Human Genetics

, Volume 122, Issue 3, pp 261–273

A novel locus for autosomal dominant “uncomplicated” hereditary spastic paraplegia maps to chromosome 8p21.1-q13.3

Authors

  • Sylvain Hanein
    • INSERM, Unit 679, 47 Bd de l’Hôpital
    • Institut Fédératif de Recherche en Neurosciences (IFR70), Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6, UMR S679
    • INSERM, Unit 679, 47 Bd de l’Hôpital
    • Institut Fédératif de Recherche en Neurosciences (IFR70), Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6, UMR S679
    • Département de Génétique et CytogénétiqueAP-HP, Groupe Hospitalier Pitié-Salpêtrière
  • Pascale Ribai
    • INSERM, Unit 679, 47 Bd de l’Hôpital
    • Institut Fédératif de Recherche en Neurosciences (IFR70), Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6, UMR S679
  • Sylvie Forlani
    • INSERM, Unit 679, 47 Bd de l’Hôpital
    • Institut Fédératif de Recherche en Neurosciences (IFR70), Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6, UMR S679
  • Anne-Louise Leutenegger
    • INSERM, U535
    • Univ. Paris-Sud, IFR69, UMR-S535
  • Isabelle Nelson
    • INSERM, Unit 679, 47 Bd de l’Hôpital
    • Institut Fédératif de Recherche en Neurosciences (IFR70), Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6, UMR S679
  • Marie-Claude Babron
    • INSERM, U535
    • Univ. Paris-Sud, IFR69, UMR-S535
  • Nizar Elleuch
    • INSERM, Unit 679, 47 Bd de l’Hôpital
    • Institut Fédératif de Recherche en Neurosciences (IFR70), Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6, UMR S679
  • Christel Depienne
    • INSERM, Unit 679, 47 Bd de l’Hôpital
    • Institut Fédératif de Recherche en Neurosciences (IFR70), Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6, UMR S679
    • Département de Génétique et CytogénétiqueAP-HP, Groupe Hospitalier Pitié-Salpêtrière
  • Céline Charon
    • Centre National de Génotypage
  • Alexis Brice
    • INSERM, Unit 679, 47 Bd de l’Hôpital
    • Institut Fédératif de Recherche en Neurosciences (IFR70), Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6, UMR S679
    • Département de Génétique et CytogénétiqueAP-HP, Groupe Hospitalier Pitié-Salpêtrière
    • Fédération de NeurologieAP-HP, Groupe Hospitalier Pitié-Salpêtrière
    • Faculté de Médecine, Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6
    • INSERM, Unit 679, 47 Bd de l’Hôpital
    • Institut Fédératif de Recherche en Neurosciences (IFR70), Groupe Hospitalier Pitié-SalpêtrièreUniversité Pierre et Marie Curie, Paris 6, UMR S679
    • Département de Génétique et CytogénétiqueAP-HP, Groupe Hospitalier Pitié-Salpêtrière
Original Investigation

DOI: 10.1007/s00439-007-0396-1

Cite this article as:
Hanein, S., Dürr, A., Ribai, P. et al. Hum Genet (2007) 122: 261. doi:10.1007/s00439-007-0396-1

Abstract

Hereditary spastic paraplegias (HSPs) are genetically and phenotypically heterogeneous. Both “uncomplicated” and “complicated” forms have been described, with autosomal dominant, autosomal recessive, and X-linked inheritance. Hitherto, ten autosomal dominant “uncomplicated” HSP (ADHSP) loci have been mapped. Here, we report linkage of ADHSP with markers of the 8p21.1-q13.3 chromosomal region in a large French family, including 29 examined at-risk individuals. The age at onset varied from 8 to 60 years with a mean of 31.6 ± 16.4 years. Multipoint and two-point LOD-score calculations as well as haplotype reconstruction in this region gave support to the location of this novel ADHSP locus (SPG37) in a 43.5 cM genetic interval flanked by loci D8S1839 and D8S1795. The region was shared by all definitely (n = 13), probably (n = 3) and possibly (n = 2) affected patients with a maximum LOD score of 4.20 at the D8S601 locus. Two candidate genes, encoding the kinesin family member 13B and neuregulin 1 (isoforms SMDF and GFF2), were screened for mutations, but no disease-causing alterations were identified. Interestingly, another region, on chromosome 10q22.3-23.31, was found to segregate in all affected patients (but not in probably or possibly affected subjects) and in a high proportion of healthy at risk individuals, suggesting that this locus might act as a modifier of the phenotype.

Introduction

The hereditary spastic paraplegias (HSPs) are a large group of clinically and genetically heterogeneous inherited neurological disorders, characterized by progressive spasticity in the lower limbs (Harding 1983). Their estimated prevalence varies from 1.27 to 9.6/100,000 according to the geographical area (Polo et al. 1991; Filla et al. 1992; Fink 2006). The disease is characterized pathologically by axonal degeneration in the long descending and ascending tracts of the spinal cord, especially in their terminal portions (Behan and Maia 1974). The corticospinal tracts are mainly affected, but the columns of Goll and the spinocerebellar tracts may also be involved. HSP is classified according to the absence (“uncomplicated” HSP) or presence (“complicated” HSP) of accompanying neurological signs/symptoms, including mental retardation, deafness, cerebellar ataxia, epilepsy, dysarthria, ichthyosis, optic atrophy, peripheral neuropathy, retinitis pigmentosa, cataract, etc., and by the mode of inheritance: autosomal dominant (AD), autosomal recessive (AR) or X-linked (Dürr and Brice 2000; Fortini et al. 2003; Fink 2006).

Each type of HSP, according to its inheritance, is genetically heterogeneous. To date 32 HSP genes have been localized, 14 of which have been identified according to the HUGO and OMIM databases (Fink 2006; Mannan et al. 2006; Valdmanis et al. 2007; Stevanin et al. 2007).

“Uncomplicated” forms are inherited most often as an autosomal dominant trait. However X-linked (SPG16, Steinmuller et al. 1997) and autosomal recessive (SPG5, 24, 28) uncomplicated forms also occur (Hentati et al. 1994; Hodgkinson et al. 2002; Bouslam et al. 2005). Ten loci for autosomal dominant “uncomplicated”/“pure” hereditary spastic paraplegia (ADPHSP) have been mapped, for eight of which the causative genes have been identified (OMIM): SPG3A (MN_606439; 14q11-q21; Atlastin), SPG4 (MN_604277; 2p22; Spastin), SPG6 (MN_600303; 15q11.1; NIPA1), SPG8 (MN_603563; 8q23-q24; KIAA0196), SPG10 (MN_604187; 12q13; KIF5A), SPG12 (MN_604805; 19q13), SPG13 (MN_605280; 2q24-34; Chaperonin 60), SPG19 (MN_607152; 9q33-34), SPG31 (MN_609139; 2p12; REEP1) and SPG33 (MN_610243, 10q24.2; ZFYVE27).

Even though patients with ADPHSP generally have the same core clinical features, namely spastic gait, lower limb spasticity, weakness, hyperreflexia, and extensor plantar responses, there is a considerable variation in age at onset and severity of spasticity, both within and between families (Harding 1981; Polo et al. 1993; Dürr et al. 1994; De Jonghe et al. 1996; Nielsen et al. 1998; Reid et al. 1999; Dürr and Brice 2000; Fortini et al. 2003; Fink 2006). In addition to spasticity, bladder dysfunction and overt or subclinical sensory abnormalities are common, and subclinical cognitive impairment may be present (Reid et al. 1999; Tallaksen et al. 2001).

In this study, we evaluated a large French kindred with ADPHSP and excluded linkage of the disease gene to most known ADPHSP loci. We subsequently performed a genome-wide search for the disease locus and detected tight linkage of the disorder to markers of the 8p21.1-q13.3 chromosomal region.

Patients and methods

Patients

We examined 29 members and their non-affected spouses (n = 8) of a non-consanguineous French kindred with an “uncomplicated” HSP segregating as an autosomal dominant trait (FSP-SAL-028; Fig. 1). Based on the clinical examination, individuals were classified as “definitely affected” (n = 13), “probably affected” (n = 3), “possibly affected” (n = 2) or “unaffected” (n = 11). In order to be as stringent as possible, we deliberately did not use brisk reflexes in the lower limbs as a definite clinical criterion. “Definitely affected” was defined as the presence of at least two of the following criteria: gait spasticity, bilateral extensor plantar reflex and/or perceived handicap reported by the patient. Patients fulfilling only one of these criteria were classified as “probably affected”. Patients with isolated clonus at ankles were classified as “possibly affected”. Age at onset of symptoms was ascertained from patients during the clinical interview.
https://static-content.springer.com/image/art%3A10.1007%2Fs00439-007-0396-1/MediaObjects/439_2007_396_Fig1a_HTML.gif
https://static-content.springer.com/image/art%3A10.1007%2Fs00439-007-0396-1/MediaObjects/439_2007_396_Fig1b_HTML.gif
Fig. 1

Pedigree structure of the FSP-SAL-028 family with haplotype reconstruction for informative markers on chromosome 8p21.1-q13.3 (a) and 10q22.3-23.31 (b). The physical positions of the microsatellite markers are indicated in million base pairs (Mb). The haplotype segregating with the disease is in black. Arrows indicate the position of critical recombination events. Hash indicates the additional markers used for refinement of the interval after the genome scan. Horizontal arrows indicate obligatory recombination events. The SPG37 locus overlaps with the SPG5 and HSP-TCC loci. The10q haplotype overlaps with the SPG27 locus. The code numbers of all sampled (asterisk) individuals are given below the symbols

Genomic DNA was extracted from the blood of each subject participating in the study after informed consent, according to the local Ethics Committee (authorization n°03.12.07 of the Comité Consultatif pour la Protection des Personnes et la Recherche Biomédicale Paris-Necker).

Conventional mutations and rearrangements in the two most frequent forms of ADPHSP (SPG3A, SPG4) had previously been excluded in the index patient (Depienne et al. 2007; Namekawa et al. 2006). In addition, no mutations were found in the SPG17/BSCL2 gene (unpublished data).

Genotyping and linkage analysis

Indirect genetic studies were performed at several of the known ADPHSP loci, SPG3A, SPG6, SPG12 and SPG13, using polymorphic markers (available on request).

Subsequently, a 10-cM genome-wide search for the disease-causing gene was undertaken using fluorescent oligonucleotides flanking the 400 polymorphic markers of the Genescan Linkage Mapping Set, version II (Applied Biosystems, Foster City, California, USA), under conditions recommended by the manufacturer. Amplified fragments were electrophoresed and analysed on an automatic sequencer (ABI 3730, Applied Biosystems). Genotypes were determined with GeneMapper 3.5 software (Applied Biosystems). One hundred and twenty eight additional markers were used to explore several regions.

Pairwise and multipoint linkage analyses were performed using Fastlink 3.0 (Cottingham et al. 1993) and Superlink Online 1.5 (Silberstein et al. 2006). Equal allele frequencies were used for the analysis of the genome scan. Subsequently, for the validation of putative loci, Caucasian allele frequencies from the Human Genome Database (GDB) were used. The disease allele frequency was set at 0.0001 and equal recombination fractions for males and females were given. Since the penetrance of ADHSP is age-dependent, liability classes were used with penetrance set at 0.05 (<10 years); 0.4 (10–20 years); 0.7 (20–40 years); 0.8 (40–60 years) and 0.9 (>60 years) [Penetrance 0.9: all individuals, age-dependent]. Additional calculations were performed using a single penetrance class set at 0.8 [Penetrance 0.8: all individuals, single penetrance class]. We also evaluated linkage by minimizing the weight of unaffected at risk subjects [Affected only like method or AOL], by setting the single penetrance class at 0.001. Patients that did not meet the full clinical criteria for definitely affected status (i.e., possibly affected and probably affected) were considered to be of unknown status in all calculations. Genetic distances were those of the Marshfield Centre for Medical Genetics and map positions were verified on the human genome sequence draft (NCBI, UCSC and Ensembl centres).

Screening for mutations in candidate genes in the 8p21.2-q13.3 region

The genes encoding the kinesin family member 13B (KIF13B, NM_015254; n = 40 exons; physical position at chr8: 28,980,715–29,176,560 bp); the neuregulin 1 (NRG1) isoforms SMDF (NM_013959; n = 3 exons; physical position at chr8: 32,624,286–32,720,285 bp) and GGF2 (NM_013962; n = 5 exons; physical position at chr8: 31,616,810–32,720,310 bp); were screened for mutations using primers designed to flank each exon, including intron-exon boundaries, in two affected members (II-16 and III-21; primers available on request). Amplified products were purified and directly sequenced using the Big Dye Terminator Cycle Sequencing Kit V3.1 on an ABI 3730 automated sequencer (Applied Biosystems).

Results

Clinical features

Twenty-nine patients were examined at a mean age of 39.9 ± 18.5 years (mean ± SEM; range: 14–79 years). Thirteen patients were classified as definitely affected. In this group, age at onset could only be ascertained in 8 patients (mean 31.6 ± 16.4 years; range 8–60 years). Reflexes were increased in the upper limbs (UL) in 8/13 patients and weakness of the iliopsoas was mild to moderate in 9/13. None of the 13 patients in this group required a walking aid. Vibration sense was decreased in 5/13 (38%) and bladder disturbances were also present in 5/13 (38%). Cerebral MRI (patient III-30), CT-scan (patient II-18), electromyography and conduction velocities (patients II-16, II-18, III-27 and III-30) as well as brainstem visual and somatosensory evoked potentials (patients II-18 and III-30) were normal in tested patients.

Three patients were classified as probably affected. In this group, spasticity was absent, but two patients were functionally impaired and in one patient plantar reflexes were extensor. Two other members were classified as possibly affected since they had no spastic gait and plantar reflexes were flexor but both had ankle clonus and very brisk reflexes that could not be considered “normal” (Fig. 1; Table 1).
Table 1

Clinical characteristics of the 29 examined individuals in family FSP-SAL-028

Individual no.

Age on examination

Age at onset

Spasticity

Extensor plantar reflex

Functional disability

Clinical status

Remarks

II-10

74

NA

0

++

+

Definitely affected

Decreased vibration sense at ankles

II-12

72

60

+

0

+++

Definitely affected

Decreased vibration sense at ankles

II-14

79

40

+++

++

+++

Definitely affected

Decreased vibration sense at ankles

II-16

67

45

+++

++

+++

Definitely affected

Decreased vibration sense at ankles, urinary urgencies

II-18

64

30

+++

++

+++

Definitely affected

Decreased vibration sense at ankles, urinary urgencies

III-20

48

NA

0

0

0

Unaffected

 

III-21

39

30

++

++

+++

Definitely affected

Index case

III-23

51

NA

0

0

0

Unaffected

 

III-25

42

NA

0

0

0

Unaffected

Scoliosis at age 13

III-27

39

NA

++

++

++

Definitely affected

 

III-28

38

NA

0

++

0

Probably affected

Death at age 49

III-29

48

NA

0

0

0

Unaffected

 

III-30

18

8

++

++

+++

Definitely affected

Urinary urgency

III-31

19

NA

+

0

+

Definitely affected

 

III-32

26

NA

0

0

0

Unaffected

 

III-33

48

20

+++

++

++

Definitely affected

Urinary urgency

III-42

56

NA

0

0

+

Probably affected

Ankle clonus

III-44

48

NA

0

++

+

Definitely affected

Ankle clonus, falls, urinary urgencies

IV-34

24

NA

0

0

0

Unaffected

 

IV-35

22

NA

0

0

+

Probably affected

Ankle clonus

IV-36

20

NA

+

0

+

Definitely affected

Ankle clonus

IV-46

37

NA

0

0

0

Unaffected

 

IV-47

35

NA

0

0

0

Unaffected

 

IV-48

33

NA

0

0

0

Unaffected

Thigh cramps, lower back pain

IV-50

22

20

+

+

+

Definitely affected

Ankle clonus, spasticity at rest

IV-52

14

NA

0

0

0

Possibly affected

Ankle clonus

IV-54

26

NA

0

0

0

Possibly affected

Ankle clonus, fatigue

IV-55

24

NA

0

0

0

Unaffected

Shoes showed signs of abnormal wear

IV-56

22

NA

0

0

0

Unaffected

 

0 = absent; + = mild; ++ = moderate; +++ severe. Disability was assessed as mild (slight stiffness while walking, but running possible), moderate (moderate gait stiffness, running difficult or impossible) severe (marked gait stiffness, unable to walk without help)

NA data not available

Genetic study

Prior to the genome scan, all tested ADPHSP loci (SPG3A, SPG6, SPG12 and SPG13) were excluded in family FSP-SAL-028 on the basis of negative LOD scores below the threshold of −2 (data not shown). In addition, mutations in SPG3A, SPG4 and SPG17 were previously excluded (Depienne et al. 2007; Namekawa et al. 2006).

A genome-wide search using 400 microsatellite markers was subsequently undertaken to map the disease gene in family FSP-SAL-028. One hundred additional polymorphic markers were used to explore uninformative or insufficiently covered regions. When only affected individuals were considered for LOD score calculations (AOL), positive and significant values were only reached in two genomic regions on chromosome 8p and 10q. Additional markers of these regions, chosen from the Genethon Linkage Map and from the human genome sequence draft, were then analysed. Positive LOD score values were obtained for 19 and 9 successive markers located on chromosome 8p21.2-q13.3 and 10q22.3-23.31, respectively (Fig. 1a, b).

On chromosome 8, maximum pairwise LOD scores of Zmax = 3.10 at loci D8S1820 (AFM051zb1) and D8S1718 (AFMa191ye1), and 3.02 at D8S1803 (AFM312xf1) were obtained at θ = 0 using definitely affected individuals only (Table 2, AOL). Linkage was confirmed when non-affected carriers were taken into account. Two-point LOD scores >3 at θ = 0 were obtained for seven loci, namely, D8S1820 (AFMa051zb1); D8S1769 (AFM220yf8); D8S1810 (AFM345td5); D8S1803 (AFM312xf1); D8S1722 (AFMa219zg9); D8S601 (AFM276xe9) and D8S1718 (AFMa191ye1) using penetrance classes with a maximal penetrance set at 0.9 (Table 2). When a single penetrance was used and set at 0.8, the two-point LOD scores increased for loci D8S1820 (AFMa051zb1; Zmax = 4.28 at θ = 0), D8S1769 (AFM220yf8; Zmax = 4.14 at θ = 0); D8S1810 (AFM345td5; Zmax = 4.46 at θ = 0); D8S1803 (AFM312xf1; Zmax = 4.06 at θ = 0) and D8S1722 (AFMa219zg9; Zmax = 4.67 at θ = 0). Similar results were obtained using Caucasian allele frequencies (GDB) or equal allele frequencies. Multipoint linkage analysis in this region produced LOD scores >3 for all markers between D8S1769 and D8S1767 with a maximal value of 4.20 at the D8S601 locus when the penetrance was set at 0.90 (Fig. 2a). Haplotype reconstruction showed that all definitely affected patients, and all those classified as probably and possibly affected, carried a common chromosomal region on 8p21.1-q13.3. Obligatory recombination events were observed in 3/13 of the definitely affected patients (individuals II-12, III-27 and III-44), indicating mapping of the disease gene in a 43.5 cM genetic interval flanked by the D8S1839 (AFMb320va5) and D8S1795 (AFM281wb1) loci, respectively (Fig. 1a). Marker D8S255 would be considered as the centromeric boundary of the interval by taking into account the recombination observed in the probably affected individual III-28.
Table 2

Results of two-point linkage analysis for microsatellite polymorphisms on chromosome 8p21.2-q13.3 according to the penetrance set at 0.9 (all individuals, age-dependent) or 0.8 (all individuals, single penetrance class) or using the affected only like approach (AOL)

Markers

Mb

Penetrance

Two-point LOD score at Θ = 

0.0

0.01

0.05

0.1

0.2

0.3

0.4

D8S1771 (AFMb320va5)

25.4

0.9

−2.35

−2.10

−1.08

−0.57

−0.13

0.03

0.06

0.8

−2.30

−2.08

−1.07

−0.54

−0.09

0.07

0.09

AOL

−1.72

−1.61

−0.85

−0.40

−0.05

0.06

0.06

D8S1839 (AFMb320va5)

27.4

0.9

−0.45

−0.19

0.75

1.11

1.19

0.91

0.43

0.8

−0.60

−0.33

0.65

1.05

1.16

0.91

0.44

AOL

−0.69

−0.55

−0.24

0.60

0.74

0.58

0.27

D8S1820 (AFMa051zb1)

28

0.9

3.90

3.83

3.54

3.16

2.38

1.54

0.67

0.8

4.28

4.20

3.88

3.48

2.63

1.73

0.77

AOL

3.10

3.05

2.82

2.52

1.87

1.17

0.45

D8S1769 (AFM220yf8)

31.2

0.9

4.05

3.99

3.73

3.38

2.59

1.70

0.72

0.8

4.14

4.08

3.85

3.52

2.74

1.82

0.79

AOL

2.70

2.64

2.43

2.15

1.58

0.98

0.39

D8S1810 (AFM345td5)

31.8

0.9

4.09

4.01

3.71

3.32

2.49

1.62

0.70

0.8

4.46

4.38

4.05

3.63

2.74

1.80

0.80

AOL

2.70

2.64

2.43

2.15

1.58

0.98

0.39

D8S283 (AFM238yh12)

33.7

0.9

2.70

2.66

2.46

2.21

1.66

1.07

0.43

0.8

2.67

2.63

2.46

2.23

1.69

1.07

0.42

AOL

1.85

1.81

1.65

1.45

1.05

0.63

0.24

D8S1750 (AFMb036yd1)

35.5

0.9

1.55

1.53

1.42

1.27

0.92

0.55

0.23

0.8

1.65

1.63

1.55

1.41

1.07

0.67

0.30

AOL

1.49

1.46

1.32

1.15

0.79

0.46

0.19

D8S1803 (AFM312xf1)

36.4

0.9

3.97

3.91

3.65

3.30

2.53

1.65

0.70

0.8

4.06

4.01

3.78

3.45

2.68

1.78

0.78

AOL

3.02

2.97

2.74

2.45

1.81

1.13

0.43

D8S1722 (AFMa219zg9)

37.7

0.9

4.30

4.22

3.91

3.50

2.63

1.70

0.72

0.8

4.67

4.59

4.25

3.82

2.89

1.88

0.82

AOL

2.91

2.85

2.63

2.34

1.72

1.07

0.41

D8S255 (AFM023xc1)

40

0.9

1.07

1.05

0.99

0.89

0.64

0.37

0.14

0.8

0.91

0.92

0.92

0.87

0.68

0.42

0.17

AOL

1.20

1.17

1.04

0.89

0.58

0.30

0.11

D8S532 (AFM081yd11)

40.9

0.9

2.97

2.93

2.73

2.46

1.87

1.21

0.50

0.8

2.69

2.65

2.50

2.28

1.76

1.15

0.47

AOL

2.32

2.27

2.08

1.83

1.33

0.81

0.30

D8S601 (AFM276xe9)

53.8

0.9

3.19

3.14

2.93

2.68

2.07

1.37

0.59

0.8

3.16

3.12

2.93

2.49

1.92

1.26

0.53

AOL

2.34

2.30

2.10

1.86

1.35

0.83

0.31

D8S1763 (AFMb307xb9)

60.2

0.9

1.77

1.75

1.66

1.51

1.16

0.73

0.29

0.8

1.83

1.82

1.75

1.63

1.28

0.84

0.35

AOL

1.71

1.67

1.52

1.32

0.93

0.53

0.19

D8S1177 (AFMa156yc5)

61.4

0.9

2.28

2.25

2.10

1.90

1.46

0.95

0.39

0.8

2.53

2.49

2.33

2.11

1.62

1.07

0.45

AOL

1.68

1.64

1.51

1.33

0.97

0.59

0.21

D8S1718 (AFMa191ye1)

62.1

0.9

3.35

3.31

3.12

2.86

2.23

1.46

0.62

0.8

3.44

3.41

3.25

3.00

2.38

1.59

0.69

AOL

3.10

3.04

2.81

2.51

1.87

1.17

0.47

D8S1696 (AFMa108wb5)

64

0.9

1.78

1.75

1.62

1.45

1.06

0.64

0.22

0.8

1.77

1.74

1.62

1.44

1.06

0.64

0.22

AOL

1.36

1.33

1.20

1.03

0.70

0.38

0.11

D8S544 (AFM294vd5)

65.7

0.9

2.46

2.42

2.25

2.02

1.53

0.98

0.39

0.8

2.70

2.66

2.48

2.23

1.70

1.10

0.46

AOL

1.85

1.81

1.66

1.45

1.04

0.62

0.22

D8S1841 (AFMa084wc5)

66.2

0.9

2.82

2.79

2.63

2.40

1.87

1.24

0.54

0.8

2.91

2.89

2.75

2.55

2.02

1.36

0.61

AOL

2.57

2.52

2.32

2.06

1.51

0.95

0.39

D8S553 (AFM326te5)

67

0.9

2.83

2.79

2.63

2.41

1.87

1.24

0.54

0.8

2.92

2.89

2.76

2.55

2.02

1.37

0.61

AOL

2.58

2.53

2.32

2.06

1.52

0.95

0.39

D8S1797 (AFMc031wc5)

67.4

0.9

2.89

2.85

2.69

2.45

1.90

1.26

0.55

0.8

2.97

2.95

2.81

2.60

2.05

1.38

0.62

AOL

2.63

2.58

2.37

2.11

1.55

0.97

0.39

D8S1767 (AFMb311yf9)

68.8

0.9

2.48

2.46

2.36

2.19

1.73

1.14

0.48

0.8

2.57

2.56

2.49

2.34

1.88

1.26

0.55

AOL

2.93

2.87

2.65

2.36

1.74

1.08

0.42

D8S1795 (AFM281wb1)

70.9

0.9

−3.80

−0.73

−0.05

−0.20

0.37

0.34

0.20

0.8

−3.42

−0.64

−0.03

0.28

0.44

0.40

0.23

AOL

−1.23

−0.33

0.22

0.39

0.45

0.36

0.19

D8S530 (AFM259yg5)

72.7

0.9

−5.14

−0.76

−0.07

−0.19

0.33

0.27

0.13

0.8

−5.03

−1.00

−0.26

0.05

0.25

0.23

0.11

AOL

−1.37

0.19

0.72

0.81

0.69

0.46

0.20

Θ = recombination fractions. Figures in bold are the maximal lod scores for linked markers. Physical positions of microsatellite markers are indicated in million base pairs (Mb)

https://static-content.springer.com/image/art%3A10.1007%2Fs00439-007-0396-1/MediaObjects/439_2007_396_Fig2_HTML.gif
Fig. 2

Multipoint linkage analysis for microsatellite polymorphisms on chromosome 8p21.1-q13.3 (a) and 10q22.3- 23.31 (b) in FSP-SAL-028 family, with the penetrance set at 0.9 (Similar results were obtained when the penetrance was set at 0.8)

At the other possible location on chromosome 10q, a maximal pairwise LOD score of 3.18 was only reached for marker D10S1686(AFM191va9) using only definitely affected patients. When the genetic penetrance was set at 0.9 and 0.8, evidence of linkage decreased as the result of an excess number of non-affected carriers: the maximum pairwise LOD scores at θ = 0 dropped to 1.90 and 1.21, respectively (Table 3). Some of the adjacent markers even dropped below zero. Of note, none of the probably or possibly affected individuals but seven of unaffected ones carried the potential disease associated haplotype (Fig. 1b). The maximal multipoint LOD score only reached the value of +2 in this region (Fig. 2b).
Table 3

Results of two-point linkage analysis for microsatellite polymorphisms on chromosome 10q22.3-23.31 according to the penetrance set at 0.9 (all individuals, age-dependent) or 0.80 (all individuals, single penetrance class) or using the affected only like approach (AOL)

Markers

Mb

Penetrance

Two-point LOD score at Θ = 

0.0

0.01

0.05

0.1

0.2

0.3

0.4

D10S605 (AFMa120xc5)

78.9

0.9

−6.41

−1.64

−0.29

0.21

0.50

0.44

0.23

0.8

−7.02

−2.38

−0.91

−0.29

0.19

0.26

0.15

AOL

−4.23

−0.73

0.46

0.80

0.85

0.60

0.24

D10S201 (AFM155zc3)

80.7

0.9

−1.75

0.25

0.92

1.13

1.10

0.80

0.36

0.8

−2.44

−0.41

0.38

0.70

0.86

0.67

0.31

AOL

−0.69

1.26

1.74

1.76

1.45

0.96

0.40

D10S1786 (AFMa070xe9)

83.9

0.9

1.40

1.39

1.35

1.26

0.99

0.63

0.24

0.8

0.64

0.67

0.74

0.78

0.69

0.47

0.18

AOL

2.35

2.31

2.11

1.87

1.36

0.83

0.31

D10S246 (Humrpt38A)

84.1

0.9

1.89

1.89

1.85

1.75

1.41

0.94

0.39

0.8

1.21

1.24

1.31

1.33

1.17

0.81

0.34

AOL

2.95

2.90

2.67

2.38

1.76

1.10

0.43

D10S532 (AFM192xa5)

85.4

0.9

1.00

1.00

0.98

0.92

0.73

0.46

0.17

0.8

0.82

0.83

0.83

0.80

0.65

0.42

0.15

AOL

1.80

1.76

1.61

1.41

1.02

0.61

0.21

D10S1686 (FM191va9)

85.5

0.9

1.90

1.90

1.88

1.79

1.47

0.98

0.39

0.8

0.74

0.79

0.94

1.03

0.98

0.70

0.27

AOL

3.18

3.13

2.92

2.60

1.94

1.22

0.48

D10S1769 (AFM353ta5)

86.9

0.9

0.33

0.33

0.33

0.33

0.29

0.22

0.13

0.8

−0.66

−0.61

−0.46

−0.32

−0.12

−0.01

0.03

AOL

0.80

0.79

0.72

0.63

0.47

0.32

0.16

D10S1744 (AFM063xb10)

88.3

0.9

0.44

0.45

0.48

0.48

0.43

0.30

0.15

0.8

−0.55

−0.49

−0.32

−0.16

0.02

0.07

0.05

AOL

0.92

0.91

0.86

0.79

0.61

0.41

0.19

D10S1765 (AFM337xf9)

89.6

0.9

1.35

1.36

1.36

1.31

1.08

0.73

0.31

0.8

0.19

0.24

0.41

0.55

0.60

0.45

0.19

AOL

2.63

2.58

2.38

2.11

1.56

0.98

0.40

D10S1735 (AFM269xf9)

90.6

0.9

−0.11

−0.08

0.02

0.10

0.15

0.13

0.07

0.8

−1.10

−1.02

−0.77

−0.54

−0.25

−0.10

−0.03

AOL

0.37

0.38

0.41

0.41

0.34

0.23

0.11

D10S1739 (AFMb362yg5)

90.8

0.9

−0.37

−0.35

−0.30

−0.24

−0.16

−0.09

−0.04

0.8

−1.53

−1.47

−1.24

−1.00

−0.64

−0.37

−0.16

AOL

0.62

0.61

0.54

0.47

0.32

0.20

0.09

D10S1753 (AFM287ze1)

92.4

0.9

−4.45

−1.44

−0.65

−0.29

0.01

0.08

0.04

0.8

−4.77

−2.05

−1.16

−0.70

−0.25

−0.07

−0.02

AOL

−1.45

0.06

0.60

0.72

0.63

0.42

0.17

D10S564 (AFM029xh12)

92.6

0.9

−3.81

−0.81

−0.06

0.24

0.41

0.32

0.12

0.8

−4.84

−2.09

−1.15

−0.63

0.15

0.01

0.01

AOL

−0.87

0.63

1.13

1.19

0.97

0.60

0.22

θ = recombination fractions. Figures in bold are the maximal lod scores for linked markers. Physical positions of microsatellite markers are indicated in million base pairs (Mb)

Candidate gene analysis

Exons and intron-exon boundaries of both known genes, KIF13B and NRG1 (SMDF and GGF2 isoforms), were screened for mutations. No disease-causing alteration was found in two affected patients of family FSP-SAL-028. Only already reported single nucleotide polymorphisms (Human Genome Project working draft at UCSC) were identified in this family (not shown; available on request).

Discussion

The present study establishes the existence of a new locus for autosomal dominant “uncomplicated” HSP, in a 43.5 cM region on chromosome 8p21.1-q13.3, between loci D8S1839 (AFMb320va5) and D8S1795 (AFM281wb1). We propose that the chromosome 8p21.1-q13.3 locus for ADPHSP be designated “SPG37” in accordance with HUGO nomenclature. This adds a new locus to the ten loci for ADPHSP already identified (for review, see Fink 2006).

As in SPG4 (Santorelli et al. 2000; Svenson et al. 2001; Fonknechten et al. 2000), age at onset varied widely (8–60 years). Interestingly, many definitely affected individuals were not able to assess their age at onset (5/13), probably because of the slowly progressive nature of the disease and the absence of other neurological signs.

In family FSP-SAL-028, the whole genome linkage analysis was performed using a panel of 500 polymorphic markers. When affected individuals were considered alone, only two different chromosomal regions on chromosome 8p21.1-q13.3 and 10q22.3-23.31 were identified with positive pairwise LOD score values ranging from 3.10 to 3.18 (Tables 2, 3).

It is now well established that the penetrance of autosomal dominant “uncomplicated” HSP is incomplete and is in any case age dependent (Hedera et al. 1999; Fonknechten et al. 2000; Fontaine et al. 2000; Dürr et al. 2004; D’Amico et al. 2004; Fink 2006). Based on this assumption, markers on chromosome 8p21.1-q13.3 generated pairwise and multipoint LOD scores above +3 with a maximum multipoint LOD score of +4.20 (Table 2; Fig. 2) while markers of the 10q22.3-23.31 region produced pairwise and multipoint LOD scores that barely exceeded the value of +2, using 0.9 and 0.8 maximal penetrance (Table 3; Fig. 2b). These data strongly support the location of the disease gene on chromosome 8p21.1-q13.3 rather than on chromosome 10q, by odds of at least 1:100. This hypothesis was reinforced by haplotype reconstructions at both chromosomal regions (Fig. 1a, b). With regard to the 10q22.3-23.31 region, the putative disease haplotype was carried by 20 individuals (13 affected patients, 7 unaffected, Fig. 1a). None of the “probably” or “possibly” affected individuals carried the putative disease haplotype, leading to a calculated penetrance of 0.63 if they were considered as unaffected. This penetrance is much lower than the accepted range of 0.80–0.90 (Hedera et al. 1999; Fonknechten et al. 2000; Fontaine et al. 2000; Dürr et al. 2004; D’Amico et al. 2004; Fink 2006). Conversely, on chromosome 8, the putative disease haplotype was shared by also 20/29 family members (between the D8S1839 and D8S255 loci) that included all definitely, probably and possibly affected patients as well as two unaffected subjects only, both in the youngest generations (Fig. 1a). Considering the stringency of our clinical criteria used to define the clinical status, if possibly and probably affected subjects were considered as affected, penetrance would be 0.89, a value closer to the accepted 0.80–0.90 values. This strongly supports the location of a new locus for autosomal dominant “uncomplicated” spastic paraplegia on chromosome 8p21.1-q13.3.

Interestingly, the gene for autosomal recessive HSP with thin corpus callosum and epilepsy has recently been mapped to chromosome 8p12-p11.21 (HSP-TCC; Al-Yahyaee et al. 2006). This locus and SPG37 overlap in the 13 cM region flanked by the D8S1820 (AFMa051zb1; physical position at chr8: 28 Mb) and D8S532 (AFM081yd11; physical position at chr8: 40.9 Mb) loci, respectively. Along the same lines, the gene for autosomal recessive HSP, SPG5, maps to chromosome 8q11.22-q13.2 (Wilkinson et al. 2003) but does not overlap with the novel HSP-TCC locus. Conversely, the SPG5 and SPG37 loci share a 20 cM region flanked by the D8S589 (GATA12H01; physical position at chr8: 51.08 Mb) and D8S1795 (AFM281wb1; physical position at chr8: 70.9 Mb) loci, respectively (Fig. 1a). Finally, the 10q22.3-23.31 region overlaps SPG27 (Meijer et al. 2004; Ribai et al. 2006) in the 6.9 cM region flanked by the D10S1786 (AFMa070xe9; physical position at chr10: 83.9 Mb) and D10S1739 (AFMb362yg5; physical position at chr10: 90.8 Mb) loci, respectively (Fig. 1b). Though different in terms of their clinical presentation or their mode of transmission, the four phenotypes share pyramidal signs with spasticity. The mapping data reported here cannot exclude allelism between SPG37 and SPG5 or the locus for HSP-TCC.

Interestingly, the disease haplotype on chromosome 8 was found in all affected, three probably affected (patients III-42; III-28 and IV-35) and two possibly affected individuals (IV-52 and IV-54), as well as in two unaffected subjects (IV-47 and IV-48), whereas the linked haplotype on chromosome 10q was carried by all affected patients, as well as by seven unaffected family members (IV-47 and IV-48 excluded) but not the five probably and possibly affected individuals (Fig. 1a, b). It can therefore be postulated that the locus on chromosome 10q might contain a modifier gene. In this case, the combination of a mutation on 8p21.1-q13.3 and a polymorphism on 10q22.3-23.31 would lead to a more severe phenotype than the mutation on 8p21.1-q13.3 alone. The modifier gene on 10q might be the gene responsible for SPG27. To test this hypothesis we compared the likelihood of an effect of both chromosomes versus an effect of chromosome 8 only, using TLinkage, a linkage package allowing for 2 trait loci (Lathrop and Ott 1990). For this analysis, having the at-risk genotypes on chromosome 10q was set to increase the probability of an individual to be definitely affected (see Supplementary material). We found significant evidence (P < 0.013) for a modifier gene on 10q22.3-23.31 only when its presence, in addition to chromosome 8, gave a more than two-fold increase in the risk of being definitely affected compared to having 8p21.1-q13.3 only. This result was dependant on the risk allele frequencies, however. If a rare risk allele on chromosome 8 seems a valid hypothesis, the frequency of the risk allele for the modifier gene on chromosome 10 may not be so low. The higher the modifier gene risk allele frequency, the lower the increase in the risk of being definitely affected need be.

The SPG37 genetic interval contains 146 known genes and 25 predicted genes (Human Genome Project working draft at UCSC). This candidate region could be strongly reduced between the D8S1839 and D8S255 loci considering the recombination event in probably affected patient III-28 (Fig. 1a). In addition, the critical interval defined by the D8S1839 (AFMb320va5; 27.4 Mb) and D8S255 (AFM023xc1; 40 Mb) loci does not overlap with the SPG5 locus (Fig. 1a). In this interval that overlaps with the HSP-TCC locus recently reported, 71 genes have been assigned.

It is worth noting that all HSP genes hitherto identified are involved in four different physiological pathways, suggesting that diverse biochemical abnormalities may be responsible for axon degeneration in various genetic HSP types (Fink 2006). These abnormalities encompass disturbance in Golgi function, dysmyelination, mitochondrial abnormality and axonal transport dysfunction. Several genes in this new ADPHSP locus could thus be considered attractive candidates for HSP. Two of them encoding for neuregulin 1 (NRG1) isoform SMDF and kinesin heavy chain 13B (KIF13B) are strong functional candidates for HSP according to their homology with genes previously implicated in HSP, such as KIF5A (NM_602821), a microtubule motor of axonal transport, implicated in SPG10 (Reid et al. 2002). The KIF13B gene, ubiquitously expressed, may be involved in reorganization of the cortical cytoskeleton and may be functionally important for the intracellular trafficking of MAGUK (membrane-associated GUK) proteins and associated protein complexes (Ishikawa et al. 1998; Hanada et al. 2000). In addition, it is known that a wide variety of different isoforms are produced from the NRG1 gene by alternative splicing. These isoforms include heregulins (HRGs), glial growth factors (GGFs) and sensory and motor neuron-derived factor (SMDF). They are tissue-specifically expressed and differ significantly in their structure. The isoform SMDF may play a role in motor and sensory neuron development and is expressed in nervous system, spinal cord motor neurons, dorsal root ganglion neurons, and brain (Meyer et al. 1997).

These two genes were first considered in view of their functional characteristics but the screening of the 40 exons of KIF13B and the 8 exons of isoforms SMDF and GGF2 of NRG1 in family FSP-SAL-028 failed to identify any mutation. The fact that no mutation was identified in any of these candidate genes does not mean that none of them is involved in the disease, however. The disease-causing mutation could lie in unscreened regions of these genes, such as promoter regions (i.e. PAX6, Okladnova et al. 1998; Sander et al. 1999; Zheng et al. 2001), intronic sequences (NPHP6, NM_610142, den Hollander et al. 2006), 3′ untranslated regions (e.g. alpha-globin, Waggoner et al. 2003) or supplementary undescribed exons (Gerber et al. 2001; 2002). Along the same lines, it is worth noting that standard methods for genetic analyses do not allow detection of heterozygous rearrangements. For example, exon deletions in SPG4 are almost as frequent as point mutations (Depienne et al. 2007).

The mapping of a novel ADPHSP locus further demonstrates the extensive genetic heterogeneity of this condition. This outstanding genetic and physiological diversity, which could greatly increase in the coming years, already hinders the molecular diagnosis in HSP patients. Nevertheless, elucidation of the full extent of genetic heterogeneity in ADPHSP will make it easier to identify ADPHSP genes and will also allow more accurate genetic counselling of affected families.

Databases

UCSC Genome Browser: http://www.genome.ucsc.edu/

Ensembl Genome Browser: http://www.ensembl.org

OMIM—Online Mendelian Inheritance in Man: http://www.ncbi.nlm.nih.gov/omim/

GDB—Genome Data Base: http://www.gdb.org

Superlink—Online: http://www.bioinfo.cs.technion.ac.il/superlink-online/index.shtml

HUGO—Human Genome Organisation: http://www.gene.ucl.ac.uk/nomenclature/

Acknowledgments

We thank Elodie Denis, Estelle Fedirko, Claire-Sophie Rime-Davoine and Drs B. Fontaine, C. Paternotte, E. Reid and J. Hazan for their help and Drs C. Goizet and S. Cogilnicean for clinical examinations. We are grateful to N. Barton and Drs M. Ruberg and J.M. Rozet for critical reading of the manuscript. This study was supported financially by the VERUM foundation (Germany) and the Programme Hospitalier de Recherche Clinique (France).

Supplementary material

439_2007_396_MOESM1_ESM.doc (24 kb)
Supplementary Material: Use of TLinkage (DOC 25 kb).

Copyright information

© Springer-Verlag 2007