Introduction

Huntington’s disease (HD) is an autosomal dominant neurodegenerative disorder characterized by presenile dementia, motor disturhance and psychiatric manifestations. It affects about 1 in 10,000 individuals of European origin and symptoms usually appear in the fourth to fifth decade [1, 2]. There is good correlation for age of onset between individuals from the same family, but paternal transmission of the disorder is often associated with earlier onset of symptoms [3]. The genetic defect causing HD has been mapped to chromosome band 4pl6.3 by linkage analysis, making presymptomatic testing for this disorder possible in an affected family setting [4, 5]. Recently a new gene, IT15, was isolated containing an expanded and unstable trinucleotide repeat ((CAG)n) in HD individuals [6]. We report on the expansion and transmission of this repeat in 248 individuals from 41 families. This new information will be important for presymptomatic testing and prenatal diagnosis.

Patients and Methods

Clinical information and genomic DNA were olµl tained from 248 individuals belonging to 41 families with HD as part of a presymptomatic testing programme at our center. From this group a total of 103 individuals were diagnosed with HD, and 38 asymptomatic individuals had entered the presymptomatic testing program. Age of onset information for HD was collected as part of a retrospective study and we estimate the maximum error for age of onset to be 2 years. All individuals concerned by this study had consented at the time of blood sampling to make this DNA also available for research in the future, under the condition that the results of such tests would not be made available to them. To protect the rights of the individuals (patients and controls) all samples were coded and provided for research only with the essential clinical information. As a result a new counselling session and a new blood sample will have to be taken from any individual who wants to confirm the result of the predictive testing by this new assay.

Polymerase chain reaction (PCR) amplification of the (CAG)n repeat in IT15 was performed on each sample using genomic primers [6], one of which was conjugated with fluorescein isothiocyanate (FITC). A 50 µl PCR reaction consisted of 500 ng genomic DNA, 10 pmol of each primer, 200 µM each dNTP, 1.2 mM MgCl2, 10% dimethylsulfoxide (DMSO), 16.6 mM (NH4)2SO4, 67 mM Tris-HCl (pH 8.8), 10 mM β-mercaptoethanol, 6.7 µM EDTA, 0.85 µg BSA and 1.25 units of Taq polymerase (Cetus). A total of 40 cycles were performed at the following temperatures: 95 °C for 1 min, 60° C for 1 min and 72° C for 2 min after an initial denaturation of 3 min at 95 °C. Denatured PCR products were electrophoresed on an automatic sequencing device (ALF, Pharmacia), sized according to a reference sequence, and the corresponding length of the (CAG)n repeat was calculated, knowing that a PCR fragment of 247 bp contains 18 CAG repeats [6]. DNA samples from the 8 individuals from table 3 were also analyzed with a different primer set immediately flanking the (CAG)n repeat (sense primer: 5′-ATG AAG GCC TTC GAG TCC CTC AAG TCC TTC-3′ and anti-sense primer: 5′-CGG CGG CGG CGG TGG CGG CTG T-3′.

Table 1 The different observed (CAG)n alleles in the IT15 gene are shown for a group of 65 founders without HD
Full size table
Table 2 a Difference in repeat length between parent and offspring shown for mother-child and father-child transmissions, and b maximal difference in repeat length between sibs of the same sibship is listed for sibships with maternal transmission vs. sibships with paternal transmission
Full size table
Table 3 Summary of clinical data for 8 individuals with an expanded repeat smaller than 42 is shown
Full size table

Results

In a sample of 65 founders (unrelated spouses), without HD, we observed 15 different alleles (table 1) with a heterozygote fraction of 0.893 and a polymorphism information content (PIC) value of 0.883. Affected individuals from every HD family showed an increase in (CAG)n repeat length ranging from 42 to 62 on the HD allele (102 HD alleles).

The length of the expanded (CAG)n repeat in IT15 correlated with age of onset of HD (fig. 1) for a group of 84 HD patients, with Pearson’s correlation coefficient r = −0.72 (95% CI −0.81 < r < −0.60) and r2 = 0.49. We did not find an additional correlation between age of onset and (CAG)n repeat length on the normal IT 15 allele. For individuals with a repeat unit length in the 42 to 50 size range the correlation was lower (r = −0.46). Results were available for a total of 21 parent-to-child transmissions of an expanded (CAG)n repeat (table 2a). Paternal transmission resulted in a significantly larger average repeat length in offspring of 2.79 repeats (SE 0.74) as compared with an average decrease of 0.29 repeats in offspring (SE 0.36) in case of maternal transmission (Wilcoxon two sample test, p = 0.0018). For several pedigrees no DNA was available from the transmitting parent. For those families we examined the maximum difference in repeat length between sibs from the same sibship (table 2b). Paternal transmission resulted in a significantly larger mean difference in repeat length per sibship than maternal transmission (table 2b, Wilcoxon two sample test, p = 0.027). The maximum repeat length in a sibship also correlated with the maximum difference in repeat length between sibs (Pearson’s correlation coefficient r = 0.56, 95% CI 0.25 < r < 0.76).

Fig. 1
figure 1

Relation between age of onset of symptoms and size of expanded (CAG)n repeat in 84 individuals with HD. Regression: Y = 124 − 1.73X, where Y is age of onset in years and X is the number of CAG repeats. CI

Full size image

In these 41 HD pedigrees, 8 individuals carried an expanded (CAG)n repeat smaller than 42 repeats (table 3). None of these individuals had a definite diagnosis of HD, but 3 were tentatively labeled as probably early HD, and individual 5.1 showed aspecific mental deterioration from the age of 80 years on.

It is known that the region of the (CAG)n repeat contains 2 CCG repeats that can be polymorphic [7]. To find out if this influenced the findings in the 8 individuals described in table 3, we analyzed their DNA samples also with PCR primers immediately flanking the (CAG)n repeat. The results of these tests are also listed in table 3 ((CAG)n;B). It is clear that there might be some variation in the CCG repeats but these 8 individuals clearly have repeats expanded outside the normal range but without the symptoms of typical HD.

Discussion

Analysis of the (CAG)n repeat in the IT15 gene in these 40 families supports the idea of genetic homogeneity in HD, because an expanded (CAG)n repeat was observed in every HD family tested. Analysis of the genetic transmission of this repeat also explains several known genetic characteristics of HD. The repeat size itself correlates with age of onset of HD (fig. 1) and 49% of the variance in age of onset is explained by variance in repeat length (r2 = 0.49), at least for the observed range in repeat lengths in HD patients. However, the correlation is weaker in the 42 to 50 repeat unit length range, and the length of the expanded repeat cannot be used to predict age of onset. This is also observed in other studies [8–10]. The observed expansion in repeat size in case of paternal transmission and the nearly unchanged repeat size in case of maternal transmission (table 2a, b), may explain why paternal transmission of HD is associated on average with an earlier age of onset than maternal transmission, and why the parent-offspring age-of-onset correlation is stronger in case of maternal transmission [3]. The difference in repeat length between sibs is larger in case of a longer repeat, suggesting that a longer expansion of the (CAG)n repeat is more unstable in meiosis than a shorter expansion. This phenomenon may be responsible for the very long expansions (> 100 repeats) observed in some cases of juvenile HD [6] after paternal transmission.

The most important finding in this study however is the presence of 8 individuals positive for the haplotype segregating with HD in their family, but without symptoms or only atypical symptoms of HD at an advanced age. These 8 individuals show an expanded (CAG)n repeat length between 39 and 41 (table 3) as measured with the primers published by the Huntington’s Disease Collaborative Research Group [6]. The expanded repeats have a length between 36 and 39 (CAG)n repeats when measured with primers immediately flanking the CAG repeat in IT15 [7]. These preliminary data would indicate that, at least in the examined population using the described technique, there might be a continuum in the number of expanded (CAG)n repeats in the IT15 gene with apparent normal mental and physical health, over repeats associated with minor motor symptoms or aspecific mental deterioration beginning after the age of 60, to longer repeats with increasing severity of symptoms and decreasing age of onset of typical HD. Affected relatives of individuals 16.3, 16.5, 18.5 and 39.4 (table 3) showed expanded (CAG)n repeats of 42, 43 and 44. Although the HD parents were not available for study in these cases, there is indirect evidence for a contraction of the repeat to a premutation size in individuals listed in table 3. On the other hand, we observed expansion from 34 (CAG)n repeats to 44 repeats after paternal transmission. The father with 34 (CAG)n repeats was asymptomatic. This indicates that a repeat length of 34 can be unstable and might expand in the disease range of repeat length after transmission. The expanded (CAG)n repeat in the 8 individuals listed in table 3 compares best with asymptomatic premutations observed in the fragile-X syndrome and in myotonic dystrophy [11]. It is tempting to speculate that expanded (CAG)n repeats in the range of 40 might also be responsible for other forms of hereditary chorea without dementia [12].

From these data it appears that precise measurement of the (CAG)n repeat may become very important for accurate genetic counselling in cases of presymptomatic or prenatal testing for HD. Presymptomatic testing with linked markers only will not detect a decrease in repeat length and might result in a very high risk for developing HD, whereas in some individuals with a (CAG)n repeat expansion of 39 or lower, the risk for developing HD within a normal life span may be close to zero. On the other hand, children from a normal 70-year-old individual belonging to a HD family might still be at risk for developing HD in case the repeat expanded and this may require DNA testing. The (CAG)n repeat test will also allow to test sporadic cases of HD, and presymptomatic testing will be possible for children of such an individual. However, we are still convinced that all presymptomatic testing should be done as part of a presymptomatic testing programme, and ‘urgent’ PCR-based presymptomatic testing outside a testing programme should be avoided because of the possible negative psychological impact for the tested individual and partner. A collaborative effort should be set up to assess the precise meaning of our observations on borderline expansions of the (CAG)n repeat in HD. In the mean time the (CAG)n test should be used with caution and other diseases, which share some characteristics with Huntington’s chorea, should be studied.