Brain, Vol. 122, No. 9, 1741-1755,
September 1999
© 1999 Oxford University Press
Locus–phenotype correlations in autosomal dominant pure hereditary spastic paraplegia
A clinical and molecular genetic study of 28 United Kingdom families
1 Department of Medical Genetics, University of Cambridge, Cambridge and 2 Institute of Medical Genetics, University of Wales, Cardiff, UK
Correspondence to:
D. C. Rubinsztein, Cambridge Institute for Medical Research, Wellcome/MRC Building, Addenbrooke's Hospital, Hills Road, Cambridge CB2 2XY, UK E-mail: dcr1000{at}cam.cus.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
This study aimed to describe the clinical phenotype of a large collection of families with autosomal dominant pure hereditary spastic paraplegia (ADPHSP), to examine the relative frequency of each of the three known ADPHSP genes within this population, to assess locus–phenotype correlation in ADPHSP and to ascertain whether there are clinical subgroups within genetically defined populations of ADPHSP families. We examined 306 family members, 144 affected, from 28 families with ADPHSP. Linkage analysis at the three known ADPHSP loci allowed us to categorize the families into three groups: (i) those families showing linkage to the chromosome 2 ADPHSP locus (seven families); (ii) those in which linkage to all known loci was excluded (five families); and (iii) those in which linkage results were inconclusive. There was a correlation between linkage group and clinical features, with chromosome 2-linked families having a later age at onset of symptoms (P = 0.001) and later age before commencing walking stick use (P = 0.007) than those families in which linkage to all known ADPHSP loci was excluded. There were no clinical differences between the families showing linkage to the chromosome 2 locus, but there were clinical differences between the families in which linkage to all of the known loci had been excluded (P < 0.0001). We conclude that the chromosome 2 ADPHSP gene is a frequent cause of ADPHSP in UK families, that the responsible gene has not yet been mapped in a significant proportion of families and that certain clinical features of ADPHSP, including age at onset, are at least in part determined by genetic locus.
autosomal dominant pure hereditary spastic paraplegia; locus–phenotype correlation; linkage
ADPHSP = autosomal dominant pure hereditary spastic paraplegia
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The hereditary spastic paraplegias are a group of conditions which are characterized by progressive spasticity, predominantly affecting the legs (Harding, 1984
; Fink et al., 1996; Kobayashi et al., 1996a; Reid, 1997). They conventionally are divided into two types, pure and complicated, depending on the presence of additional neurological or non-neurological features (Harding, 1984; Fink et al., 1996; Kobayashi et al., 1996a; Reid, 1997). Autosomal recessive, autosomal dominant and X-linked recessive inheritance patterns have been described for each type (Harding, 1984; Fink et al., 1996; Kobayashi et al., 1996a; Reid, 1997).
Pure hereditary spastic paraplegia is genetically heterogeneous. It is most commonly inherited in an autosomal dominant pattern (Harding, 1981; Polo et al., 1993), with loci mapped on chromosomes 2p (SPG4) (Hazan et al., 1994; Hentati et al., 1994a), 14q (SPG3) (Hazan et al., 1993) and 15q (SPG6) (Fink et al., 1995). In the majority of dominant families showing linkage, the responsible gene maps to the chromosome 2 locus, with small numbers of families showing linkage to the chromosome 14 and chromosome 15 loci (Fink et al., 1996). Some families are not linked to any of these loci, strongly suggesting the presence of at least one further autosomal dominant pure hereditary spastic paraplegia (ADPHSP) gene (Kobayashi et al., 1996b; Bruyn et al., 1997). There are two loci for autosomal recessive PHSP. One has been mapped on chromosome 8 (SPG5) (Hentati et al., 1994b). The second recessive gene (SPG7), at chromosome 16q24, is a nuclear-encoded mitochondrial metalloprotease gene (Casari et al., 1998). Mutations in the proteolipid protein gene (PLP) on Xq21–22 have been found in families with X-linked PHSP (Cambi et al., 1996).
The principal clinical feature of ADPHSP is a slowly progressive spastic gait abnormality, accompanied by lower limb hypertonicity, pyramidal weakness, hyperreflexia and extensor plantar responses (Harding, 1981; Polo et al., 1993; Dürr et al., 1994; De Jonghe et al., 1996; Nielsen et al., 1998). There is considerable variation in age of 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). The disease process is not entirely restricted to the motor system. Many patients experience bladder dysfunction (Harding, 1981; Polo et al., 1993; Dürr et al., 1994), clinical and subclinical sensory abnormalities are common (Harding, 1981; Schady and Sheard, 1990; Polo et al., 1993; Dürr et al., 1994) and cognitive impairment may be present (Tedeschi et al., 1991; Webb and Hutchinson, 1998).
No definite correlation between genetic locus and clinical phenotype has been described for ADPHSP, although a correlation between genetic locus and age at onset has been suggested. Members of chromosome 2-linked families do not have a consistent age of onset (Hazan et al., 1994; Hentati et al., 1994a; Bürger et al., 1996; De Jonghe et al., 1996; Dürr et al., 1996; Raskind et al., 1997; Nielsen et al., 1998). Although the small numbers of chromosome 14 families described make comparison difficult, chromosome 14-linked families appear to have a younger average age of onset than chromosome 2-linked families (Hazan et al., 1993; Hentati et al., 1994a; Gispert et al., 1995; Huang et al., 1997). The single chromosome 15-linked family described had a unimodal age of onset (mean 22 years) and severe disease, with 30% of family members being chairbound (Fink et al., 1995). Few data have been published on the phenotype of families where linkage to the three known loci has been excluded.
Apart from locus heterogeneity, variations in age at onset could be caused by allelic heterogeneity at each individual locus. If this were the case, differences in age of onset might be expected between different families from the same linkage group. Despite the considerable variation in age of onset within chromosome 2-linked families, two separate studies involving a total of 17 chromosome 2-linked families have found no significant difference in age of onset between families (Dürr et al., 1996; Nielsen et al., 1998). There are no reported data on differences in age of onset between chromosome 14-linked families, or between families where linkage to all three known loci has been excluded.
In this study, we present a clinical and genetic analysis of 28 families with ADPHSP. Seven of the families are likely to be linked to the chromosome 2 ADPHSP locus, whereas in five of the families we excluded linkage to all of the known ADPHSP loci. We describe and compare the clinical features of each of these two linkage groups and search for evidence of clinical heterogeneity within each group.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Clinical ascertainment
Twenty-eight families were ascertained as part of a UK-wide clinical and genetic study of ADPHSP (Fig. 1
). All of the families had been referred to regional clinical genetics departments, and in all the pattern of transmission was consistent with autosomal dominant inheritance. After gaining informed consent, all available affected and apparently unaffected family members were assessed neurologically by a single physician (E.R.), using a standard protocol. Diagnostic criteria for ADPHSP were based on those of Harding (Harding, 1981, 1984). Individuals were classified as being affected if they had lower limb hyperreflexia in addition to at least one of the following: progressive spastic gait abnormality, bilateral extensor plantar reflex, bilateral sustained (
5 beats) ankle or knee clonus. Exclusion criteria were: (i) cerebellar ataxia greater than mild upper limb inco-ordination; (ii) overt dementia; (iii) mental retardation; (iv) severe amyotrophy; (v) ophthalmoplegia, optic neuropathy or retinal pigment degeneration; and (vi) extrapyramidal signs. Mild upper limb inco-ordination, mild distal amyotrophy, sphincter symptoms and sensory examination findings were not exclusion criteria. Subjects were classified as affected if they met the above diagnostic criteria. They were deemed to have `suggestive signs' if they had bilateral lower limb hyperreflexia and a unilateral extensor plantar reflex or unilateral clonus. They were classified as being possibly affected if lower limb hyperreflexia was present without other abnormal signs, and as being normal if they had an entirely normal neurological examination. Disease severity was graded in affected individuals using a disability scoring system devised for this study (Table 1). Ethical approval for the study was granted by the Addenbrooke's Hospital ethical committee.
|
|
DNA analysis
DNA samples were isolated from blood of all consenting individuals taking part in the study. In addition, DNA samples were available from a further 10 affected family members. Four of these subjects were obligate gene carriers, and the affected status of the remainder was confirmed by review of hospital records. DNA samples were also available from 43 unrelated spouses. For each ADPHSP locus, potentially informative subjects from all of the families except 15, 16, 19 and 20, which were too small to be useful for linkage, were genotyped by polymerase chain reaction (PCR) amplification at three microsatellite markers, two of which flanked each candidate region. The markers used were D2S400, D2S352, D2S367, D14S288, D14S269, D14S281, D15S128, D15S122 and D15S156. Additional markers D2S2255, D2S2283, D2S2203, D2S2351, D2S2325, D2S2347, D14S1053, D14S1013 and D14S276 were genotyped in certain families. Primer sequences for all of the markers are available from the Généthon microsatellite linkage map (Dib et al., 1996
). Approximately 50–200 ng of DNA was amplified in a 10 µl reaction volume containing 10 mM Tris–HCl pH 8, 50 mM KCl, 50 ng of each primer and 0.5 U of Taq polymerase (Gibco-BRL, Paisley, UK). MgCl2 concentrations were optimized for each marker, and were between 1.0 and 2.0 mM. Reaction product was radiolabelled either by incorporation of [
-32P]dCTP or by end labelling primers with [
-32P]dATP. The standard thermocycling profile consisted of an initial denaturation of 4.5 min at 94°C, followed by 30 cycles with denaturation at 94°C for 30 s, annealing for 30 s at a temperature optimized for each marker, synthesis at 72°C for 30 s, followed by a final extension step of 72°C for 5 min. Alleles were resolved in 6% denaturing polyacrylamide gels.
Linkage analysis
Two-point linkage analysis was carried out using the MLINK program of the FASTLINK version 3.0P package. Multipoint linkage analysis was carried out using the LINKMAP program of the FASTLINK package, or with the program Vitesse, on each family individually (Lathrop and Lalouel, 1984; Cottingham et al., 1993; Schaffer et al., 1994; O'Connell and Weeks, 1995). For linkage examination, subjects' phenotypes were scored as affected if they met the diagnostic criteria above, `unknown' if they had been classified as being possibly affected or as having suggestive signs, or as being `at risk' if clinically normal.
Two strategies were used for linkage analysis. Because of the age-dependent penetrance of ADPHSP, the disease status of young, unaffected family members is uncertain, and so two-point and multipoint linkage analyses were undertaken initially on an `affecteds-only' basis. With this type of analysis, phenotypic data are included on only affected individuals, with all other family members being classified as phenotypically `unknown'. Genotypic data are included for all individuals. Exclusion of linkage was declared if the multipoint lod score throughout an entire candidate region was less than –2. As a second strategy, we performed linkage analysis by assigning potentially informative clinically normal cases to liability classes based on age. We used the two-point lod score results from the first analysis strategy to generate two separate age of onset of symptoms/signs curves (for asymptomatic affected subjects, age at examination was taken as age at onset of signs), one for families where linkage to the chromosome 2 locus was likely (two-point lod score for any chromosome 2 marker 1.0), and a second for the remainder of the families. This second curve also contained age of onset data for two affected subjects from an additional family (family 14) not otherwise included in this paper. In this family, two siblings were affected by PHSP. While there was no other definite family history of PHSP, one parent was unavailable for examination, and Harding has pointed out that there is a significant chance that such families have ADPHSP (Harding, 1981). However, recessive PHSP could not be excluded definitely in this family. The curves were used to generate penetrance values for six liability classes. [Classes derived from the first curve (families where chromosome 2 linkage is likely): age at examination; 0–9 years, 7.8%; 10–19 years, 33.3%; 20–29 years, 64.7%; 30–39 years, 90.2%; 40–49 years, 96.1%; 50 years, 99%. Classes derived from the second curve (all families except those in which chromosome 2 linkage is likely): age at examination; 0–9 years, 39.4%; 10–19 years, 57.4%; 20–29 years, 76.6%; 30–39 years, 89.3%; 40–49 years, 98.9%; 50 years, 99%. Removal of family 14 had virtually no effect on the latter group of penetrance classes, as follows: 0–9 years, 40.2%; 10–19 years, 56.5%; 20–29 years, 76.0%; 30–39 years, 89.3%; 40–49 years, 98.9%; 50 years, 99%.] The penetrance values generated from the first curve (chromosome 2 linkage likely) were used with chromosome 2 locus markers, and the values generated for the second curve were used with chromosome 14 and 15 loci markers. These penetrance classes were assigned to `at risk' individuals (i.e. those who were clinically normal) on the basis of their age at examination. Maximum disease penetrance was assumed to be 99%, as previously suggested (Hazan et al., 1994). Subjects who had `suggestive signs' or who were `possibly affected' were classified as `unknown'. We also used these penetrance classes in recalculation of multipoint lod scores for certain families at loci where linkage had not been excluded definitely by the affecteds-only analysis.
Allele frequencies for all of the markers used were determined by genotyping a panel of unrelated spouses. In linkage calculations, the gene frequency for ADPHSP was assumed to be 10–4 and male and female recombination rates were assumed equal.
Statistical methods
Parametric data were analysed using the unpaired t test or factorial ANOVA (analysis of variance). Non-parametric data were analysed using either the Mann–Whitney U test or Spearman rank correlation.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Subjects and pedigree analysis
Three hundred and six family members were examined from the 28 participating families. Of these, 144 were classified as affected by ADPHSP, 10 had `suggestive' signs, 52 were `possibly affected' and 99 were clinically normal. One 77-year-old subject (II:1, family 18) who had no definite signs of spastic paraplegia when examined as part of this study, but had clearly documented findings of bilateral lower limb hyperreflexia, ankle clonus and extensor plantar responses on previous physical examination by an experienced neurologist, was classified as affected. It is difficult to explain these findings and so clinical information on this patient is not included in the study. No clinical data are presented on the 10 affected individuals who were not interviewed as part of the present study, with the exception of age of onset for three cases where this was documented. In each family, the disease transmission was consistent with autosomal dominant inheritance. The male to female ratio of examined, affected patients was 75 : 69 and male-to-male transmission was present in all of the families except 1, 6, 9, 10, 11, 12, 13, 19, 20, 24 and 28. There were affected females in all of the families lacking male-to-male transmission.
Linkage results
The two-point and multipoint lod scores were used to divide the families into three groups (Appendices 1 and 2). The first group consisted of families in which there was significant evidence of linkage to the chromosome 2 ADPHSP locus. Taking into account a prior probability of linkage to the chromosome 2 ADPHSP locus of 0.35, Dubé et al. (1997) have considered a lod score of 1.55 to indicate linkage, and this threshold lod score was reached, using either affecteds-only or penetrance class analysis, in families 2, 3, 5, 7, 22, 24, 25 and 27. For all of the families except family 3, penetrance class analysis gave a more positive maximum lod score than affecteds-only analysis. In family 3, affecteds-only analysis gave a maximum lod score of 1.73 with D2S2347, which decreased to –0.44 with penetrance class analysis, and so this family was not included in the chromosome 2-linked group. The second group consisted of families 1, 4, 8, 26 and 28, in which linkage to all three known loci was formally excluded. The third group consisted of families in which linkage results were inconclusive.
Haplotype analysis
The chromosome 2 ADPHSP locus previously has been narrowed to a candidate region of 3 cM, bounded by marker D2S367 at its centromeric end and D2S352 at its telomeric end (Scott et al., 1997). Three recombination events were found in affected members of our chromosome 2-linked families. All of these recombination events (subjects IV:6 in family 5, III:12 in family 7 and III:4 in family 25) involved marker D2S367, confirming this marker as the centromeric boundary of the chromosome 2 locus critical region. More telomeric markers (D2S2347, D2S2325, D2S2351 and D2S2203) in these three subjects were either uninformative or non-recombinant.
Clinical results
General features
The families were ascertained by referral from regional clinical genetics centres, as part of a UK-wide study. Although the mode of ascertainment may bias against smaller families, several of these are included. The families are therefore likely to be fairly representative of ADPHSP in the UK, and so a summary description of the disease phenotype is given in Tables 2 and 3. To determine whether large families were more mildly affected than small families, families with 5 examined, affected family members were compared with those with <5 examined, affected family members. There was no significant difference in overall disability score between these two groups, and this was not confounded by disease duration, which also did not differ between the two groups.
|
|
Comparison of chromosome 2-linked families with families excluded from linkage to all known ADPHSP loci
Clinical features were compared between the chromosome 2-linked families and the families where linkage to all three known loci had been excluded, to ascertain whether the clinical heterogeneity which has been described for ADPHSP is related to disease locus. Patients from linkage-excluded families had a younger age at onset of symptoms than chromosome 2-linked family patients (Fig. 2
, Table 3
). Affected patients from linkage-excluded families were also younger when they commenced using a walking stick, and tended to be younger when they commenced using a wheelchair.
|
With the exception of a marginally significant increase in the frequency of abnormal lower limb joint position sense in the linkage-excluded group, there was no significant difference in the proportion of patients from either linkage group suffering from a variety of disease features, summarized in Table 2
. In view of the discordant two-point lod scores generated for family 3 using affecteds-only versus liability class conditions, the above analyses were repeated with the inclusion of this family in the chromosome 2-linked group. Significant differences in age at onset of symptoms and age at commencing walking stick use remained. The marginally significant increase in abnormal lower limb joint position sense in the linkage-excluded group was still observed (P = 0.023) and there were significantly more subjects in the linkage-excluded group who had dysmetria (P = 0.03). All other comparisons gave similar results to those obtained when family 3 was excluded from analysis.
Disease progression was similar for the two linkage groups. Overall disability score did not differ, despite a similar mean disease duration (Table 3). Disease duration to commencement of stick use did not differ between the two linkage groups. There was a marginally significant difference in duration to commencement of wheelchair use, which might be explained by small sample size in the linkage-excluded group (Table 3). As a measure of rate of progression, disability score was divided by disease duration, and again there was no significant difference between the two linkage groups.
In both linkage groups, there were significant correlations between disease duration and increased lower limb tone, lower limb weakness, abnormal vibration sensation and disability score (Table 4). However, only the chromosome 2-linked group had a significant correlation between disease duration and bladder involvement, abnormal sensation in modalities other than vibration sense, lower limb hyperreflexia and absent lower limb reflexes. On the other hand, a significant correlation between upper limb hyperreflexia and disease duration was found only in the linkage-excluded group of families.
|
Harding used an age at onset cut-off of 35 years to divide ADPHSP families into early or late onset types and suggested that later onset families tend to have faster progression, a greater degree of lower limb weakness, more frequent bladder involvement and more frequent sensory symptoms (Harding, 1981
). Subsequent studies have proposed that a division at 20 years age at onset is more appropriate, and have not consistently found clinical differences between early and late onset groups of patients (Schady and Sheard, 1990; Dürr et al., 1994). We looked at rate of progression as a function of age at onset, by either comparing groups with an age at onset cut-off of 20 years or 35 years, or by Spearman rank correlation analysis. The total patient group and the chromosome 2-linked group showed significantly faster progression in older patients, irrespective of analytical procedure. (Total patient group: age 20 years cut-off, P < 0.0001; age 35 years cut-off, P = 0.004; Spearman correlation, P < 0.0001. Chromosome 2-linked group: age 20 years cut-off, P = 0.0015; age 35 years cut-off, P = 0.035; Spearman correlation, P < 0.0001.) The linkage-excluded group showed similar significant trends (age 20 years cut-off, P = 0.022; Spearman correlation, P = 0.012). The age 35 years cut-off data in the excluded group were not appropriate for this analysis, since only two patients had age at onset >35 years.
Testing for evidence of subgroups within linkage groups
Many neurodegenerative disorders display phenotypic variation, even when analysis is restricted to a single genetic locus. We therefore examined our chromosome 2-linked families and linkage-excluded families for evidence of statistically significant interfamilial variation. There was no significant difference in age at onset of symptoms between the chromosome 2-linked families, whether family 3 was included or excluded. However, there was a statistically significant difference in age at onset of symptoms between the excluded families (P < 0.0001), with the mean age at onset being lower in families 1 (7.0 ± 4.0 years, range 3–15 years), 4 (9.3 ± 8.5 years, range 1–28 years) and 28 (6.9 ± 6.2 years, range 2–22 years) than in families 8 (29.4 ± 11.4 years, range 18–57 years) and 26 (17.4 ± 3.6 years, range 2–22 years).
Segregation of disease-associated haplotypes amongst clinically unaffected subjects from chromosome 2-linked families
We analysed the segregation of the putative disease-associated haplotype in the seven chromosome 2-linked families. This allowed us to ascertain the genetic status of 49 subjects from these families who were clinically normal or who had an ambiguous phenotype. Three of 28 clinically normal subjects, seven of 19 possibly affected subjects and two of two subjects with suggestive signs had inherited a disease-associated haplotype (Table 5). These subjects may be pre-symptomatic, may represent variable expression of the disease phenotype (in those subjects with an ambiguous phenotype) or may represent true non-penetrance (in those subjects who were clinically normal). It is also possible that they may be true recombinants in families with a false-positive lod score.
|
Anticipation analysis
In several inherited neurodegenerative conditions, the underlying mutation involves a trinucleotide repeat expansion. Anticipation (decreased age at onset or increased disease severity with subsequent generations) is a clinical hallmark of these conditions, and has been suggested to occur in chromosome 2-linked and chromosome 14-linked ADPHSP (Hazan et al., 1994
; Bürger et al., 1996; de Jonghe et al., 1996; Scott et al., 1997). We examined the change in age at onset of symptoms in all available parent–child pairs from our families, and found a significant decrease in age at onset of symptoms from parent to child (mean –6.5 ± 12.2 years, range –48 to +23 years, P = 0.0002). When split by linkage group, there was a significant decrease in age at onset from parent to child in the group of linkage-excluded families (mean –7.7 ± 10.1 years, range –26 to +10 years, P = 0.0005), but not in the group of families linked to chromosome 2. Assessment of the presence of anticipation is problematic, since many biases may contribute to an erroneous impression of its existence (McInnis and Margolis, 1998). For example, eight of the 12 clinically unaffected subjects with disease-associated haplotypes described above are older than the mean age at onset for patients from chromosome 2-linked families. It is likely that these subjects, if they develop disease, will have an older age at onset than their parents, but such subjects cannot be controlled for in most assessments of anticipation. One approach which minimizes these problems is to compare the change in age at onset for parent–child pairs with a male affected parent versus parent–child pairs with a female affected parent, since there frequently is a parental sex bias in the degree of anticipation seen in trinucleotide repeat disorders (Rubinsztein and Hayden, 1998). We found no significant difference in mean decrease of age at onset between parent–child pairs with a male affected parent and parent–child pairs with a female affected parent when we examined pairs from all of our families, or from only chromosome 2-linked families. We found a marginally significant difference (P = 0.031) in the linkage-excluded families, with male parent pairs having a greater mean decrease in age at onset (–12.6 ± 10.4 years) than female parent pairs (–4.3 ± 8.5 years). |
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The UK families described in this study provide a clinical picture of ADPHSP which conforms to previous descriptions: a slowly progressive spastic paraplegia of variable age at onset and rate of progression, with lower limb hyperreflexia, hypertonicity, weakness, clonus and upgoing plantar responses, often accompanied by urinary dysfunction and abnormal vibration sensation (Harding, 1981
; Polo et al., 1993; Dürr et al., 1994; De Jonghe et al., 1996; Nielsen et al., 1998). As previously described, some patients had signs atypical of a spastic paraplegia, such as downgoing plantar reflexes or absent lower limb jerks (Harding, 1981; Polo et al., 1993; Dürr et al., 1994). A few patients had abnormalities in pain and fine touch sensation, findings which are not usually associated with ADPHSP. However, their presence is not unexpected, since subclinical sensory abnormalities affecting all modalities have been described in ADPHSP (Schady and Sheard, 1990). Many of the motor and extramotor clinical features correlated with disease duration.
Our study presents linkage results from a systematically ascertained population residing in a geographically defined region. The relative frequency of different ADPHSP genes in our families is therefore likely to give an indication of the relative gene frequency in the UK ADPHSP population. The 12 families which gave definite linkage results (either chromosome 2-linked or not linked to any of the known loci, excluding family 3) were the largest in our series of 28 families, having between five and 15 affected members. Thus, our linkage analysis results suggest that, for such medium to large sized UK families, the chromosome 2 ADPHSP locus accounts for ~60% of disease genes, and an unmapped gene(s) is responsible in the remaining 40%. We found no evidence of linkage to the chromosome 14 or 15 loci in any family. It might be argued that the lod score criteria used to declare linkage at the chromosome 2 locus (1.55) versus the chromosome 14 or 15 loci (>3) might bias the results in favour of detecting chromosome 2 locus linkage and against detecting chromosome 14 or 15 locus linkage, since medium sized families might be large enough to generate a lod score 1.55, but not 3. However, using a uniform lod score criterion for linkage of 1.55 would not have changed our linkage groups, and so it is unlikely that the methods used have significantly biased our results. The relative frequency of disease loci found in the larger families should be extrapolated with caution to smaller families, since it could be argued that more severely affected families tend to be smaller, and disease severity could be correlated with genetic locus. However, in our patient sample, there was no difference in mean severity score or rate of decline between subjects from small families versus subjects from large families.
In genetically heterogeneous conditions, the relative prevalence of disease genes may vary from population to population. For example, the prevalence of different spinocerebellar ataxia genes, all associated with a similar phenotype, varies with geographical location (Leggo et al., 1997). Only one other study has attempted systematically to determine the linkage group of a large number of ADPHSP families from a geographically defined region. Fink et al. presented a summary of linkage results in 33 ADPHSP families from several North American centres (Fink et al., 1996). Their findings approximate ours, suggesting that 45% of ADPHSP families are chromosome 2-linked, 45% are not linked to any known locus and the small remainder of families are linked to either the chromosome 14 or 15 loci.
No previous study has found definite, statistically significant differences in the clinical features of ADPHSP families from different linkage groups. We observed a correlation between linkage group and clinical features, with the chromosome 2-linked group having a significantly older age at onset of symptoms, age at commencing stick use and tending to be older before commencing wheelchair use than the group in which linkage to all known ADPHSP loci was excluded. There was no other striking difference between the two groups, and it appears that chromosome 2-linked ADPHSP is qualitatively very similar to that associated with the unknown locus/loci, but that its time course is shifted to the right. It might be argued that this appearance could be generated by a selection bias, since a larger family is required to generate a lod score of 1.55 than is required to exclude linkage, and families with late onset might tend to be larger because of greater reproductive fitness. Thus, one might expect that chromosome 2-linked families would be larger than unlinked families. However, in our patient sample, the linkage-excluded families contained, on average, a greater number of affected family members than the chromosome 2-linked group, making such a bias unlikely. A correlation between very early age of onset and chromosome 14 linkage has been suggested (Hazan et al., 1993; Hentati et al., 1994a; Gispert et al., 1995). Thus, age at onset may be, at least in part, locus dependent.
Our study confirms many of the clinical features previously described in two studies on chromosome 2-linked ADPHSP (Dürr et al., 1996; Nielsen et al., 1998). All three studies have found a similar mean age at symptom onset, between 25 and 30 years, and none have found significant evidence of heterogeneity in age at onset between families. In each study, the core clinical features of slowly progressive lower limb spasticity and weakness frequently were accompanied by abnormal lower limb vibration sensation and bladder involvement. A similar range of rarer associations, such as nystagmus or upper limb inco-ordination, was also found in all three studies.
In contrast to the homogeneity amongst chromosome 2-linked families, we found significant heterogeneity in age at onset of symptoms within the linkage-excluded group of families, with some families having very early onset, while others had later onset, similar to that of the chromosome 2-linked families. This might reflect locus heterogeneity within the linkage-excluded group of families. Alternatively, there are many neurodegenerative precedents for phenotypic variability being caused by allelic heterogeneity at a single locus, e.g. trinucleotide repeat diseases such as Huntington's disease or the spinocerebellar ataxias (Schols et al., 1997; Rubinsztein and Hayden, 1998).
The chromosome 2 ADPHSP locus previously has been narrowed to a region of 3 cM, bounded by markers D2S352 at the telomeric end and D2S367 at the centromeric end. Analysis of recombination events in our families confirmed the centromeric boundary of the critical interval at D2S367. Linkage analysis in a family with spastic paraplegia complicated by dementia and epilepsy mapped the responsible gene to a 0 cM region between D2S2255 and D2S2347, within the chromosome 2 ADPHSP locus (Heinzlef et al., 1998). However, it is possible that the disease gene in this family is distinct from that which causes ADPHSP, so the family cannot be used to narrow the ADPHSP critical region reliably.
The presence of clinical differences between early and late onset ADPHSP patients has been debated, with studies supporting (Harding, 1981) or finding no evidence of (Dürr et al., 1994) their presence. We found faster disease progression in late versus early onset patients. This difference was not locus specific, since we found it in both the chromosome 2-linked and linkage-excluded groups.
In several neurodegenerative conditions, the underlying mutation involves a CAG repeat expansion which is translated and expressed as an expanded, pathogenic polyglutamine tract. Anticipation is a feature of such diseases and reflects the fact that the CAG repeat expansion may increase in size from generation to generation (Rubinsztein and Hayden, 1998). Anticipation has been described for chromosome 2- and 14-linked ADPHSP families, and recent molecular genetic data have also implicated CAG/polyglutamine expansion mutations in chromosome 2-linked ADPHSP (Hazan et al., 1994; Bürger et al., 1996; de Jonghe et al., 1996; Nielsen et al., 1997; Scott et al., 1997). Although we found evidence for anticipation when we analysed our families as a whole and when we analysed the linkage-excluded group of families, we did not find any evidence of anticipation in the chromosome 2-linked families. Analysis of anticipation is prone to a number of biases which lead to false-positive results (McInnis and Margolis, 1998). It is impossible to control wholly for these biases, and so the marginally significant anticipation results for the linkage-excluded group of families should be viewed with caution.
In summary, this study confirms that the chromosome 2 ADPHSP locus is a major disease locus in UK families, and that the disease is associated with an unmapped gene(s) in a significant number of families. We found evidence that age of disease onset is associated with linkage group. We also detected evidence of clinical heterogeneity, perhaps reflecting locus or allelic heterogeneity, within the group of families not linked to the known ADPHSP loci. Further dissection of the relationship between genetic pathology and clinical features in ADPHSP will be helped by the mapping of the remaining disease loci and cloning of disease genes.
Appendix 1
Two-point lod scores for families from the chromosome 2-linked and linkage-excluded groups, using affecteds-only and liability class analysis. Lod scores at 
= 0.1 are not reported for chromosome 2 markers, since the candidate region is only 3 cM in length. Lod scores are not reported at >0.1 for the chromosome 14 or 15 loci, since the candidate regions are 7 cM in length. Lod scores for the other families tested are available from the authors on request.
Appendix 2
Multipoint lod score results. Definite = lod score >3, excluded = lod score <–2 throughout candidate region, probable = lod score 1.55 at marker from the chromosome 2 ADPHSP locus.
|
|
|
Acknowledgments |
|---|
We wish to thank the families for participating in this study. E.R. is a Wellcome Research Training Fellow, D.C.R. is a Glaxo Wellcome Research Fellow and M.R. was supported by the Muscular Dystrophy Group of Great Britain and Northern Ireland. This work was supported by the UK Medical Research Council. E.R. is supported by a Sackler Award.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Bruyn RP, van Veen MM, Kremer H, Scheltens PH, Padberg GW. Familial spastic paraplegia: evidence for a fourth locus. Clin Neurol Neurosurg 1997; 99: 87–90.[Web of Science][Medline]
Bürger J, Metzke K, Paternotte C, Schilling F, Hazan J, Reis A. Autosomal dominant spastic paraplegia with anticipation maps to a 4-cM interval on chromosome 2p21–2p24 in a large German family. Hum Genet 1996; 98: 371–75.[Web of Science][Medline]
Cambi F, Tang X-M, Cordray P, Fain PR, Keppen LD, Barker DF. Refined genetic mapping and proteolipid protein mutation analysis in X-linked pure hereditary spastic paraplegia. Neurology 1996; 46: 1112–7.
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; 93: 973–83.[Web of Science][Medline]
Cottingham RW Jr, Idury RM, Schaffer AA. Faster sequential genetic linkage computations. Am J Hum Genet 1993; 53: 252–63.[Web of Science][Medline]
De Jonghe P, Krols L, Michalik A, Hazan J, Smeyers G, Löfgren A, et al. Pure familial spastic paraplegia: clinical and genetic analysis of nine Belgian pedigrees. Eur J Hum Genet 1996; 4: 260–6.[Web of Science][Medline]
Dib C, Fauré S, Fizames C, Samson D, Drouot N, Vignal A, et al. A comprehensive genetic map of the human genome based on 5264 microsatellites [see comments]. Nature 1996; 380: 152–4. Comment in: Nature 1996; 380: 111.[Medline]
Dubé M-P, Mlodzienski MA, Kibar Z, Farlow MR, Ebers G, Harper P, et al. Hereditary spastic paraplegia: LOD-score considerations for confirmation of linkage in a heterogeneous trait. Am J Hum Genet 1997; 60: 625–9.[Web of Science][Medline]
Dürr A, Brice A, Serdaru M, Rancurel G, Derouesné C, Lyon-Caen O, et al. The phenotype of `pure' autosomal dominant spastic paraplegia. Neurology 1994; 44: 1274–7.
Dürr A, Davoine C-S, Paternotte C, von Fellenberg J, Cogilnicean S, Coutinho P, et al. Phenotype of autosomal dominant spastic paraplegia linked to chromosome 2. Brain 1996; 119: 1487–96.
Fink JK, Wu CT, Jones SM, Sharp GB, Lange BM, Lesicki A, et al. Autosomal dominant familial spastic paraplegia: tight linkage to chromosome 15q. Am J Hum Genet 1995; 56: 188–92.[Web of Science][Medline]
Fink JK, Heiman-Patterson T, Bird T, Cambi F, Dubé MP, Figlewicz DA, et al. Hereditary spastic paraplegia: advances in genetic research. [Review]. Neurology 1996; 46: 1507–14.
Gispert S, Santos N, Damen R, Voit T, Schulz J, Klockgether T, et al. Autosomal dominant familial spastic paraplegia: reduction of the FSP1 candidate region on chromosome 14q to 7 cM and locus heterogeneity. Am J Hum Genet 1995; 56: 183–7.[Web of Science][Medline]
Harding AE. Hereditary `pure' spastic paraplegia: a clinical and genetic study of 22 families. J Neurol Neurosurg Psychiatry 1981; 44: 871–83.
Harding AE. The hereditary ataxias and related disorders. Edinburgh: Churchill Livingstone; 1984.
Hazan J, Lamy C, Melki J, Munnich A, de Recondo J, Weissenbach J. Autosomal dominant familial spastic paraplegia is genetically heterogeneous and one locus maps to chromosome 14q. Nature Genet 1993; 5: 163–7.[Web of Science][Medline]
Hazan J, Fontaine B, Bruyn RP, Lamy C, van Deutekom JC, Rime CS, et al. Linkage of a new locus for autosomal dominant familial spastic paraplegia to chromosome 2p. Hum Mol Genet 1994; 3: 1569–73.
Heinzlef O, Paternotte C, Mahieux F, Prud'homme J-F, Dien J, Madigand M, et al. Mapping of a complicated familial spastic paraplegia to locus SPG4 on chromosome 2p. J Med Genet 1998; 35: 89–93.
Hentati A, Pericak-Vance MA, Lennon F, Wasserman B, Hentati F, Juneja T, et al. Linkage of a locus for autosomal dominant familial spastic paraplegia to chromosome 2p markers. Hum Mol Genet 1994a; 3: 1867–71.
Hentati A, Pericak-Vance MA, Hung W-Y, Belal S, Laing N, Boustany R-M, et al. Linkage of `pure' autosomal recessive familial spastic paraplegia to chromosome 8 markers and evidence of genetic locus heterogeneity. Hum Mol Genet 1994b; 3: 1263–7.
Huang S, Zhuyu, Li H, Labu, Baizhu, Lo WH, et al. Another pedigree with pure autosomal dominant spastic paraplegia (AD-FSP) from Tibet mapping to 14q11.2–q24.3. Hum Genet 1997; 100: 620–3.[Web of Science][Medline]
Kobayashi H, Garcia CA, Alfonso G, Marks HG, Hoffman EP. Molecular genetics of familial spastic paraplegia: a multitude of responsible genes. [Review]. J Neurol Sci 1996a; 137: 131–8.[Web of Science][Medline]
Kobayashi H, Garcia CA, Tay P, Hoffman EP. Extensive genetic heterogeneity in the `pure' form of autosomal dominant familial spastic paraplegia (Strümpell's disease). Muscle Nerve 1996b; 19: 1435–8.[Web of Science][Medline]
Lathrop GM, Lalouel J-M. Easy calculations of LOD scores and genetic risks on small computers. Am J Hum Genet 1984; 36: 460–5.[Web of Science][Medline]
Leggo J, Dalton A, Morrison PJ, Dodge A, Connarty M, Kotze MJ, et al. Analysis of spinocerebellar ataxia types 1, 2, 3, and 6, dentatorubral-pallidoluysian atrophy, and Friedreich's ataxia genes in spinocerebellar ataxia patients in the UK. J Med Genet 1997; 34: 982–5.
McInnis MG, Margolis RL. Anticipation, triplet repeats and psychiatric disorders. In: Rubinsztein DC, Hayden MR, editors. Analysis of triplet repeat disorders. Oxford: BIOS Scientific; 1998. p. 239–46.
Nielsen JE, Koefoed P, Abell K, Hasholt L, Eiberg H, Fenger K, et al. CAG repeat expansion in autosomal dominant pure spastic paraplegia linked to chromosome 2p21–p24. Hum Mol Genet 1997; 6: 1811–1816.
Nielsen JE, Krabbe K, Jennum P, Koefoed P, Jensen LN, Fenger K, et al. Autosomal dominant pure spastic paraplegia: a clinical, paraclinical, and genetic study. J Neurol Neurosurg Psychiatry 1998; 64: 61–6.
O'Connell JR, Weeks DE. The VITESSE algorithm for rapid exact multilocus linkage analysis via genotype set-recoding and fuzzy inheritance [see comments]. Nature Genet 1995; 11: 402–8. Comment in: Nature Genet 1995; 11: 354–5.[Web of Science][Medline]
Polo JM, Calleja J, Combarros O, Berciano J. Hereditary `pure' spastic paraplegia: a study of nine families. J Neurol Neurosurg Psychiatry 1993; 56: 175–81.
Raskind WH, Pericak-Vance MA, Lennon F, Wolff J, Lipe HP, Bird TD. Familial spastic paraparesis: evaluation of locus heterogeneity, anticipation, and haplotype mapping of the SPG4 locus on the short arm of chromosome 2. Am J Med Genet 1997; 74: 26–36.[Web of Science][Medline]
Reid E. Syndrome of the month: pure hereditary spastic paraplegia. [Review]. J Med Genet 1997; 34: 499–503.
Rubinsztein DC, Hayden MR, editors. Analysis of triplet repeat disorders. Oxford: BIOS Scientific; 1998. p. 1–12.
Schady W, Sheard A. A quantitative study of sensory function in hereditary spastic paraplegia. Brain 1990; 113: 709–20.
Schaffer AA, Gupta SK, Shriram K, Cottingham RW, Jr. Avoiding recomputation in linkage analysis. Hum Hered 1994; 44: 225–37.[Web of Science][Medline]
Schols S, Amoiridis G, Buttner T, Przuntek H, Epplen JT, Riess O. Autosomal dominant cerebellar ataxia: phenotypic differences in genetically defined subtypes? Ann Neurol 1997; 42: 924–32.[Web of Science][Medline]
Scott WK, Gaskel PC, Lennon F, Wolpert C, Menold MM, Aylsworth AS, et al. Locus heterogeneity, anticipation and reduction of the chromosome 2p minimal candidate region in autosomal dominant familial spastic paraplegia. Neurogenetics 1997; 1: 95–102.[Web of Science][Medline]
Tedeschi G, Allocca S, Di Costanzo A, Carlomagno S, Merla F, Petretta V, et al. Multisystem involvement of the central nervous system in Strümpell's disease. J Neurol Sci 1991; 103: 55–60.[Web of Science][Medline]
Webb S, Hutchinson M. Cognitive impairment in families with pure autosomal dominant hereditary spastic paraparesis. Brain 1998; 121: 923–9.
Received November 1, 1998. Revised February 8, 1999. Accepted April 16, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. S. Sellick, C. Longman, J. Tolmie, R. Newbury-Ecob, L. Geenhalgh, S. Hughes, M. Whiteford, C. Garrett, and R. S. Houlston Genomewide linkage searches for Mendelian disease loci can be efficiently conducted using high-density SNP genotyping arrays Nucleic Acids Res., November 23, 2004; 32(20): e164 - e164. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Orlacchio, T. Kawarai, E. Rogaeva, Y.Q. Song, A.D. Paterson, G. Bernardi, and P.H. St. George-Hyslop Clinical and genetic study of a large Italian family linked to SPG12 locus Neurology, November 12, 2002; 59(9): 1395 - 1401. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E REID, A ESCAYG, A M DEARLOVE, D D LEE, M H MEISLER, and D C RUBINSZTEIN The spastic paraplegia SPG10 locus: narrowing of critical region and exclusion of sodium channel gene SCN8A as a candidate J. Med. Genet., January 1, 2001; 38(1): 65 - 67. [Full Text]
|
|
|
|
|
|
P. McMonagle, P.C. Byrne, B. Fitzgerald, S. Webb, N.A. Parfrey, and M. Hutchinson Phenotype of AD-HSP due to mutations in the SPAST gene: Comparison with AD-HSP without mutations Neurology, December 26, 2000; 55(12): 1794 - 1800. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J C Lindsey, M E Lusher, C J McDermott, K D White, E Reid, D C Rubinsztein, R Bashir, J Hazan, P J Shaw, and K M D Bushby Mutation analysis of the spastin gene (SPG4) in patients with hereditary spastic paraparesis J. Med. Genet., October 1, 2000; 37(10): 759 - 765. [Abstract] [Full Text]
|
|
|
|
|
|
F. M. Santorelli, C. Patrono, D. Fortini, A. Tessa, G. Comanducci, E. Bertini, A. Pierallini, G. A. Amabile, and C. Casali Intrafamilial variability in hereditary spastic paraplegia associated with an SPG4 gene mutation Neurology, September 12, 2000; 55(5): 702 - 705. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. McDermott, K White, K Bushby, and P. Shaw Hereditary spastic paraparesis: a review of new developments J. Neurol. Neurosurg. Psychiatry, August 1, 2000; 69(2): 150 - 160. [Full Text] [PDF]
|
|
|
|
|
|
E. Reid, A. M. Dearlove, M. L. Whiteford, M. Rhodes, and D. C. Rubinsztein Autosomal dominant spastic paraplegia: Refined SPG8 locus and additional genetic heterogeneity Neurology, November 1, 1999; 53(8): 1844 - 1844. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||