High frequency of SPG4 in Taiwanese families with autosomal dominant hereditary spastic paraplegia
High frequency of SPG4 in Taiwanese families with autosomal dominant hereditary spastic paraplegia
Min-Yu Lan 2 3
Yung-Yee Chang 2 3
Tu-Hseuh Yeh 0 1 6
Szu-Chia Lai 0 1
Chia-Wei Liou 2
Hung-Chou Kuo 5 6
Yih-Ru Wu 5 6
Rong-Kuo Lyu 5 6
Jen-Wen Hung 4
Ying-Chao Chang 7
Chin-Song Lu 0 1 6 8
0 Neuroscience Research Center, Chang Gung Memorial Hospital , Linkou, Taoyuan , Taiwan
1 Section of Movement Disorder, Department of Neurology, Chang Gung Memorial Hospital, Linko Medical Center and Chang Gung University College of Medicine , 5 Fu-Shin St, Kwei-Shan, Taoyuan 333 , Taiwan
2 Department of Neurology, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of Medicine , Kaohsiung , Taiwan
3 Center for Parkinson's Disease, Kaohsiung Chang Gung Memorial Hospital , Kaohsiung , Taiwan
4 Department of Rehabilitation, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of Medicine , Kaohsiung , Taiwan
5 Department of Neurology, Chang Gung Memorial Hospital Linko Medical Center and Chang Gung University College of Medicine , Taoyuan , Taiwan
6 Healthy Aging Research Center, Chang Gung University , Taoyuan , Taiwan
7 Department of Pediatrics, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of Medicine
8 School of Traditional Chinese Medicine, College of Medicine, Chang Gung University , Taoyuan , Taiwan
Background: Hereditary spastic paraplegias (HSPs) are a group of neurodegenerative diseases characterized by progressive spasticity and weakness of the lower limbs. SPG4, SPG3A and SPG31 are the three leading causes of autosomal dominant (AD) HSPs. Methods: A total of 20 unrelated AD-HSP families were recruited for clinical and genetic assessment. Detection of mutations in SPG4, SPG3A and SPG31 genes was conducted according to a standard protocol. Genotype-phenotype correlations and determinants for disease severity and progression were analyzed. Results: Mutations in the SPG4 gene (SPAST) were detected in 18 (90%) of the AD-HSP families. Mutations in SPG4, SPG3A and SPG31 genes were not detected in the remaining two families. Considerable variations in clinical features were noted, even for mutation carriers from the same family. Mutations causing complete loss of the spastin AAA cassette were associated with earlier onset of disease (20 ± 18 years) compared with those with preservation of partial or total AAA cassette (32 ± 19 years, p = 0.041). For those with SPG4 mutations, disease severity was related to the patients' current age, and the progression rate of disease was positively correlated with age at onset. Conclusions: SPG4 accounts for most of the AD-HSP cases in Taiwanese, with a frequency significantly higher than in other populations. SPAST mutations which predict complete loss of the spastin AAA cassette were associated with an earlier onset of disease.
Hereditary spastic paraplegia; SPG4; SPAST; Spastin; AAA cassette
Hereditary spastic paraplegias (HSPs) are a heterogeneous
group of neurodegenerative disorders clinically
characterized by progressive spasticity and weakness of the lower
limbs . They can be classified according to the mode of
inheritance, or by the absence (“pure form”) or presence
(“complicated form”) of additional neurological or systemic
manifestations such as dementia, epilepsy, mental
retardation, thin corpus callosum, cerebellar ataxia, peripheral
neuropathy, deafness, retinopathy, and optic atrophy [1,2].
Pure forms of HSPs are usually inherited as an autosomal
dominant trait, whereas most complicated HSPs are
autosomal recessive forms. To date, more than 70 loci for
HSPs have been mapped and at least 55 genes have been
Autosomal dominant HSPs (AD-HSPs) represent
around 70% of cases of uncomplicated HSPs . SPG4 is
the most common form of AD-HSP, accounting for
around 40% of the families in previous reports [2,5].
Spastin, the protein encoded by the SPG4 gene, is a member of
the family of AAA proteins (ATPase associated with
diverse cellular activities) which share a common functional
domain called the AAA cassette . SPG3A is the second
most frequent form of AD-HSP, representing
approximately 7-10% of all AD-HSP families [7,8]. The SPG3A
gene encodes the protein atlastin-1 which is structurally
homologous to members in the dynamin superfamily of
GTPases . SPG31 has been suggested to be the third
most common cause of AD-HSP, with an overall
frequency of 2.3-6.5% in the kindreds of primarily European
descent [10,11]. The SPG31 gene encodes receptor
expression-enhancing protein 1 which may be involved in
a chaperone-like function .
We previously reported four Taiwanese families with
SPG4-related HSP . In this study, we extended the
scope of the investigation to screen SPG4, SPG3A and
SPG31 in additional AD-HSP families. The results showed
an unexpectedly high frequency of SPG4 in the cohort of
Taiwanese AD-HSP families.
The study protocol was approved by the Chang Gung
Memorial Hospital Institutional Review Board (No.
1004460C). Clinical and genetic features of the participants
were examined after written informed consent had been
obtained from participants or a parent while participants
Patients and clinical examinations
This study included 20 unrelated AD-HSP kindreds (49
cases in total) of Han Taiwanese ethnicity collected from
the Departments of Neurology, Rehabilitation and
Pediatric, Chang Gung Memorial Hospitals, Taiwan
(Additional file 1: Figure S1). Families 1 to 4 have been
reported previously . A diagnosis of AD-HSP was
based on the following diagnostic criteria: (1) spastic
paraplegia, or (2) spastic tetraparesis with earlier and
more severe affliction of the lower limbs, or (3) spastic
paraplegia as an early and prominent sign of a
degenerative disease affecting the nervous system; (4) positive
family history of spastic paraplegia with affected
members in at least two generations; and (5) no other causes
of the presenting symptoms . Family history was
recorded according to the patients’ reports, and
inheritance was ascertained in 14 families by examining
affected family members. Clinical information was
collected and neurological examinations were performed.
Ambulatory disability was scored based on a five-point
scale: 0 asymptomatic, 1 minimally impaired (able to
run), 2 mildly impaired (able to walk independently but
not run), 3 moderately impaired (walk with an aid), and
4 severely impaired (wheelchair bound).
Genomic DNA was extracted from peripheral blood
leukocytes for genetic analysis, and mutation screenings
were conducted in the following order. First, nucleotide
substitutions and small deletions or insertions in all
exons and their adjacent splicing sites of the SPG4 gene
(SPAST) were analyzed by polymerase chain reactions
and direct sequencing. If a mutation was not detected,
multiplex ligation-dependent probe amplification (MLPA)
was performed to detect exonic deletions in SPAST.
Finally, for the cases with no detected mutations in SPAST,
sequence analysis and MLPA of the SPG3A gene (ATL1)
and sequence analysis of the SPG31 gene (REEP1) were
performed. Detected mutations were checked in 100
control DNA samples collected from unrelated ethnic Han
MLPA was performed using SALSA MLPA kits (P165-B1
HSP probemix; MRC-Holland, Amsterdam, Netherlands)
according to the manufacturer’s instructions. For each
sample, a normalized value ratio for a relative peak area
between 0.8 and 1.2 was considered normal. A heterozygous
deletion was expected with a ratio between 0.3 and 0.7, and
a heterozygous duplication between 1.3 and 1.7. Total
mRNA from the blood leukocytes was extracted using
Trizol procedure (Invitrogen, Carlsbad, CA, United States).
mRNA analysis was performed as described previously 
to assess alternations in gene transcription due to SPAST
Novelty of a detected mutation was checked with the
Human Gene Mutation Database . Amino acid
conservation in orthologues was assessed using the
HomoloGene database . Prediction of pathogenicity of a
missense mutation was analysed using the PolyPhen-2
Data were presented as mean ± SD unless otherwise
specified. Disease severity was defined as “mild” with an
ambulation score of 1 or 2, or “severe” with a score of 3
or 4. The progression rate of the disease was assessed
with a disease-progression score (DPS) calculated as the
current disability score divided by disease duration.
Mutations were classified according to the consequence,
namely, whether they caused a complete loss of the
AAA cassette (i.e. nonsense and frameshift mutations
proximal to the encoding sequence of the AAA cassette,
whole gene deletion, and mRNA decay) or not. Age at
onset (AAO) with respect to mutation consequence was
tested with t-tests (asymptomatic cases were excluded
from the analysis). The relationships between disease
severity and current age, gender, disease duration and
mutation consequence were tested using the chi-square test
or t-test. The relationships between DPS and AAO,
gender and mutation consequence were tested using the
Pearson test or t-test. Independence of the association
between disease severity (or DPS) and clinical and
genetic factors was examined using logistic (or linear)
regression analysis. A p value <0.05 was considered to be
Table 1 Summary of the families in the current study
Family Sex Age (decade) Age at onset Ambulatory DPS SPAST mutation
1‡ F 6 2 2
F* 4 2 2
2‡ F* 7 5 3
F 4 2 2
M 1 1 1
3‡ M* 7 5 4
M 6 6 2
M 6 5 3
M 4 - 0
F 3 3 1
M 3 2 3
F 3 - 0
4‡ F 8 6 4
M* 6 5 3
F 5 5 2
5 M* 7 2 3
6 M 5 - 0
M* 2 1 1
M 1 1 1
7 F* 6 5 2
M 4 1 2
F 5 1 1
M 2 1 1
M 5 2 2
8 F* 6 5 3
9 M 4 3 3
F 1 1 1
M* 1 1 1
10 F 8 4 1
M 6 1 1
F* 5 2 1
M 1 1 2
11 F* 5 4 2
Table 1 Summary of the families in the current study (Continued)
1.67 Not detected
3.33 c.1738_1740 del ATT/ins
A total of 14 different SPAST mutations were identified
in 18 families (Table 1). In this cohort, SPG4 accounted
for 90% of the AD-HSP families. Mutation types were
quite heterogeneous, including nonsense (2), missense
(2), insertion (2), deletion (2), insertion-deletion (1),
splicing site (2), and large exonic deletion (3) (Figure 1).
Four mutations (c.1361_1363 ins GGG, c.1005-1 G > C,
c.1004 + 1 G > T, and c.1738_1740 del ATT/ins GA)
were novel mutations. These mutations were neither
included in the data derived from the 1000 Genome
Project (http://www.1000genomes.org/)  nor detected
in the 100 control DNA samples. The residue leucine
affected by the missense mutation p.L461P and the
residue aspartate affected by D555G were highly
evolutionarily conserved according to the public genome database
(Additional file 2: Figure S2). The substitutions of amino
acids caused by the two missense mutations were
predicted to be damaging (a score of 1.000 and 0.988
respectively) by the PolyPhen-2 program. In general, nonsense
and missense mutations were located in the 3′ half of
SPAST exons, while frameshift (mini-insertions and
minideletions) and splicing site mutations were more evenly
distributed in the gene. Co-segregation of mutations with
the disease was confirmed in 14 families, in which at least
two affected members underwent genetic testing.
Mutations in SPAST, ATL1 and REEP1 were not detected in
two families (Families 14 and 19).
SPAST mRNA analysis
For the in-frame insertion c.1361_1363 ins GGG
detected in Family 6, only the wild type of mRNA was
detected in mRNA analysis and the decay of mutant
mRNA is suspected (Additional file 3: Figure S3). For
the other insertions and mutations of missense, nonsense,
Figure 1 SPAST mutations detected in the current study. Schematic representation of the location of SPAST mutations detected in this study.
Numbered boxes indicate transcript coding regions in exons drawn to scale. Encoding region of the AAA cassette is highlighted.
and deletion types, mRNA analysis showed nucleotide
changes identical to those in genomic DNA. The two
splicing site mutations in intron 6 led to skipping of
transcription of the first 8 bases of exon 7 (Family 9) and the whole
of exon 6 (Family 10), respectively. For the exonic deletion
revealed by MLPA in Family 2, cDNA analysis showed a
deletion of 1090 bases at the 5′-end of exon 17, as
reported previously . In Family 4, the effects of complete
loss of exon 17 on translation could not be determined
due to failure to amplify the mutant transcript.
In this AD-HSP cohort, AAO in the probands ranged
from 1 to 48 years, and six (30%) families had at least
one case whose AAO was less than 10 years. The clinical
characteristics of the SPG4 cases are summarized in
Table 2. The onset of disease ranged from 1 to 69 years
of age. Among the 47 cases carrying SPAST mutations,
44 (94%) showed spastic paraparesis and 3 (6%) were
asymptomatic. All of the symptomatic SPG4 patients
manifested a pure phenotype, except for a case in Family
7 who presented with distal amyotrophy in the upper
limbs. The most common clinical presentation was gait
disorder due to spasticity or increased tendon reflexes in
the lower limbs. Neurological examinations of the
cranial nerves and upper limbs were normal except for an
accentuated jaw-jerk and tendon reflex in some patients.
Considerable variations in AAO, rate of progression or
severity of disease were noted, even in mutation carriers
from the same family (Table 1). There was no clear
phenotype correlation with respect to the different types
of mutations. For the two families without detected
Mean age (range), years
Mean age at onset (range), years*
Mean disease duration (range), years*
Increased tendon reflexes in the upper limbs
Increased tendon reflexes in the lower limbs
Presence of extensor plantar reflex
Presence of ankle clonus
Attenuated vibration perception at the ankles
Dysfunction of bladder control†
Ambulation function, 0/1/2/3/4
28 ± 19 (1 – 69)
17 ± 12 (3 – 43)
*asymptomatic cases not included.
†urinary retention, frequency or incontinence.
‡0, asymptomatic; 1, able to run; 2, unable to run, walking independently;
3, walking with an aid; 4, wheelchair bound.
mutations, one had a pure phenotype while the other
presented with deafness in the affected members.
For the SPG4 cases, mutations causing complete loss of
the AAA cassette were associated with an earlier onset of
disease (20 ± 18 years, n =16) compared with those with
preservation of partial or total AAA cassette (32 ± 19 years,
n = 28; p = 0.041). In univariate analysis, disease severity
was related with the patients’ current age (37 ± 19 years
for mild disease vs. 55 ± 12 years for severe disease,
p <0.001), and DPS was correlated with AAO (r = 0.564,
p <0.001) (Additional file 4: Table S1 and Additional file 5:
Table S2). After adjusting for gender, disease duration and
consequence of mutation, the association of age (odds
ratio 1.10, 95% confidence interval 1.04-1.16 for every
increase of 1 year, p = 0.002) with disease severity, and that
of AAO (β = 0.038, standard error 0.010, p <0.001) with
DPS remained significant. Consequence of SPAST
mutation was not related with either disease severity or DPS.
Compared to previous series in other populations, the
current findings showed a significantly higher frequency
of SPG4 for AD-HSP in our ethnically Han Taiwanese
cohort (90% vs. 30-60%; Additional file 6: Table S3).
Previously, a study of HSP cohort in Sardinia, Italy showed
that SPG4 was responsible for all of nine AD-HSP
families receiving genetic test . However, eight of these
SPG4 families were attributable to a same multi-exonic
deletion in SPAST, suggesting a founder effect for this
mutation in this population. In contrast, the
predominance of SPG4 in our cohort could not be due to specific
founder mutations because most of the detected
mutations were unique to the respective families. Although
710% of AD-HSPs are known to be caused by SPG3A,
none of our cases had ATL1 mutations. SPG3A is
characterized by early onset of disease, usually before 10 years
of age . Failure to detect SPG3A in this study due to
ascertainment bias is unlikely in view of the wide range
of AAO (1–48 years) in the probands and the inclusion
of pediatric cases in this cohort. A possible explanation
for this phenomenon may be ethnic differences in the
genetic background. However, because of the relatively
small kindred number in the present study, this finding
needs to be confirmed with a larger cohort of AD-HSP
The current study showed that missense mutations
were located in the AAA cassette-encoding region,
while frameshift mutations which create premature
termination codons were scattered along the SPAST gene
(Figure 1). In spite of the heterogeneity in mutation
types, all of these mutations may either alter the
structure or disrupt the integrity of the AAA cassette.
Further, most of these mutations were predicted to
produce a truncated spastin or absence of mRNA
transcription. This finding suggests that haploinsufficiency due
to the abolishment of ATPase activity is the most
important pathogenicity for disease-causing SPAST
For SPG4, mutations causing a complete loss of the
spastin AAA cassette were associated with earlier onset
of disease than those with at least partial AAA cassette
preservation. However, further studies of the production
of truncated spastin proteins by the mutations
predicting partial AAA loss and function of spastin protein
carrying partial AAA cassette in cellular levels will be
needed to provide biological evidence for this finding.
Except for the common motor dysfunction due to
spastic paraparesis, remarkable variations in onset age,
progression rate and disease severity were noted in the
SPG4 patients, even for those from the same family. We
also found that later onset of the disease was associated
with faster disease progression. This finding is
consistent with Harding’s observation on the pure form
ADHSP  and has also been noted in previous studies in
SPG4 [19,20]. In addition, the patient’s current age was
the most significant determinant for disease severity.
Taken together, these findings suggest that the SPG4
phenotype is the product of interactions between SPAST
mutations and environmental or other genetic factors.
There are some limitations to this study. First, this
study was based on a relatively small cohort of AD-HSP
kindreds collected from two medical centers in Taiwan.
Our findings need to be confirmed by a more extensive
nationwide survey of AD-HSP families. Second, this is a
cross-sectional study, and disease progression would be
more accurately evaluated by a longitudinal observation.
Third, the five-graded ambulation score did not have
enough sensitivity in detecting minor differences in
disease severity. A more detailed assessment scale, such as
Spastic Paraplegia Rating Scale , would be better in
reflecting overall functional impairment. Fourth, because
the age of symptom onset was determined according to
patient reports, there could be errors in recalling the
information, especially for those with long-lasting disease.
Finally, this study did not test the genetic modifiers,
such as some sequence variants in ATL1 and HSPD1,
which have been postulated to account for the variability
in SPG4 phenotypes .
SPG4 accounted for a significantly higher proportion of
AD-HSPs in ethnic Taiwanese than in other populations.
There was no clear genotype-phenotype correlation
among the SPG4 patients, except that mutations
resulting in complete loss of the spastin AAA cassette were
associated with an earlier onset of disease.
Additional file 1: Figure S1. Pedigrees of the 20 HSP families in the
study. The pedigrees of the 20 hereditary spastic paraplegia families
included in the study. The circles represent female subjects and the
squares represent male subjects, and the diamond represents the patient
whose gender was withheld for confidentiality reasons. The filled
symbols indicate affected individuals and the question marks indicate
uncertain disease affection. The cases that were involved in the clinical
and genetic studies are indicated with a cross bar above the individual
symbol. The probands are indicated by arrows.
Additional file 2: Figure S2. Conservation of the mutated residues in
different species. Conservation of the residues affected by the mutations
p.L461P and p.D555G in different species.
Additional file 3: Figure S3. cDNA analysis of the mutations. cDNA
analysis of the SPAST mutations in Family 2 (large deletion of 5′ region of
exon 17), Family 9 (spicing acceptor mutation, the boxed sequence
representing the nucleotides spliced during transcription due to the
mutation) and Family 10 (splicing donor mutation).
Additional file 4: Table S1. Determinants of SPG4 disease severity.
Univariate and multivariate regression analyses for determinants of
disease severity in the SPG4 cases.
Additional file 5: Table S2. Determinants of SPG4 disease progression.
Univariate and multivariate analyses for determinants of disease
progression score in the SPG4 cases.
Additional file 6: Table S3. Proportions of SPG4 in different
populations. Proportions of SPG4 in autosomal dominant hereditary
spastic paraplegias (AD-HSPs) in different populations.
AAA: ATPase associated with diverse cellular activities; AAO: Age at onset;
AD: Autosomal dominant; DPS: Disease progression score; HSP: Hereditary
spastic paraplegia; MLPA: Multiplex ligation-dependent probe amplification.
The authors declare that they have no competing interests.
MYL: project conception and design, clinical and genetic data collection and
analysis, statistical analysis, writing of the first draft. YYC: project conception
and design, clinical data collection and analysis, statistical analysis. THY:
genetic analysis, organization of the study, manuscript review and critique.
SCL: genetic analysis, organization of the study. CWL, HCK, YRW, RKL, JWH,
YCC: clinical data collection and analysis. CSL: project conception and design,
organization of the study, manuscript review and critique. All authors read
and approved the final manuscript.
We are grateful to all the families with hereditary spastic paraplegia who
contributed to this research project. We are also in debt to grants CMRPG
3D0371 (to Dr. Lu), CMRPG 3D0381 (to Dr. Yeh), and CMRPD 1B0331 (to the
Healthy Aging Research Center).
1. Harding AE : Hereditary spastic paraplegias . Semin Neurol 1993 , 13 : 333 - 336 .
2. Fink JK : Advances in the hereditary spastic paraplegias . Exp Neurol 2003 , 184 (Suppl 1): S106 - S110 .
3. Lo Giudice T , Lombardi F , Santorelli FM , Kawarai T , Orlacchio A : Hereditary spastic paraplegia: clinical-genetic characteristics and evolving molecular mechanisms . Exp Neurol 2014 , 261C : 518 - 539 .
4. Fink JK , Heiman-Patterson T , Bird T , Cambi F , Dubé MP , Figlewicz DA , Fink JK , Haines JL , Heiman-Patterson T , Hentati A , Pericak-Vance MA , Raskind W , Rouleau GA , Siddique T : Hereditary spastic paraplegias: advances in genetic research: hereditary spastic paraplegia working group . Neurology 1996 , 46 : 1507 - 1514 .
5. Proukakis C , Auer-Grumbach M , Wagner K , Wilkinson PA , Reid E , Patton MA , Warner TT , Crosby AH : Screening of patients with hereditary spastic paraplegia reveals seven novel mutations in the SPG4 (Spastin) gene . Hum Mutat 2003 , 21 : 170 .
6. Hazan J , Fonknechten N , Mavel D , Paternotte C , Samson D , Artiguenave F , Davoine CS , Cruaud C , Dürr A , Wincker P , Brottier P , Cattolico L , Barbe V , Burgunder JM , Prud'homme JF , Brice A , Fontaine B , Heilig B , Weissenbach J : Spastin, a new AAA protein, is altered in the most frequent form of autosomal dominant spastic paraplegia . Nat Genet 1999 , 23 : 296 - 303 .
7. Sauter SM , Engel W , Neumann LM , Kunze J , Neelsen J : Novel mutations in the Atlastin gene (SPG3A) in families with autosomal dominant hereditary spastic paraplegia and evidence for late onset forms of HSP linked to the SPG3A locus . Hum Mutat 2004 , 23 : 98 .
8. Namekawa M , Ribai P , Nelson I , Forlani S , Fellmann F , Goizet C , Depienne C , Stevanin G , Ruberg M , Dürr A , Brice A : SPG3A is the most frequent cause of hereditary spastic paraplegia with onset before age of 10 years . Neurology 2006 , 66 : 112 - 114 .
9. Zhao X , Alvarado D , Rainier S , Lemons R , Hedera P , Weber CH , Tukel T , Apak M , Heiman-Patterson T , Ming L , Bui M , Fink JK : Mutations in a newly identified GTPase gene cause autosomal dominant hereditary spastic paraplegia . Nat Genet 2001 , 29 : 326 - 331 .
10. Züchner S , Wang G , Tran-Viet KN , Nance MA , Gaskell PC , Vance JM , AshleyKoch AE , Pericak-Vance MA : Mutations in the novel mitochondrial protein REEP1 cause hereditary spastic paraplegia type 31 . Am J Hum Genet 2006 , 79 : 365 - 369 .
11. Hewamadduma C , McDermott C , Kirby J , Grierson A , Panayi M , Dalton A , Rajabally Y , Shaw P : New pedigrees and novel mutation expand the phenotype of REEP-1 associated hereditary spastic paraplegia (HSP) . Neurogenetics 2009 , 10 : 105 - 110 .
12. Lan MY , Fu SC , Chang YY , Wu-Chou YH , Lai SC , Chen RS , Lu CS : Clinical and genetic analysis of four Taiwanese families with autosomal dominant hereditary spastic paraplegia . J Formos Med Assoc 2012 , 111 : 380 - 385 .
13. Cooper DN , Ball EV , Stenson PD , Phillips AD , Howells K , Mort ME : http://www.hgmd. cf.ac.uk/ac/index.php (accessed 01/06/ 2014 ).
14. National Center of Biotechnology Information, U.S. Library of Medicine: http://www.ncbi. nlm.nih.gov/homologene (accessed 15/06/ 2014 ).
15. Adzhubei IA , Schmidt S , Peshkin L , Ramensky VE , Gerasimova A , Bork P , Kondrashov AS , Sunyaev SR : A method and server for predicting damaging missense mutations . Nat Methods 2010 , 7 : 248 - 249 .
16. 1000 Genomes Project Consortium, Abecasis GR , Auton A , Brooks LD , DePristo MA , Durbin RM , Handsaker RE , Kang HM , Marth GT , McVean GA : An integrated map of genetic variation from 1,092 human genomes . Nature 2012 , 491 : 56 - 65 .
17. Racis L , Tessa A , Di Fabio R , Storti E , Agnetti V , Casali C , Santorelli FM , Pugliatti M : The high prevalence of hereditary spastic paraplegia in Sardinia, insular Italy . J Neurol 2014 , 261 : 52 - 59 .
18. Harding AE : Hereditary “pure” spastic paraplegia: a clinical and genetic study of 22 families . J Neurol Neurosurg Psychiatry 1981 , 44 : 871 - 883 .
19. Loureiro JL , Brandão E , Ruano L , Brandão AF , Lopes AM , Thieleke-Matos C , Miller-Fleming L , Cruz VT , Barbosa M , Silveira I , Stevanin G , Pinto-Basto J , Sequeiros J , Alonso I , Coutinho P : Autosomal dominant spastic paraplegias: a review of 89 families resulting from a Portuguese survey . JAMA Neurol 2013 , 70 : 481 - 487 .
20. Fonknechten N , Mavel D , Byrne P , Davoine CS , Cruaud C , Bönsch D , Samson D , Coutinho P , Hutchinson M , McMonagle P , Burgunder JM , Tartaglione A , Heinzlef O , Feki I , Deufel T , Parfrey N , Brice A , Fontaine B , Prud'homme JF , Weissenbach J , Dürr A , Hazan J : Spectrum of SPG4 mutations in autosomal dominant spastic paraplegia . Hum Mol Genet 2000 , 9 : 637 - 644 .
21. Schüle R , Holland-Letz T , Klimpe S , Kassubek J , Klopstock T , Mall V , Otto S , Winner B , Schöls L : The Spastic Paraplegia Rating Scale (SPRS): a reliable and valid measure of disease severity . Neurology 2006 , 67 : 430 - 434 .
22. Svenstrup K , Bross P , Koefoed P , Hjermind LE , Eiberg H , Born AP , Vissing J , Gyllenborg J , Nørremølle A , Hasholt L , Nielsen JE : Sequence variants in SPAST, SPG3A and HSPD1 in hereditary spastic paraplegia . J Neurol Sci 2009 , 284 : 90 - 95 .