and one of the most frequent forms is caused by mutations in the SPG11/KIAA1840 gene (Stevanin et al., 2007; Pippucci et al., 2009; Schüle and Schöls, 2011; Finsterer et al., 2012).

The phenotype of patients with SPG11 is characterized by an early onset progressive spastic paraparesis, mental deterioration, peripheral nerve involvement, and by the presence, at neuroradiological examination, of extremely thin corpus callosum and periventricular white matter abnormalities. These clinical and radiological features, initially recognized in Japanese families (Nakamura et al., 1995), were supposed to be distinguishing for this genetic subtype, and the newly identified protein, encoded by the KIAA1840 gene, was called ‘spatacsin’ (SPAsticity with Thin or Atrophied Corpus callosum SYNdrome; Stevanin et al., 2007). Subsequent genetic studies showed that this phenotype was associated with a great genetic heterogeneity, and several other autosomal recessive SPG loci and genes were identified. The second most common genotype associated with autosomal recessive hereditary SPG with thin corpus callosum and white matter lesions is SPG15, caused by mutations in ZFYVE26 gene encoding spastizin (Hanein et al., 2008). Several patients with SPG15, from different ethnic backgrounds, have been described, and their clinical presentation may include pigmentary maculopathy, cerebellar signs and dystal amyotrophy (Boukhris et al., 2009; Schüle et al., 2009).

Both thin corpus callosum and white matter abnormalities have also been observed in a few patients with SPG21/ACP33 (Simpson et al., 2003), SPG47/AP4B1 (Abou Jamra et al., 2011) and SPG54/DDHD2 gene mutations (Schuurs-Hoeijmakers et al., 2012). Complex hereditary SPGs with white matter lesions have been described in SPG35/FA2H (Dick et al., 2010; Pierson et al., 2012), and SPG48/AP5Z1 genetic subtypes (Słabicki et al., 2010; Finsterer et al., 2012; Fink, 2013).

The variability in clinical phenotype and the vast genetic heterogeneity make it difficult to perform an efficient diagnostic procedure for this group of hereditary SPGs or to determine ‘a priori’ which gene is likely to be mutated in a given family or patient. For most of the recently identified genotypes, only limited clinical and molecular descriptions are available, and their frequencies among the group of the complicated autosomal recessive hereditary SPG forms still need to be precisely determined. In this study we performed an extensive molecular screening in six of the genes causing complicated autosomal recessive SPGs with thin corpus callosum and/or white matter lesions abnormalities (SPG11, SPG15, SPG21, SPG35, SPG48 and SPG54). Our aims were to diagnose a large cohort of patients with hereditary SPG, and to compare molecular findings, clinical characteristics, and relative frequencies of these hereditary conditions.

Materials and methods

Patients

We studied a cohort of 61 consecutive unrelated patients with complicated hereditary SPG referred to our centre for clinical evaluation and genetic testing. Inclusion criteria were progressive spastic paraplegia and at least one of the following clinical and radiological features: cognitive impairment, thin corpus callosum, and white matter lesions. Subjects with autosomal dominant family history were excluded. Age at onset was based on the first motor symptom, such as leg stiffness, gait unsteadiness or frequent falls. Patients underwent neurological examination, and standard 1.5 T MRI brain imaging. The subjects with SPG35 underwent a second structural imaging brain MRI, including the following sequences: 3D sagittal T1 turbo field echo, axial T2 turbo spin echo, axial and coronal FLAIR, axial T2 fast field echo (yielding T2*-weighted images). Neurophysiological assessments included electroneuronography, EMG and multimodal evoked potentials. Detailed neurophysiological methods are described in the Supplementary material.

Molecular analyses

Genomic DNA was extracted from venous peripheral blood lymphocytes, according to a standard phenol-chloroform procedure. Written informed consent for DNA analyses were obtained from all patients and family members. DNA samples from all the patients were initially screened for mutations in the SPG11/KIAA1840 (exons 1–40), and SPG15/ZFYVE26 (exons 1–42) genes. The negative cases were further screened for mutations in the SPG21/ACP33 (exons 1–8), SPG35/FA2H (exons 1–7), SPG48/AP5Z1 (exons 1–17) and SPG54/DDHD2 (exons 1–18) genes (Fig. 1). Exons and intron–exon boundaries were analysed by direct sequence analysis using an automated sequencing system (ABI 3130 XL). The primers are available on request. Patients carrying heterozygous sequence variants were also screened for exon deletion or duplication by multiplex ligation-dependent probe amplification (MLPA, for SPG11/KIAA1840 gene), or by real-time PCR (for SPG35/FA2H and SPG48/AP5Z1 genes). MLPA was performed using the SALSA MLPA kit P306-A1 1 (MRC). Real-time PCR was performed on 7300-Real-Time PCR System Instrument (Applied Biosystems), with Fast-Start Universal SYBR Green Master (ROX, Roche).

Figure 1

Flow diagram of the study phases, including selection of the probands and molecular analyses of six SPG genes associated with recessive complex hereditary spastic paraplegia.

Figure 1

Flow diagram of the study phases, including selection of the probands and molecular analyses of six SPG genes associated with recessive complex hereditary spastic paraplegia.

Nucleotides and amino acid residues were numbered according to the reference gene sequence of the transcript GenBank (NCBI): SPG11 (Homo sapiens NM_025137.3); SPG15 (NM_015346.3); SPG21 (NM_016630.3); SPG35 (NM_024306.4); SPG48 (NM_014855.2); and SPG54 (NM_001164232.1). Sequence variations and predicted protein changes were described according to nomenclature recommendations (den Dunnen and Antonarakis, 2000). Segregation of genetic variants and compound heterozygosity or homozygosity was demonstrated through validation in all available family members. Frequencies of novel missense variants were determined by direct sequencing in at least 228 chromosomes from unrelated healthy Italian controls. In silico analyses of missense variants were performed using PolyPhen-2 [Human Diversity (HumDiv) and Human Variation (HumVar) prediction models, http://genetics.bwh.harvard.edu/pph2/], and SIFT prediction test (http://sift.jcvi.org/). The Human Gene Mutation Database (http://www.hgmd.cf.ac.uk/ac/index.php), the NCBI dbSNP132ver (http://www.ncbi.nlm.nih.gov/projects/SNP/) and the Exome Variant Server of the National Heart, Lung and Blood Institute Exome Sequencing Project (http://evs.gs.washington.edu/EVS/) were examined for single nucleotide polymorphisms (SNPs) and minor allele frequency (MAF).

Results

Sixty-one patients were included in the study (58 Italian, two from Morocco, and one from Chile). The majority presented mental retardation or cognitive deterioration (89%); 80% had brain white matter alterations; 79% thin corpus callosum; and 42% peripheral neuropathy. The mean age at examination was 27 ± 11 years (range 5–57 years). Forty-nine were sporadic cases, and 12 had a family history compatible with autosomal recessive inheritance.

A final genetic diagnosis was established in 30 index cases (49%): 16 patients carried SPG11/KIAA1840 gene variants, nine patients SPG15/ZFYVE26, three patients SPG35/FA2H, and two patients SPG48/AP5Z1 gene variants (Fig. 1). None of the examined cases carried mutations in the SPG21/ACP33 and in the SPG54/DDH2H genes. Genetic variants identified in this study are described in Tables 1–3. Principal clinical features are summarized in Table 4.

Table 1

SPG11/KIAA1840 genetic variants identified in 16 Italian patients

SPG11 cDNA Change SPG11 protein change Predicted effect Ex (intron) Patient code status# Comments, family segregation, references§ 
c.258-2A>C IVS1-2A>C Splice-site (intron 1) Patient 803 ComHet Compound heterozygosity for IVS1-2A>C (Schule, 2009), and p.Asn169Ile, demonstrated heterozygous in both parents. One unaffected brother did not carry any of the variants (Supplementary Fig. 1A). 
c.130delC p.Arg44Glyfs*13 Frameshift exon 1 Patient 1146 ComHet Compound heterozygosity for p.Arg44Glyfs*13 and p.Arg945Glyfs*5, demonstrated heterozygous in both parents. One affected brother not tested. 
c.165del19nc. p.Ser56Alafs*6 Frameshift exon 1 Patient 965 ComHet Compound heterozygosity for p.Ser56Alafs* and IVS17-1G>C demonstrated heterozygous in both parents. One unaffected brother heterozygous for IVS17-1G>C. 
c.434insC p.Gln145Profs*17 Frameshift exon 2 Patient 1253 Homoz Consanguineous parents, both tested heterozygous. Two healthy siblings (4 and 26 years) not tested. 
c.506A>T p.Asn169Ile Missense exon 3 Patient 803 ComHet See comment above for the same patient. 
c.662A>G p.Trp221* Nonsense exon 3 Patient 1518 ComHet Patient 846 ComHet Patient1518: compound heterozygosity for p.Trp221* (demonstrated in the father) and p.Gln1436Argfs*7 (demonstrated in the mother). 
Patient846: compound heterozygosity for p.Trp221* (demonstrated in the father) and IVS39-2A>G (demonstrated in the mother). No other siblings. 
The fathers of these two patients, heterozygous for p.Trp221*, did not share common intragenic polymorphisms. 
c.733_734delAT p.Met245Valfs*2 Frameshift exon 4 Patient 906 Homoz Consanguineous parents originating from Sicily, both tested heterozygous. Previously described in different subjects (Del Bo, 2007; Hehr et al.,2007; Stevanin et al., 2007, 2008; Boukhris et al., 2008; Liao et al., 2008; Denora et al., 2009a). In Italian patients this mutation is associated with p.Val2053Met variant (see text). 
c.1203delA p.Asp402Ilefs*13 Frameshift exon 6 Patient 1288 ComHet Previosuly described in Italian patients (Stevanin et al., 2007; Denora et al., 2009a). 
Patient1288: compound heterozygosity for p.Asp402Ilefs*13* (heterozygous mother and one healthy brother) and p.Leu950Phefs*3 (father and one healthy brother). 
Patient 347 ComHet Patient347: compound heterozygosity for p.Asp402Ilefs*13 (demonstrated in the father) and p.Val948Glyfs*5 (mother not tested). 
Haplotype analysis was not informative. Not known relations between families, but same geographical origin (South of Italy). 
c.1492C>T p.Gln498* Nonsense exon 7 Patient 1105 ComHet Compound heterozygosity for p.Gln498* (demonstrated in the mother) and p.Leu2271Aspfs*68 (in the father). One healthy sister, non-carrier. 
c.1951C>T p.Arg651* Nonsense exon 10 Patient 1804 Homoz Recurrent mutation (p.R651X) (Hehr et al., 2007; Stevanin et al., 2008; Denora et al., 2009a). The patients did not share common haplotypes (see also text). 
Patient 2055 Homoz  
Patient 1465 ComHet  
c.2277delC p.Cys760Alafs*16 Frameshift exon 12 Patient 856 Homoz Consanguineous parents, both tested heterozygous. Two healthy siblings, both heterozygous for the mutation. 
c.2833A>G p.Arg945Glyfs*5 Splice-site exon 15 Patient 1146 ComHet See comment above for this patient. 
Missense affecting splicing (p.R945GfsX950), previously described (Stevanin et al., 2008; Crimella et al., 2009; Denora et al., 2009a). 
c.2842delG p.Val948Phefs*5 Frameshift exon 16 Patient 1637 ComHet Compound heterozygosity for p.Val948Phefs*5 and p.Leu1623Tyrfs*16, demonstrated heterozygous in both parents. 
c.2843insG p.Val948Glyfs*5 Frameshift exon 16 Patient 347 ComHet See comment above for the same patient. 
Previously described (p.V948GfsX953) (Stevanin et al., 2007). 
c.2850insT p.Leu950Phefs*3 Frameshift exon 16 Patient 1288 ComHet See comment above for the same patient. 
Previously described (p.L950FfsX953) (Denora et al., 2009a). 
c.3146-1G>C IVS17-1G>C Splice-site (intron 17) Patient 965 ComHet See comment above for the same patient. 
c.4307_4308delAA p.Gln1436Argfs*7 Frameshift exon 25 Patient 1518 ComHet See comment above for the same patient. 
Previously described (p.Q1436RfsX1442) (Stevanin et al., 2008; Zhang et al., 2008; Denora et al., 2009a). 
c.4868delT p.Leu1623Tyrfs*16 Frameshift exon 28 Patient 1637 ComHet See comment above for this patient. 
c.6789delA p.Gln2263Hisfs*3 Frameshift exon 37 Patient 2103 Homoz Consanguineous parents, demonstrated heterozygosity in the mother. 
c.6811_6812delCT p.Leu2271Aspfs*68 Frameshift exon 37 Patient 1105 ComHet See comment above for the same patient. 
c.6998A>G p.Gln2333Arg Missense exon 38 Patient 1465 ComHet Compound heterozygosity for p.Gln2333Arg (possible splice site alteration) and p.Arg651*(recurrent mutation). One healthy sister heterozygous for the p.Gln2333Arg. Parents not tested (Supplementary Fig.1A). 
c.7152-2A>G IVS39-2A>G Splice-site (intron 39) Patient 846 ComHet See comment above for the same patient. 
SPG11 cDNA Change SPG11 protein change Predicted effect Ex (intron) Patient code status# Comments, family segregation, references§ 
c.258-2A>C IVS1-2A>C Splice-site (intron 1) Patient 803 ComHet Compound heterozygosity for IVS1-2A>C (Schule, 2009), and p.Asn169Ile, demonstrated heterozygous in both parents. One unaffected brother did not carry any of the variants (Supplementary Fig. 1A). 
c.130delC p.Arg44Glyfs*13 Frameshift exon 1 Patient 1146 ComHet Compound heterozygosity for p.Arg44Glyfs*13 and p.Arg945Glyfs*5, demonstrated heterozygous in both parents. One affected brother not tested. 
c.165del19nc. p.Ser56Alafs*6 Frameshift exon 1 Patient 965 ComHet Compound heterozygosity for p.Ser56Alafs* and IVS17-1G>C demonstrated heterozygous in both parents. One unaffected brother heterozygous for IVS17-1G>C. 
c.434insC p.Gln145Profs*17 Frameshift exon 2 Patient 1253 Homoz Consanguineous parents, both tested heterozygous. Two healthy siblings (4 and 26 years) not tested. 
c.506A>T p.Asn169Ile Missense exon 3 Patient 803 ComHet See comment above for the same patient. 
c.662A>G p.Trp221* Nonsense exon 3 Patient 1518 ComHet Patient 846 ComHet Patient1518: compound heterozygosity for p.Trp221* (demonstrated in the father) and p.Gln1436Argfs*7 (demonstrated in the mother). 
Patient846: compound heterozygosity for p.Trp221* (demonstrated in the father) and IVS39-2A>G (demonstrated in the mother). No other siblings. 
The fathers of these two patients, heterozygous for p.Trp221*, did not share common intragenic polymorphisms. 
c.733_734delAT p.Met245Valfs*2 Frameshift exon 4 Patient 906 Homoz Consanguineous parents originating from Sicily, both tested heterozygous. Previously described in different subjects (Del Bo, 2007; Hehr et al.,2007; Stevanin et al., 2007, 2008; Boukhris et al., 2008; Liao et al., 2008; Denora et al., 2009a). In Italian patients this mutation is associated with p.Val2053Met variant (see text). 
c.1203delA p.Asp402Ilefs*13 Frameshift exon 6 Patient 1288 ComHet Previosuly described in Italian patients (Stevanin et al., 2007; Denora et al., 2009a). 
Patient1288: compound heterozygosity for p.Asp402Ilefs*13* (heterozygous mother and one healthy brother) and p.Leu950Phefs*3 (father and one healthy brother). 
Patient 347 ComHet Patient347: compound heterozygosity for p.Asp402Ilefs*13 (demonstrated in the father) and p.Val948Glyfs*5 (mother not tested). 
Haplotype analysis was not informative. Not known relations between families, but same geographical origin (South of Italy). 
c.1492C>T p.Gln498* Nonsense exon 7 Patient 1105 ComHet Compound heterozygosity for p.Gln498* (demonstrated in the mother) and p.Leu2271Aspfs*68 (in the father). One healthy sister, non-carrier. 
c.1951C>T p.Arg651* Nonsense exon 10 Patient 1804 Homoz Recurrent mutation (p.R651X) (Hehr et al., 2007; Stevanin et al., 2008; Denora et al., 2009a). The patients did not share common haplotypes (see also text). 
Patient 2055 Homoz  
Patient 1465 ComHet  
c.2277delC p.Cys760Alafs*16 Frameshift exon 12 Patient 856 Homoz Consanguineous parents, both tested heterozygous. Two healthy siblings, both heterozygous for the mutation. 
c.2833A>G p.Arg945Glyfs*5 Splice-site exon 15 Patient 1146 ComHet See comment above for this patient. 
Missense affecting splicing (p.R945GfsX950), previously described (Stevanin et al., 2008; Crimella et al., 2009; Denora et al., 2009a). 
c.2842delG p.Val948Phefs*5 Frameshift exon 16 Patient 1637 ComHet Compound heterozygosity for p.Val948Phefs*5 and p.Leu1623Tyrfs*16, demonstrated heterozygous in both parents. 
c.2843insG p.Val948Glyfs*5 Frameshift exon 16 Patient 347 ComHet See comment above for the same patient. 
Previously described (p.V948GfsX953) (Stevanin et al., 2007). 
c.2850insT p.Leu950Phefs*3 Frameshift exon 16 Patient 1288 ComHet See comment above for the same patient. 
Previously described (p.L950FfsX953) (Denora et al., 2009a). 
c.3146-1G>C IVS17-1G>C Splice-site (intron 17) Patient 965 ComHet See comment above for the same patient. 
c.4307_4308delAA p.Gln1436Argfs*7 Frameshift exon 25 Patient 1518 ComHet See comment above for the same patient. 
Previously described (p.Q1436RfsX1442) (Stevanin et al., 2008; Zhang et al., 2008; Denora et al., 2009a). 
c.4868delT p.Leu1623Tyrfs*16 Frameshift exon 28 Patient 1637 ComHet See comment above for this patient. 
c.6789delA p.Gln2263Hisfs*3 Frameshift exon 37 Patient 2103 Homoz Consanguineous parents, demonstrated heterozygosity in the mother. 
c.6811_6812delCT p.Leu2271Aspfs*68 Frameshift exon 37 Patient 1105 ComHet See comment above for the same patient. 
c.6998A>G p.Gln2333Arg Missense exon 38 Patient 1465 ComHet Compound heterozygosity for p.Gln2333Arg (possible splice site alteration) and p.Arg651*(recurrent mutation). One healthy sister heterozygous for the p.Gln2333Arg. Parents not tested (Supplementary Fig.1A). 
c.7152-2A>G IVS39-2A>G Splice-site (intron 39) Patient 846 ComHet See comment above for the same patient. 

Novel variants are highlighted in bold. # Zygosity status: Homoz = homozygous; ComHet = compound heterozygous. § = references for previously described SPG11/KIAA1840 mutations.

Table 2

SPG15/ZFYVE26 novel genetic variants identified in nine Italian patients

SPG15 cDNA change SPG15 protein change Predicted effect Ex (intron) Patient code status# Comments, family segregation§ 
c.1471C>T p.Gln491* Nonsense Patient 1256* Patient carrying the p.Gln491* and the p.2248delLys variants. Compound heterozygosity was demonstrated through validation in both parents. 
Exon 10 ComHet. 
c.1523T>A p.Ile508Asn Missense Patient 378 Family analysis: one affected sister was also homozygous for this mutation, and both healthy parents were heterozygous (Supplementary Fig. 1B). 
Exon 10 Homoz 
c.1730delA+c.1731C>T p.Asn577Ilefs*36 Frameshift Patient SLA766 Patient carried the p.Asn577Ilefs*36 and the p.Ser1312* variants. Compound heterozygosity was demonstrated through validation in both parents. No other affected family members. 
Exon 11 ComHet 
c.2254C>T p.Gln752* Nonsense Patient 730; Patient 930; Patient 688; Homoz Present in homozygous form in three apparently unrelated patients, originating form the same geographical region in the South of Italy, carrying also the c.735G>A (p.Glu245Glu) variant. 
Exon 12 
c.3935C>A p.Ser1312* Nonsense Patient 654,Homoz; Homozygous case had consanguineous parents. For the compound heterozygous see also comment above. 
Exon 21 Patient SLA766 ComHet 
c.4132C>T p.Arg1378* Nonsense Patient 1310 Sporadic. Demonstrated heterozygous in both parents. 
Exon 21 Homoz 
c.5203C>T p.Gln1735* Nonsense Patient 2079 Consanguineous parents, not tested. 
Exon 26 Homoz 
c.6744_6746delGAA p.2248delLys In-frame single amino acid deletion Exon 36 Patient 1256* ComHet See comment above for Patient 1256*. 
SPG15 cDNA change SPG15 protein change Predicted effect Ex (intron) Patient code status# Comments, family segregation§ 
c.1471C>T p.Gln491* Nonsense Patient 1256* Patient carrying the p.Gln491* and the p.2248delLys variants. Compound heterozygosity was demonstrated through validation in both parents. 
Exon 10 ComHet. 
c.1523T>A p.Ile508Asn Missense Patient 378 Family analysis: one affected sister was also homozygous for this mutation, and both healthy parents were heterozygous (Supplementary Fig. 1B). 
Exon 10 Homoz 
c.1730delA+c.1731C>T p.Asn577Ilefs*36 Frameshift Patient SLA766 Patient carried the p.Asn577Ilefs*36 and the p.Ser1312* variants. Compound heterozygosity was demonstrated through validation in both parents. No other affected family members. 
Exon 11 ComHet 
c.2254C>T p.Gln752* Nonsense Patient 730; Patient 930; Patient 688; Homoz Present in homozygous form in three apparently unrelated patients, originating form the same geographical region in the South of Italy, carrying also the c.735G>A (p.Glu245Glu) variant. 
Exon 12 
c.3935C>A p.Ser1312* Nonsense Patient 654,Homoz; Homozygous case had consanguineous parents. For the compound heterozygous see also comment above. 
Exon 21 Patient SLA766 ComHet 
c.4132C>T p.Arg1378* Nonsense Patient 1310 Sporadic. Demonstrated heterozygous in both parents. 
Exon 21 Homoz 
c.5203C>T p.Gln1735* Nonsense Patient 2079 Consanguineous parents, not tested. 
Exon 26 Homoz 
c.6744_6746delGAA p.2248delLys In-frame single amino acid deletion Exon 36 Patient 1256* ComHet See comment above for Patient 1256*. 

§All genetic variants identified in this study were novel. # Zygosity status: Homoz = homozygous; ComHet = compound heterozygous.

(*) and ()indicate the two compound heterozygous subjects of this series.

Table 3

SPG35/FA2H and SPG48/AP5Z1 genetic variants identified in the study

cDNA change Protein change Predicted effect Patient code status# Comments, family segregation§ 
SPG35/FA2H 
c.137G>A p.Gly46Asp Missense Patient H2788 Consanguineous parents. One unaffected sister demonstrated heterozygous. 
Homoz (Supplementary Fig. 1C
c.101A>G p.Tyr34Cys Missense Patient 73 Two affected brothers, compound heterozygous for the two missense variants. The mother and one healthy brother heterozygous for one of the variant. 
c.620C>T p.Thr207Met Missense ComHet (Supplementary Fig. 1C
c. 677G>A p.Trp226* Nonsense Patient 1891 Only one mutation identified. Consistent clinical phenotype. 
Heteroz 
    
SPG48/AP5Z1 
c.412C>T p.Arg138* Nonsense Patient 1738 p.Arg138* novel mutation. 
c.1322G>A p.Trp441* Nonsense ComHet Not reported in patients, but in exome variant database (rs3739194089, 1/12419 European alleles, MAF = 0.0081). Two unaffected daughters heterozygous for p.Trp441*. 
c. 616C>T p.Arg206Trp Missense Patient 1457 Homoz Consanguineous parents. Consanguineous parents, demonstrated heterozygosity in the father. 
(Supplementary Fig. 1D
cDNA change Protein change Predicted effect Patient code status# Comments, family segregation§ 
SPG35/FA2H 
c.137G>A p.Gly46Asp Missense Patient H2788 Consanguineous parents. One unaffected sister demonstrated heterozygous. 
Homoz (Supplementary Fig. 1C
c.101A>G p.Tyr34Cys Missense Patient 73 Two affected brothers, compound heterozygous for the two missense variants. The mother and one healthy brother heterozygous for one of the variant. 
c.620C>T p.Thr207Met Missense ComHet (Supplementary Fig. 1C
c. 677G>A p.Trp226* Nonsense Patient 1891 Only one mutation identified. Consistent clinical phenotype. 
Heteroz 
    
SPG48/AP5Z1 
c.412C>T p.Arg138* Nonsense Patient 1738 p.Arg138* novel mutation. 
c.1322G>A p.Trp441* Nonsense ComHet Not reported in patients, but in exome variant database (rs3739194089, 1/12419 European alleles, MAF = 0.0081). Two unaffected daughters heterozygous for p.Trp441*. 
c. 616C>T p.Arg206Trp Missense Patient 1457 Homoz Consanguineous parents. Consanguineous parents, demonstrated heterozygosity in the father. 
(Supplementary Fig. 1D

§All genetic variants, except p.Trp441* are novel. # Zygosity status: Homoz = Homozygous; ComHet = Compound Heterozygous; MAF = minor allele frequency.

Table 4

Demographic and clinical characteristics of genetically diagnosed subjects

 SPG11 SPG15 SPG35 SPG48 
n, patients/families 18/16 9/9 4/3 2/2 
Male/female 5M/13F 4M/5F 3M/1F 2F 
Family history 
    Sporadic 71% 75% 2/3 2/2 
    Affected siblings/cousins 18% 25% 1/3 
    Consanguineous parents 18% 50% 1/3 1/2 
     
Age at onset (Gait impairment)* 20.5 ± 7.7 (12–36) 16.3 ± 3.6 (11–23) 30.3 ± 18.3 (4–45) 2 and 47 years 
Age at examination* 28.9 ± 8.0 (20–49) 30.8 ± 10.0 (19–47) 41.0 ± 15.9 (19–57) 7 and 51 years 
Intellectual disability 
    Learning difficulties/mental retardation 67.0% 100% 2/4 1/2 
    Severe cognitive impairment 50% 88% 2/4 1/2 
Gait 
    Mild-moderate spastic gait 53% 44% 2/4 1/2 
    Walk with help 29% 22% 1/4 1/2 
    Wheelchair-bound 11% 33% 1/4 
Limb spasticity 
    Upper limbs 6% 33% 1/4 
    Lower limbs 100% 100% 4/4 2/2 
Limb muscle weakness 
    Upper limbs 38% 44% 
    Lower limbs 100% 66% 1/4 1/2 
Increased DTR 
    Upper limbs 80% 88% 2/4 2/2 
    Knee jerk 100% 100% 4/4 2/2 
    Ankle jerk 87% 66% 4/4 2/2 
Dysarthria 59% 55% 2/4 0/2 
Dysphagia 7% 38% 0/2 
Sphinteric dysfunctions 19% 44% 1/4 1/2 
Extrapyramidal signs 44% 2/4 
Epilepsy 20% 
Increased CK (>195 U/l) 47% 50% 1/2 
Brain MRI 
    TCC 100% 100% 3/4 2/2 
    WML 100% 100% 4/4 2/2 
    Cerebellar atrophy 4/4 
 SPG11 SPG15 SPG35 SPG48 
n, patients/families 18/16 9/9 4/3 2/2 
Male/female 5M/13F 4M/5F 3M/1F 2F 
Family history 
    Sporadic 71% 75% 2/3 2/2 
    Affected siblings/cousins 18% 25% 1/3 
    Consanguineous parents 18% 50% 1/3 1/2 
     
Age at onset (Gait impairment)* 20.5 ± 7.7 (12–36) 16.3 ± 3.6 (11–23) 30.3 ± 18.3 (4–45) 2 and 47 years 
Age at examination* 28.9 ± 8.0 (20–49) 30.8 ± 10.0 (19–47) 41.0 ± 15.9 (19–57) 7 and 51 years 
Intellectual disability 
    Learning difficulties/mental retardation 67.0% 100% 2/4 1/2 
    Severe cognitive impairment 50% 88% 2/4 1/2 
Gait 
    Mild-moderate spastic gait 53% 44% 2/4 1/2 
    Walk with help 29% 22% 1/4 1/2 
    Wheelchair-bound 11% 33% 1/4 
Limb spasticity 
    Upper limbs 6% 33% 1/4 
    Lower limbs 100% 100% 4/4 2/2 
Limb muscle weakness 
    Upper limbs 38% 44% 
    Lower limbs 100% 66% 1/4 1/2 
Increased DTR 
    Upper limbs 80% 88% 2/4 2/2 
    Knee jerk 100% 100% 4/4 2/2 
    Ankle jerk 87% 66% 4/4 2/2 
Dysarthria 59% 55% 2/4 0/2 
Dysphagia 7% 38% 0/2 
Sphinteric dysfunctions 19% 44% 1/4 1/2 
Extrapyramidal signs 44% 2/4 
Epilepsy 20% 
Increased CK (>195 U/l) 47% 50% 1/2 
Brain MRI 
    TCC 100% 100% 3/4 2/2 
    WML 100% 100% 4/4 2/2 
    Cerebellar atrophy 4/4 

*mean ± standard deviation, (range); DTR = deep tendon reflexes; CK = creatine kinase; TCC = thin corpus callosum; WML = white matter lesions.

The 31 undiagnosed patients had ages at onset and clinical findings similar to the genetically diagnosed subjects (mean age at onset 15.2 ± 10.9, range 1–47 years). Progressive lower limb spasticity was associated with mental retardation, thin corpus callosum and white matter in 12 of 31 patients, and six of them had peripheral neuropathy. Nine cases had mental retardation associated with either thin corpus callosum or white matter lesions (29%), and the remaining negative cases had spasticity plus thin corpus callosum or white matter lesions, variably associated with peripheral neuropathy (eight cases), epilepsy (four), cerebellar atrophy (four), and retinopathy (two). In all negative cases, mutations in SPG4/SPAST and in SPG7/SPG7 genes were excluded (Casari et al., 1998; Hazan et al., 1999).

Spastic paraplegia 11

Molecular findings

Twenty-two SPG11/KIAA1840 different variants were identified: 14 were novel, and eight were previously reported (Del Bo et al., 2007; Hehr et al., 2007; Stevanin et al., 2007, 2008; Boukhris et al., 2008; Liao et al., 2008; Zhang et al., 2008; Crimella et al., 2009; Denora et al., 2009a; Schüle et al., 2009) (Table 1). The majority of the novel variants predicted the truncation of the SPG11 protein and a consequent loss of function mechanism. Only two missense variants were identified: the p.Asn169Ile (c.506A>T), and the p.Gln2333Arg (c.6998A>G). The patient with the p.Asn169Ile missense variant (Patient 803), also carried the splice-site variant c.258-2A>C, previously described in another SPG11 case (Schüle et al., 2009). The patient with the missense p.Gln2333Arg, carried the p.Arg651* nonsense mutation on the second allele (Stevanin et al., 2008). These novel missense variations involved well-conserved amino acids in other vertebrates (Supplementary Fig. 1), were not reported as SNPs (according to Human Gene Mutation Database, NCBI–dbSNP132ver and Exome Variant Server databases), and were not found in healthy Italian controls. In silico analyses predicted both the mutations as probably damaging: scores for p.Asn169Ile were 1.000 (PolyPhen-2-HumDiv), 0.991 (PolyPhen-2-HumVar), and 0 (SIFT). For p.Gln2333Arg scores were 1.000 (HumDiv), 0.999 (HumVar) and 0.01 (SIFT). Splice site prediction analysis (Mutation T@sting and Alternative Splice Site Predictor) indicated that the p.Gln2333Arg could cause an alteration of the normal splicing by abolishing the physiological acceptor site.

In addition, we found frequent SPG11 gene polymorphisms, previously reported in the literature, (dbSNP rs3759875, p.Phe463Ser, and dbSNP rs3759873, p.Tyr2341Tyr) (Stevanin et al., 2008), and sequence changes already described in the exome variant database: c.1108G>A (p.Glu370Lys, rs77697105), c.2083G>A (p.Ala695Thr, rs78183930), and c.3037A>G (p.Lys1013Glu, rs111347025). The frequencies of these latter variants range from 0.98 to 1.5% in control populations and are likely to be polymorphisms.

We used these intragenic SNPs to recognize possible common haplotypes in the patients sharing the same mutation. The novel p.Trp221* was present in two unrelated patients (Patients 1518 and 846), both inheriting the variant from the heterozygous fathers (Table 1). In one case the mutation was transmitted in association with p.Phe463Ser and the p.Tyr2341Tyr polymorphisms, whereas no polymorphisms were present in the other family. In two families with the p.Arg651*mutation (Patients 1804 and 2055) no known polymorphisms were found, and haplotype analysis was performed using four markers in linkage with the SPG11 gene (RH79982, SGC33828, STS-T97010, D15S1325). Also in these cases the alleles carrying the p.Arg651* variant showed different haplotypes in the patients, suggesting the presence of multiple ancestors and the possibility of a mutational hot-spot. This finding is consistent with previous studies describing different haplotypes associated with this recurrent mutation (Hehr et al., 2007; Stevanin et al., 2008; Denora et al., 2009a). In two families carrying the recurrent p.Asp402Ilefs*13 variant (Families 1288 and 347), SNPs analysis was not informative (Table 1).

Finally, we identified two rare SPG11 variants: the c.6224A>G (p.Asn2075Ser, rs140824939, MAF of 0.36%), and the c.6157G>A (p.Val2053Met, rs149003934, MAF of 0.023%). The patient carrying the first variant did not present other SPG11 pathogenic mutations, and thus was classified as a SPG11-negative case. The patient carrying the p.Val2053Met variant, also carried in homozygous form the c.733_734delAT mutation (p.Met245Valfs*2), as previously reported in another Italian SPG11 case (Del Bo et al., 2007). This finding confirms that the p.Val2053Met has not a pathogenic role for the SPG11 disease, but it represents, in Italian descents, a variant in linkage disequilibrium with the frame shift c.733_734delAT mutation. The p.Met245Valfs*2 mutation has been reported in other SPG11 families from various geographical origin and with different haplotypes (Denora et al., 2009a).

Clinical features

Clinical characteristics of the patients with SPG11 are summarized in Table 4. All patients with SPG11 belong to newly diagnosed Italian families, except one [DNA code of Family 347, previously described by Stevanin et al. (2007), Family PE, in the early stage of the disease]. The majority of the cases (12 of 16 patients) presented signs of cognitive impairment several years before the recognition of movement dysfunctions. Nine patients presented moderate to severe cognitive decline (IQ range 55–72), and seven cases had an IQ score within the normal range, but presented deficit in memory and in calculation tests. No extrapyramidal signs were observed, 50% had pes cavus, and 36% had nystagmus and mild limb dysmetria.

Electroneuronography assessments showed normal or only mildly reduced sensory and motor conduction velocities, whereas ∼25% of cases had signs of axonal, sensory and motor, neuropathy. EMG studies revealed chronic neurogenic alterations in the majority of the patients (62% at lower limbs, 33% at upper limbs). Moderate to severe spontaneous activity was detected in ∼20% of cases at lower limbs (Fig. 2A and Supplementary Table 1).

Figure 2

Electrophysiological results in patients with SPG11 and with SPG15. In A, the graph shows the percentage of SPG11 and SPG15 patients with electrophysiological abnormalities at motor nerve conduction studies (electroneuronography) and EMG. (B) Results of multimodal evoked potentials (for details, see Supplementary Tables 1 and 2). ↓SNAP amp = reduction of sensory nerve action potential amplitude; SNCV = sensory nerve conduction velocity; CMAP amp = compound muscle action potential amplitude; MNCV = motor nerve conduction velocity; UL = upper limbs; LL = lower limb; CMCT = central motor conduction time; MEP amp = motor evoked potential amplitude; SEP CCT = somatosensory evoked potential central conduction time; VEP 30' lat. P100 = visual evoked potential 30' P100 latency; VEP 30' P100 amp = 30' P100 amplitude.

Figure 2

Electrophysiological results in patients with SPG11 and with SPG15. In A, the graph shows the percentage of SPG11 and SPG15 patients with electrophysiological abnormalities at motor nerve conduction studies (electroneuronography) and EMG. (B) Results of multimodal evoked potentials (for details, see Supplementary Tables 1 and 2). ↓SNAP amp = reduction of sensory nerve action potential amplitude; SNCV = sensory nerve conduction velocity; CMAP amp = compound muscle action potential amplitude; MNCV = motor nerve conduction velocity; UL = upper limbs; LL = lower limb; CMCT = central motor conduction time; MEP amp = motor evoked potential amplitude; SEP CCT = somatosensory evoked potential central conduction time; VEP 30' lat. P100 = visual evoked potential 30' P100 latency; VEP 30' P100 amp = 30' P100 amplitude.

Motor evoked potentials showed increased central conduction time and/or reduced amplitude in 90% of patients at lower limbs, and in 10% of cases at upper limbs. Sensory evoked potentials were altered at lower limbs (62% of cases). Visual evoked potentials showed moderate abnormalities in 28% of cases, whereas brainstem auditory evoked potentials were normal in the all examined patients (Fig. 2B and Supplementary Table 2). None of the patients complained of decreased visual acuity. Electroretinogram was performed in eight patients, and in three cases the scotopic and photopic responses were found decreased in amplitude, but with latency times within the normal range.

Brain MRI showed in all cases a moderate to severe thin corpus callosum, particularly evident in the anterior portion with relatively preserved or only mildly reduced volume of the splenium. White matter hyperintensity was diffuse in the frontal horns and in the peritrigonal regions in 82% of cases; and in 18% it was visible only in the region of the frontal horns (‘ears of the lynx’ shape). Mild to severe enlargement of the lateral ventricles and of the cerebral sulci was observed in 81%, and mild reduction of the cerebral white matter in 82%. Cerebellum, brainstem and cervical part of the spinal cord were normal (Fig. 3A–C).

Figure 3

Brain MRI characteristics. Sagittal T1- (A, D, G and J), coronal T2- (B, E, H and K), and axial T2-weighted spin echo images (C, F, I and L). Axial and coronal sections in SPG11 (A–C, Patient 2055) show the increased signal intensity in the region of the forceps minor of the corpus callosum, (arrows in B and C), with a ‘ears of the lynx’ appearance (C), not visible in SPG15 (D–F Patient 654). In SPG11 a dramatic enlargement of the interhemispheric and frontal sulci (B), despite a quite normal cerebellar volume is visible. In SPG35 (G–I, Patient 73), diffuse cerebral atrophy, global cerebellar atrophy (H and I), thin corpus callosum (G), and periventricular white matter hyperintensity is visible (arrowheads in I). In SPG48 (J–L, Patient 1738), mild cerebral and cerebellar atrophy is recognizable, associated with a vanish anterior periventricular white matter hyperintensity (arrows in L). The thinning of the corpus callosum (A, D and J), is more marked in SPG11.

Figure 3

Brain MRI characteristics. Sagittal T1- (A, D, G and J), coronal T2- (B, E, H and K), and axial T2-weighted spin echo images (C, F, I and L). Axial and coronal sections in SPG11 (A–C, Patient 2055) show the increased signal intensity in the region of the forceps minor of the corpus callosum, (arrows in B and C), with a ‘ears of the lynx’ appearance (C), not visible in SPG15 (D–F Patient 654). In SPG11 a dramatic enlargement of the interhemispheric and frontal sulci (B), despite a quite normal cerebellar volume is visible. In SPG35 (G–I, Patient 73), diffuse cerebral atrophy, global cerebellar atrophy (H and I), thin corpus callosum (G), and periventricular white matter hyperintensity is visible (arrowheads in I). In SPG48 (J–L, Patient 1738), mild cerebral and cerebellar atrophy is recognizable, associated with a vanish anterior periventricular white matter hyperintensity (arrows in L). The thinning of the corpus callosum (A, D and J), is more marked in SPG11.

Spastic paraplegia 15

Molecular findings

Nine unrelated patients carried mutations in the SPG15/ZFYVE26 gene (Table 2). Three apparently unrelated sporadic cases carried the same nonsense p.Gln752* mutation (c.2254C>T). Family history reveals that these patients (Patients 730, 930, 688; Table 2) originated from the same geographical area in southern Italy, suggesting the possibility of a founder effect mechanism. This hypothesis was also supported by the finding, in all three patients, of a novel synonymous substitution p.Glu245Glu (c.735G>A), not present in other SPG15 patients or in control population, and likely in linkage dysequilibrium with the p.Gln752* mutation (Table 2). One patient was homozygous for a novel missense variant, p.Ile508Asn, also present, in homozygous forms, in the affected sister. The variant was absent in 540 control chromosomes from Italian control subjects and in SNP databases (Supplementary Fig. 1B). In silico analyses predicted the p.Ile508Asn missense variant as possibly damaging (PolyPhen-2 scores, HumDiv and HumVar, 0.925 and 0.667, respectively; and SIFT score 0) (Supplementary Fig. 1B). One compound heterozygote patient (Patient 1256, Table 2) carried a novel nonsense mutation (p.Gln491*, c.1471C>T) and a trinucleotide deletion, causing an in-frame skipping of a single amino acid codon for lysine. The second compound heterozygote patient (Patient SLA766) carried a truncation mutation (c.1730delA+c.1731C>T, p.Asn577Ilefs*36), and the nonsense p.Ser1312* mutation (Table 2).

Finally, we identified several previously reported SPG15/ZFYVE26 nucleotide changes causing synonymous substitutions (p.Pro704Pro, p.Leu853Leu, p.Pro1070Pro, p.Leu1253Leu), or single amino acid changes (p.Pro1103Leu, p.Arg1241Gln, p.Cys1457Tyr, p.Asn1891Ser, and p.Ser1893Ile) (MAF ranging from 0.007 to 0.94%).

Clinical features

Clinical characteristics of the SPG15 patients are summarized in Table 4. Motor symptoms were noticed in the second decade of life, and were always preceded by cognitive impairment (mental retardation in seven patients, and learning difficulties in two patients). None of our patients with SPG15 complained of decreased visual acuity, formal ocular examination did not reveal retinal changes, whereas electroretinogram showed mild signs of retinal degeneration in one case. Four subjects presented bradykinesia and rigidity at upper limbs. Complete neurophysiological assessments were performed in seven patients with SPG15. All patients had normal or mildly reduced sensory and motor nerve conduction velocities. Reduced sensory amplitudes were found only at lower limbs (28% of cases). Axonal motor neuropathy was detected in 25% of the case at upper limbs, and in 42% at lower limbs. EMG studies revealed chronic neurogenic alterations in the majority of the cases, equally involving upper and lower limbs (71%), and severe spontaneous muscle activity in 28% of case (at four limbs) (Fig. 2A and Supplementary Table 1). Motor evoked potentials showed severe abnormalities, both at upper and lower limbs, in 83% of cases. Sensory evoked potentials were predominantly altered at lower limbs (80% of cases), and only mildly affected at upper limbs (20%). Electroretinogram was performed in four patients, and visual evoked potentials showed moderate abnormalities in three cases, whereas brainstem auditory evoked potentials were normal in all examined patients (Fig. 2B and Supplementary Table 2).

Brain MRI showed severe thinning of anterior portion of the corpus callosum, and mild thinning of the splenium region. White matter hyperintensity was always present. In four cases white matter hyperintensity was localized in the frontal horns (‘ears of the lynx’ shape), and in the remaining cases was diffuse in the frontal horns and in the peritrigonal regions. Mild to severe enlargement of the lateral ventricles and cerebral sulci was observed in all cases, and mild diffuse reduction of the cerebral white matter in four patients. Mild atrophy in the cerebellar culmen was observed in three cases, and in one mild brainstem atrophy was present (Fig. 3D–F).

Spastic paraplegia 35

We found two SPG35/FA2H missense variants (p.Tyr34Cys and p.Thr207Met) in a compound heterozygous subject presenting leg spasticity at the age of 32, and subsequently developing severe paraparesis and progressive dementia (at age 40) (Table 3). This patient had one brother with similar clinical characteristics and disease progression. Segregation analysis in the family showed that both affected siblings had the same genotype, whereas the mother and one unaffected brother were heterozygous for one of the two mutations (Supplementary Fig. 1C). In both affected individuals, brain MRI showed a severe enlargement of the lateral ventricles and of the cerebral sulci, extremely thin corpus callosum, diffuse white matter hyperintensity in the centrum semiovale, severe cerebellar atrophy, both of the vermis and cerebellar hemispheres, and no sign of iron accumulation (Fig. 3G–I and Table 4).

The second SPG35 diagnosed patient was a 19-year-old male, from Morocco, carrying the novel missense p.Gly46Asp (c.137G>A) variant in homozygous form. His consanguineous parents (first cousins) and two siblings were unaffected (Supplementary Fig.1C). The patient presented mental retardation and gait difficulties since the age of 4 years. At latest examination, he presented moderate cognitive impairment (IQ = 51), mild ophthalmoparesis in the vertical gaze, spastic gait, mildly increased muscle tone at lower limbs, increased deep tendon reflexes with ankle clonus, and Babinski sign. Brain MRI showed a slight enlargement of the lateral ventricles and sulci, mild hyperintensity of the periventricular white matter, and mild diffuse cerebellar atrophy and thin corpus callosum. Electroneuronography was normal, and no spontaneous muscle activity was detected. Motor evoked potentials were altered at lower limbs in all three subjects. Bilateral Visual Evoked Potentials (VEP) alterations were found in one patient, and mild sensorineural hearing loss in two cases.

Finally, we identified a putative nonsense heterozygous FA2H gene mutation, c.677G>A (p.Trp226*), in a 57-year old female, presented late-onset progressive spastic paraparesis, moderate cognitive deficit, ophthalmoplegia in the vertical gaze, and saccadic pursuit eye movement in the horizontal gaze (Table 4). Brain MRI findings were similar to those described in the first two patients.

Spastic paraplegia 48

Sequence analysis of the SPG48/AP5Z1 gene revealed the presence two nonsense genetic variants p.Trp441* and p.Arg138* in a compound heterozygous patient of Italian origin, and a missense variant, p.Arg206Trp, in a homozygous patient originating from Morocco (Table 3).

The first patient was a 51-year-old female with adult onset slowly progressive spastic paraplegia, wide-based gait, mild dysmetria at upper limbs and urinary incontinence. No cognitive deficit was observed. Motor Evoked Potentials (MEP) showed severe abnormalities at lower limbs, whereas electroneuronography and EMG examinations were normal. Brain MRI showed severe thin corpus callosum in the anterior part (splenium appeared normal), and mild white matter hyperintensity at the frontal horns of the lateral ventricles (Fig. 3L–N). Cerebral ventricles, sulci, cerebellum, and brainstem were normal.

The second patient, homozygous for the missense variant p.Arg206Trp (HumDiv = 1; HumVar = 0.976), had consanguineous parents, and presented since the first year of life with mild psychomotor delay and walking abnormalities (tiptoes gait). First observation, at age 3, showed spastic paraparesis, and mild mental retardation. Visual Evoked Potentials (VEP), electroretinogram, electroneuronography and EMG examinations were normal. Brain MRI showed mild thin corpus callosum only at the isthmus part, and mild white matter periventricular hyperintensities. At the age of 7 years, clinical features and neuroradiological imaging were unchanged. Neurocognitive evaluation confirmed mild intellectual impairment (IQ score = 62).

We also identified three novel missense SPG48/AP5Z1 gene variants: p.Arg110Trp (c.328C>T), p.Arg179Trp (c.535C>T), p.Tyr779Asn (c.2335T>A), and one synonymous variant p.Gly378Gly (c.1134G>T) in heterozygous patients, in whom exon deletion or duplication were excluded. In silico analyses predicted a possible or probable damaging effect on the protein, but a final genetic diagnosis in these patients was not achieved.

Discussion

In this study we describe and compare clinical and genetic features of a large series of patients with SPG11 and with SPG15, and additional cases associated with SPG35/FA2H, and SPG48/AP5Z1 gene mutations. We started to genetically characterize our patients after the identification of the SPG11/KIAA1840 and SPG15/ZFYVE26 genes (Stevanin et al., 2007; Hanein et al., 2008). Subsequently, in parallel with the discovery of other SPG genes responsible for autosomal recessive hereditary SPG with the thin corpus callosum phenotype, we continued screening our undiagnosed cases.

In agreement with literature data, we found that SPG11 was the most frequent genotype, accounting for 26.2% of our cases; and the SPG15 was second most frequent form with 14.8% of the diagnosed cases (Elleuch et al., 2007; Denora et al., 2009b; Goizet et al., 2009). The majority of the SPG11/KIAA1840 and SPG15/ZFYVE26 genetic variants predict the truncation of the corresponding proteins, thus confirming a predominant loss-of-function mechanism. Missense pathogenic mutations are very rare in patients with SPG11 (Denora et al., 2009a), and they may represent polymorphisms or nucleotide changes determining an alteration in the splice consensus site, e.g. for the p.Arg945Glyfs*5 and the p.Arg815Met mutations (Stevanin et al., 2008; Crimella et al., 2009; Denora et al., 2009a). We identified a novel missense change (p.Gln2333Arg) predicted in silico to cause an alteration in the splice site; however, we could not test this effect on mRNA from living cells. A second missense variation, p.Asn169Ile, was not previously reported as polymorphism, but its possible pathogenic mechanism remains to be elucidated (Supplementary Fig. 1A). We hypothesized that this missense change represents a rare polymorphism in linkage disequilibrium with a second pathogenic mutation not yet identified, as it has demonstrated for the recurrent p.Val2053Met variation, also present in one of our patients with SPG11 in association with the p.Met245Valfs*2 mutation.

As compared with cases with SPG11, that were predominantly compound heterozygous, the SPG15 cases were mostly caused by homozygous private mutations, indicating a predominant effect of consanguinity. Up to date, all described SPG15/ZFYVE26 variants were truncating mutations. We identified the first missense variant (p.Ile508Asn) in two homozygous affected siblings. Segregation analysis in the family and the analysis of a large control population supported the hypothesis of a disease-causing mutation; however, in the absence of additional elements in favour of its pathogenicity, this variant should be considered as a putative mutation.

At clinical examination, the phenotype of SPG11 and SPG15 patients was mostly indistinguishable: learning difficulties or mental retardation in childhood was the first symptom in the great majority of the cases, followed by the appearance, during the second or third decades, of walking difficulties and progressive spasticity (Stevanin et al., 2008; Denora et al., 2009a, b; Schüle et al., 2009). Maculopathy, cerebellar ataxia and peripheral neuropathy were not constant features, whereas cognitive impairment, thin corpus callosum and white matter abnormalities represent the most typical findings (Elleuch et al., 2007; Hanein et al., 2008). Extrapyramidal signs were observed only in four patients with SPG15. Levodopa-responsive parkinsonism has been described as the first symptom of the disease both in patients with SPG11 and with SPG15; however, in our cases, the extrapyramidal signs were observed during the course of the disease, always associated with the typical signs of SPG (Anheim et al., 2009; Paisán-Ruiz et al., 2010; Schicks et al., 2011; Mallaret et al., 2014).

In this study, electrophysiological investigations demonstrated motor axonal neuropathy in ∼60% of cases with SPG11 and 70% of cases with SPG15. Cases with SPG11 presented a predominant involvement of lower limbs, whereas those with SPG15 showed a more widespread axonal motor neuropathy, equally involving upper and lower limbs (Fig. 2, Supplementary Tables 1 and 2). Spontaneous activity at EMG needle examination was also present in a small percentage of both cases with SPG11 and with SPG15. The evidence of involvement of both upper and lower motor neuron may conform to the possible diagnosis of amyotrophic lateral sclerosis (ALS). The hypothesis that SPG11/KIAA1840 could be an ALS-causative gene has been supported by the finding of mutations in families with autosomal recessive slowly progressive ALS, thus widening the clinical spectrum of SPG11 (Orlacchio et al., 2010). We also performed a genetic screening in a series of 47 cases with ALS (eight sporadic, and 39 with autosomal recessive inheritance), presenting a disease duration of 1.8 years (range 1–9 years), and including both juvenile and adult late-onset cases. In our series of patients with ALS we did not find SPG11/KIAA1840 gene mutations (personal communication).

Interestingly, we found, three unrelated cases associated with novel SPG35/FA2H gene variants. SPG35 was mapped to chromosome 16q21-q23, in an Omani family with autosomal recessive hereditary SPG, intellectual disability, seizures and leukodystrophy (Dick et al., 2008). The disease gene, FA2H, encodes the enzyme fatty acid 2-hydroxylase that catalyses the 2-hydroxylation of membrane myelin lipids. SPG35/FA2H mutations were found in several families with infantile onset progressive spasticity, dystonia, cognitive deterioration, cerebellar signs, leukodystrophy, thin corpus callosum and brain iron accumulation (Edvardson et al., 2008; Dick et al., 2010; Kruer et al., 2010; Cao et al., 2013). Our cases with SPG35 presented atypical characteristics with respect to the previous cases, with clinical and radiological signs partially overlapping those of the SPG11–SPG15 cohort. The most obvious clinical differences observed in our cases with SPG35 were the presence of cerebellar atrophy at brain MRI and the late onset disease in three of four subjects. Only two late-onset SPG35 cases, from a single family, have been reported so far, and they showed cerebral and cerebellar atrophy, but neither iron accumulation nor white matter abnormalities (Tonelli et al., 2012). In our cases with SPG35 we found both thin corpus callosum and white matter hyperintensities, severe cerebellar atrophy and no brain iron accumulation. In FA2H related-disorders it has been hypothesized that disease severity may be influenced by the residual enzyme activity. Thus, truncating mutations could be associated with severe manifestations, whereas a missense mutation causes a milder phenotype (Edvardson et al., 2008; Dick et al., 2010; Kruer et al., 2010). In our small group of patients with SPG35, however, we could not confirm this type of genotype–phenotype correlation. In fact, the young patient presenting the symptoms in early childhood, carried a homozygous missense variant (p.Gly46Asp), and the SPG35 families with late onset disease, carried missense and truncating FA2H variants.

In our study, we also observed peculiar phenotypes in two novel cases associated with SPG48/AP5Z1 gene variants. At present, only one SPG48 French family has been described (Słabicki et al., 2010). These patients presented with late-onset SPG, normal brain MRI and white matter changes only in the spinal cord. One of our patients with SPG48 was a 5-year-old child carrying a homozygous p. Arg206Trp missense variant, and presenting a clinical and MRI phenotype completely overlapping with that of the SPG11–SPG15 patients. The second case with SPG48 was a 51-year-old female carrying two different nonsense variants (p.Arg138*; p. Trp441*), and also sharing most of typical SPG11–SPG15 features, except for a late disease onset and only mild cognitive deficit.

Also, we would like to remark on the large group of cases (51%) that were not genetically diagnosed in our cohort of patients. None of the cases were caused by mutation in SPG21/ACP33 or in SPG54/DDHD2 genes. SPG21 (Mast syndrome) has been described, so far, only in Amish patients (Simpson et al., 2003). Mutations in SPG54/DDHD2 gene, encoding one of the three mammalian intracellular phospholipases A1, have been recently identified in four families with complex autosomal recessive hereditary SPG with thin corpus callosum and periventricular white matter hyperintensity (Schuurs-Hoeijmakers et al., 2012). The large group of non-diagnosed subjects suggests further genetic heterogeneity. During the course of the study, we progressively updated and expand our genetic screening by sequencing one-by-one, in the chronological order of discovery, most of the new genes associated with the phenotypes of our patients. The use of more advanced gene screening techniques, such as the high-throughput sequencing platforms and whole-exome sequencing, could probably achieve a more successful ratio of genetic diagnosis, as demonstrated by the recent study of Novarino et al. (2014). From the clinical point of view, our data confirm a consistent and indistinguishable clinical and radiological phenotype in patients with SPG11 and with SPG15 (independently of the type of mutation), and their relative frequencies among the group of the autosomal recessive hereditary SPGs. SPG35 and SPG48 genotypes are rare, and manifest with more variable age at onset and phenotypes. Common clinical features in this group of disorders not necessarily imply similar functions of the hereditary spastic paraplegia proteins (Fink, 2013). Though, the protein encoded by the SPG48/AP5Z1 gene has been recently shown to interact and to co-localize with both SPG11 and SPG15 proteins (Hirst et al., 2013), suggesting the possibility of related molecular and biochemical pathways.

Acknowledgements

We are grateful to Drs Carlo Antozzi, Angelo Sghirlanzoni, Roberto Fancellu, Cinzia Andrigo, Graziella Uziel, Anna Rita Giovagnoli, Silvia Baratta, and Massimo Plumari for their help in providing clinical and/or genetic data of some of the patients.

Funding

This work was partially supported by AriSLA, grant NOVALS 2010 and Healthcare Research of the Ministry of Health to V.P, B.C, C.G.

Supplementary material

Supplementary material is available at Brain online.

Abbreviations

    Abbreviations
  • ALS

    amyotrophic lateral sclerosis

  • SNP

    single nucleotide polymorphism

  • SPG

    spastic paraplegia

References

Abou Jamra
R
Philippe
O
Raas-Rothschild
A
Eck
SH
Graf
E
Buchert
R
, et al.  . 
Adaptor protein complex 4 deficiency causes severe autosomal-recessive intellectual disability, progressive spastic paraplegia, shy character, and short stature
Am J Hum Genet
 , 
2011
, vol. 
88
 (pg. 
788
-
95
)
Anheim
M
Lagier-Tourenne
C
Stevanin
G
Fleury
M
Durr
A
Namer
IJ
, et al.  . 
SPG11 spastic paraplegia. A new cause of juvenile parkinsonism
J Neurol
 , 
2009
, vol. 
256
 (pg. 
104
-
8
)
Boukhris
A
Stevanin
G
Feki
I
Denis
E
Elleuch
N
Miladi
MI
, et al.  . 
Hereditary spastic paraplegia with mental impairment and thin corpus callosum in Tunisia: SPG11, SPG15, and further genetic heterogeneity
Arch Neurol
 , 
2008
, vol. 
65
 (pg. 
393
-
402
)
Boukhris
A
Stevanin
G
Feki
I
Denora
P
Elleuch
N
Miladi
MI
, et al.  . 
Tunisian hereditary spastic paraplegias: clinical variability supported by genetic heterogeneity
Clin Genet
 , 
2009
, vol. 
75
 (pg. 
527
-
36
)
Cao
L
Huang
XJ
Chen
CJ
Chen
SD
A rare family with Hereditary Spastic Paraplegia Type 35 due to novel FA2H mutations: a case report with literature review
J Neurol Sci
 , 
2013
, vol. 
329
 (pg. 
1
-
5
)
Casari
G
De Fusco
M
Ciarmatori
S
Zeviani
M
Mora
M
Fernandez
P
, et al.  . 
Spastic paraplegia and OXPHOS impairment caused by mutations in paraplegin, a nuclear-encoded mitochondrial metalloprotease
Cell
 , 
1998
, vol. 
93
 (pg. 
973
-
83
)
Crimella
C
Arnoldi
A
Crippa
F
Mostacciuolo
ML
Boaretto
F
Sironi
M
, et al.  . 
Point mutations and a large intragenic deletion in SPG11 in complicated spastic paraplegia without thin corpus callosum
J Med Genet
 , 
2009
, vol. 
46
 (pg. 
345
-
51
)
Del Bo
R
Di Fonzo
A
Ghezzi
S
Locatelli
F
Stevanin
G
Costa
A
, et al.  . 
SPG11: a consistent clinical phenotype in a family with homozygous spatacsin truncating mutation
Neurogenetics
 , 
2007
, vol. 
8
 (pg. 
301
-
5
)
den Dunnen
JT
Antonarakis
SE
Mutation nomenclature extensions and suggestions to describe complex mutations: a discussion
Hum Mutat
 , 
2000
, vol. 
15
 (pg. 
7
-
12
)
Denora
PS
Schlesinger
D
Casali
C
Kok
F
Tessa
A
Boukhris
A
, et al.  . 
Screening of ARHSP-TCC patients expands the spectrum of SPG11 mutations and includes a large scale gene deletion
Hum Mutat
 , 
2009a
, vol. 
30
 (pg. 
E500
-
19
)
Denora
PS
Muglia
M
Casali
C
Truchetto
J
Silvestri
G
Messina
D
, et al.  . 
Spastic paraplegia with thinning of the corpus callosum and white matter abnormalities: further mutations and relative frequency in ZFYVE26/SPG15 in the Italian population
J Neurol Sci
 , 
2009b
, vol. 
277
 (pg. 
22
-
5
)
Dick
KJ
Al-Mjeni
R
Baskir
W
Koul
R
Simpson
MA
Patton
MA
, et al.  . 
A novel locus for an autosomal recessive hereditary spastic paraplegia (SPG35) maps to 16q21-q23
Neurology
 , 
2008
, vol. 
71
 (pg. 
248
-
52
)
Dick
KJ
Eckhardt
M
Paisán-Ruiz
C
Alshehhi
AA
Proukakis
C
Sibtain
NA
, et al.  . 
Mutation of FA2H underlies a complicated form of hereditary spastic paraplegia (SPG35)
Hum Mutat
 , 
2010
, vol. 
31
 (pg. 
E1251
-
60
)
Edvardson
S
Hama
H
Shaag
A
Gomori
JM
Berger
I
Soffer
D
, et al.  . 
Mutations in the fatty acid 2-hydroxylase gene are associated with leukodystrophy with spastic paraparesis and dystonia
Am J Hum Genet
 , 
2008
, vol. 
83
 (pg. 
643
-
8
)
Elleuch
N
Bouslam
N
Hanein
S
Lossos
A
Hamri
A
Klebe
S
, et al.  . 
Refinement of the SPG15 candidate interval and phenotypic heterogeneity in three large Arab families
Neurogenetics
 , 
2007
, vol. 
8
 (pg. 
307
-
15
)
Fink
JK
Hereditary spastic paraplegia: clinico-pathologic features and emerging molecular mechanisms
Acta Neuropathol
 , 
2013
, vol. 
126
 (pg. 
307
-
28
)
Finsterer
J
Löscher
W
Quasthoff
S
Wanschitz
J
Auer-Grumbach
M
Stevanin
G
Hereditary spastic paraplegias with autosomal dominant, recessive, X-linked, or maternal trait of inheritance
J Neurol Sci
 , 
2012
, vol. 
318
 (pg. 
1
-
18
)
Goizet
C
Boukhris
A
Maltete
D
Guyant-Maréchal
L
Truchetto
J
Mundwiller
E
, et al.  . 
SPG15 is the second most common cause of hereditary spastic paraplegia with thin corpus callosum
Neurology
 , 
2009
, vol. 
73
 (pg. 
1111
-
9
)
Hanein
S
Martin
E
Boukhris
A
Byrne
P
Goizet
C
Hamri
A
, et al.  . 
Identification of the SPG15 gene, encoding spastizin, as a frequent cause of complicated autosomal-recessive spastic paraplegia, including Kjellin syndrome
Am J Hum Genet
 , 
2008
, vol. 
82
 (pg. 
992
-
1002
)
Harding
AE
Classification of the hereditary ataxias and paraplegias
Lancet
 , 
1983
, vol. 
1
 (pg. 
1151
-
5
)
Hazan
J
Fonknechten
N
Mavel
D
Paternotte
C
Samson
D
Artiguenave
F
, et al.  . 
Spastin, a new AAA protein, is altered in the most frequent form of autosomal dominant spastic paraplegia
Nat Genet
 , 
1999
, vol. 
23
 (pg. 
296
-
303
)
Hehr
U
Bauer
P
Winner
B
Schule
R
Olmez
A
Koehler
W
, et al.  . 
Long-term course and mutational spectrum of spatacsin-linked spastic paraplegia
Ann Neurol
 , 
2007
, vol. 
62
 (pg. 
656
-
65
)
Hirst
J
Borner
GH
Edgar
J
Hein
MY
Mann
M
Buchholz
F
, et al.  . 
Interaction between AP-5 and the hereditary spastic paraplegia proteins SPG11 and SPG15
Mol Biol Cell
 , 
2013
, vol. 
24
 (pg. 
2558
-
69
)
Kruer
MC
Paisán-Ruiz
C
Boddaert
N
Yoon
MY
Hama
H
Gregory
A
, et al.  . 
Defective FA2H leads to a novel form of neurodegeneration with brain iron accumulation (NBIA)
Ann Neurol
 , 
2010
, vol. 
68
 (pg. 
611
-
8
)
Liao
SS
Shen
L
Du
J
Zhao
GH
Wang
XY
Yang
Y
, et al.  . 
Novel mutations of the SPG11 gene in hereditary spastic paraplegia with thin corpus callosum
J Neurol Sci
 , 
2008
, vol. 
275
 (pg. 
92
-
9
)
Mallaret
M
Lagha-Boukbiza
O
Biskup
S
Namer
IJ
Rudolf
G
Anheim
M
, et al.  . 
SPG15: a cause of juvenile atypical levodopa responsive parkinsonism
J Neurol
 , 
2014
, vol. 
261
 (pg. 
435
-
7
)
Nakamura
A
Izumi
K
Umehara
F
Kuriyama
M
Hokezu
Y
Nakagawa
M
, et al.  . 
Familial spastic paraplegia with mental impairment and thin corpus callosum
J Neurol Sci
 , 
1995
, vol. 
131
 (pg. 
35
-
42
)
Novarino
G
Fenstermaker
AG
Zaki
MS
Hofree
M
Silhavy
JL
Andrew
D
, et al.  . 
Exome sequencing links corticospinal motor neuron disease to common neurodegenerative disorders
Science
 , 
2014
, vol. 
343
 (pg. 
506
-
11
)
Orlacchio
A
Babalini
C
Borreca
A
Patrono
C
Massa
R
Basaran
S
, et al.  . 
SPATACSIN mutations cause autosomal recessive juvenile amyotrophic lateral sclerosis
Brain
 , 
2010
, vol. 
133
 
Pt 2
(pg. 
591
-
8
)
Paisán-Ruiz
C
Guevara
R
Federoff
M
Hanagasi
H
Sina
F
Elahi
E
, et al.  . 
Early-onset L-dopa responsive parkinsonism with pyramidal signs due to ATP13A2, PLA2G6, FBXO7 and spatacsin mutations
Mov Disord
 , 
2010
, vol. 
25
 (pg. 
1791
-
800
)
Pierson
TM
Simeonov
DR
Sincan
M
Adams
DA
Markello
T
Golas
G
, et al.  . 
Exome sequencing and SNP analysis detect novel compound heterozygosity in fatty acid hydroxylase-associated neurodegeneration
Eur J Hum Genet
 , 
2012
, vol. 
20
 (pg. 
476
-
9
)
Pippucci
T
Panza
E
Pompilii
E
Donadio
V
Borreca
A
Babalini
C
, et al.  . 
Autosomal recessive hereditary spastic paraplegia with thin corpus callosum: a novel mutation in the SPG11 gene and further evidence for genetic heterogeneity
Eur J Neurol
 , 
2009
, vol. 
16
 (pg. 
121
-
6
)
Schicks
J
Synofzik
M
Petursson
H
Huttenlocher
J
Reimold
M
Schols
L
, et al.  . 
Atypical juvenile parkinsonism in a consanguineous SPG15 family
Mov Disord
 , 
2011
, vol. 
26
 (pg. 
564
-
6
)
Schüle
R
Schöls
L
Genetics of hereditary spastic paraplegias
Semin Neurol
 , 
2011
, vol. 
31
 (pg. 
484
-
93
)
Schüle
R
Schlipf
N
Synofzik
M
Klebe
S
Klimpe
S
Hehr
U
, et al.  . 
Frequency and phenotype of SPG11 and SPG15 in complicated hereditary spastic paraplegia
J Neurol Neurosurg Psychiatry
 , 
2009
, vol. 
80
 (pg. 
1402
-
4
)
Schuurs-Hoeijmakers
JH
Geraghty
MT
Kamsteeg
EJ
Ben-Salem
S
de Bot
ST
Nijhof
B
, et al.  . 
Mutations in DDHD2, encoding an intracellular phospholipase A(1), cause a recessive form of complex hereditary spastic paraplegia
Am J Hum Genet
 , 
2012
, vol. 
91
 (pg. 
1073
-
81
)
Simpson
MA
Cross
H
Proukakis
C
Pryde
A
Hershberger
R
Chatonnet
A
, et al.  . 
Maspardin is mutated in mast syndrome, a complicated form of hereditary spastic paraplegia associated with dementia
Am J Hum Genet
 , 
2003
, vol. 
73
 (pg. 
1147
-
56
)
Słabicki
M
Theis
M
Krastev
DB
Samsonov
S
Mundwiller
E
Junqueira
M
, et al.  . 
A genome-scale DNA repair RNAi screen identifies SPG48 as a novel gene associated with hereditary spastic paraplegia
PLoS Biol
 , 
2010
, vol. 
8
 pg. 
e1000408
 
Stevanin
G
Santorelli
FM
Azzedine
H
Coutinho
P
Chomilier
J
Denora
PS
, et al.  . 
Mutations in SPG11, encoding spatacsin, are a major cause of spastic paraplegia with thin corpus callosum
Nat Genet
 , 
2007
, vol. 
39
 (pg. 
366
-
72
)
Stevanin
G
Azzedine
H
Denora
P
Boukhris
A
Tazir
M
Lossos
A
, et al.  . 
Mutations in SPG11 are frequent in autosomal recessive spastic paraplegia with thin corpus callosum, cognitive decline and lower motor neuron degeneration
Brain
 , 
2008
, vol. 
131
 
Pt 3
(pg. 
772
-
84
)
Tonelli
A
D'Angelo
MG
Arrigoni
F
Brighina
E
Arnoldi
A
Citterio
A
, et al.  . 
Atypical adult onset complicated spastic paraparesis with thin corpus callosum in two patients carrying a novel FA2H mutation
Eur J Neurol
 , 
2012
, vol. 
19
 (pg. 
e127
-
9
)
Zhang
SS
Chen
Q
Chen
XP
Wang
JG
Burgunder
JM
Shang
HF
, et al.  . 
Two novel mutations in the SPG11 gene causing hereditary spastic paraplegia associated with thin corpus callosum
Mov Disord
 , 
2008
, vol. 
23
 (pg. 
917
-
9
)

Author notes

*These authors contributed equally to this work.