|
|
| |||||||||||||||||||||||||||||||
Background Huntington's disease is associated with an expanded sequence of CAG repeats in a gene on chromosome 4p16.3. However, neither the sensitivity of expanded CAG repeats in affected persons of different ethnic origins nor the specificity of such repeats for Huntington's disease as compared with other neuropsychiatric disorders has been determined.
Methods We studied 1007 patients with diagnosed Huntington's disease from 565 families and 43 national and ethnic groups. In addition, the length of the CAG repeat was determined in 113 control subjects with a family history of Alzheimer's disease (44 patients), schizophrenia (39), major depression (16), senile chorea (5), benign hereditary chorea (5), neuroacanthocytosis (2), and dentatorubropallidoluysian atrophy (2). The number of CAG repeats was also assessed in 1595 control chromosomes, with the size of adjacent polymorphic CCG trinucleotide repeats taken into account.
Results Of 1007 patients with signs and symptoms compatible with a diagnosis of Huntington's disease, 995 had an expanded CAG repeat that included from 36 to 121 repeats (median, 44) (sensitivity, 98.8 percent; 95 percent confidence interval, 97.7 to 99.4 percent). There were no significant differences among national and ethnic groups in the number of repeats. No CAG expansion was found in the 113 control subjects with other neuropsychiatric disorders (specificity, 100 percent; 95 percent confidence interval, 95.2 to 100 percent). In 1581 of the 1595 control chromosomes (99.1 percent), the number of CAG repeats ranged from 10 to 29 (median, 18). In 12 control chromosomes (0.75 percent), intermediate-sized CAG sequences with 30 to 35 repeats were found, and 2 normal chromosomes unexpectedly had expanded CAG sequences, of 39 and 37 repeats.
Conclusions CAG trinucleotide expansion is the molecular basis of Huntington's disease worldwide and is a highly sensitive and specific marker for inheritance of the disease mutation.
Recently a novel gene was identified on chromosome 4p16.3 with a CAG trinucleotide repeat that is expanded on chromosomes bearing the mutation for Huntington's disease3. Huntington's disease is one of seven diseases now known to be associated with increased numbers of triplet repeats in novel genes4,5,6,7,8,9,10. The gene for Huntington's disease encodes two transcripts of approximately 10.4 and 13.5 kb11 and is widely expressed inside and outside the central nervous system12,13. The biochemical defect underlying the disease remains unknown, and no treatment halts the progression of this illness.
Before the cloning of the gene, DNA markers were used in testing programs to determine with high probability whether a person at risk for Huntington's disease had inherited the mutant gene14,15. A substantial limitation of the predictive test, however, was that the accuracy of the result was less than 100 percent, because of potential recombination between the DNA marker used and the site of the mutation for Huntington's disease. In addition, DNA from numerous family members was required in order to determine which phase of the marker was segregating with the gene. As a result, a substantial proportion of those who sought the test (up to 25 percent) were excluded from having it because blood from crucial relatives was unavailable14,15,16.
Before CAG expansion can be used as a marker for the inheritance of Huntington's disease, the limits of accuracy of this approach need to be assessed. We present the results of an assessment of CAG expansion in 1022 patients with diagnosed Huntington's disease from 43 national and ethnic groups and six continents. In addition, we examine the range in the number of CAG repeats in persons with other neuropsychiatric disorders commonly misdiagnosed as Huntington's disease.
Methods
Selection of Patients
DNA samples and clinical records for persons with diagnoses of Huntington's disease have been collected in the Canadian DNA Bank for Huntington's Disease since 1984, from families in Canada and many parts of the world, including those of European, Asian, black South African, Arab, and Native American descent. The ethnic origins of the patients were considered to be defined by the country from which the ancestor with Huntington's disease originated. The racial and geographic origins of the DNA samples from these families are shown in Table 1. The clinical diagnoses of Huntington's disease were made locally by neurologists or geneticists. In addition, clinical details about these patients were derived from extensive study of records, documented neurologic examinations, and special investigations, such as reports of positron-emission tomography and autopsy.
|
DNA Analysis and Assessment of CAG Repeats
Genomic DNA was extracted from leukocytes by standard procedures17. In the past, methods of CAG trinucleotide assessment have used primers that encompassed not only the CAG repeat but also a flanking CCG repeat18,19,20. These prior estimates of CAG size were made with the assumption that the CCG repeat invariably included seven triplets. However, we and others have recently shown that the CCG repeat is polymorphic, ranging from 7 to 12 triplets21,22. This variation would have particular importance in the assessment of persons estimated to have 36 to 42 CAG repeats, which is in the affected range. Therefore, after the initial estimate of the number of CAG repeats in persons thought to have from 36 to 42 such repeats,18 a precise assessment of the number of CAG repeats was performed with the CCG repeats excluded21. This method allowed an accurate delineation of the lower limit of CAG expansion in patients with Huntington's disease.
Results
CAG Repeats in Huntington's Disease and Other Neuropsychiatric Disorders
A total of 1007 patients had signs and symptoms compatible with a diagnosis of Huntington's disease. In 995 patients (98.8 percent), an expanded number of CAG repeats was demonstrated, ranging from 36 to 121 (median, 44). The 995 affected persons belonged to 565 separate pedigrees and 43 different national or ethnic groups. Among those affected, no obvious differences in allele size were found in persons from different national and ethnic groups. An increased median number of CAG repeats was seen in affected black South Africans, as well as in affected persons of Chinese, Japanese, Saudi Arabian, and Native American descent, but the sample was not large enough for a statistical analysis to be undertaken (Table 1).
In the case of 12 affected persons in whom the number of CAG repeats was in the normal range (Figure 1A), additional samples of DNA were requested, and the CAG repeats were measured again. Further examination of the clinical records as well as additional information obtained by neuropathological examination and positron-emission tomography revealed clinical features atypical for Huntington's disease in 11 patients. An extensive analysis of the clinical presentations of these 12 patients appears elsewhere23.
|
|
The normal CAG repeat, as determined by the study of 599 of 600 chromosomes from normal control subjects, ranges from 10 to 29 trinucleotides (Figure 1B). The remaining chromosome studied in this group had a CAG repeat with 39 trinucleotides. Repeats with 18 trinucleotides were relatively underrepresented, whereas those with 17 or 19 triplets were found at a peak frequency.
In addition to these 600 control chromosomes, the 995 chromosomes from the affected persons that did not contain the expanded CAG repeat were used to study the normal range of CAG repeats (Figure 1C). Of these chromosomes, 982 (99 percent) had between 10 and 29 CAG repeats, a range previously determined to be normal24. Chromosomes with 18 CAG repeats were again underrepresented as compared with those with 17 and 19 repeats. Of 1595 control chromosomes studied in all, 1581 (99.1 percent) had CAG sequences in the normal range. Comparisons of CAG repeats in control subjects of white, black, and Chinese descent revealed significant differences in the number of repeats between whites and blacks (P = 0.003, by analysis of variance) and between whites and Chinese persons (P = 0.012) (Table 3). Although these differences are statistically significant, they are small. Alleles with between 30 and 35 CAG repeats ("intermediate" alleles) were found in 12 of the 1595 control chromosomes (0.75 percent).
|
One control subject was unexpectedly found to have 39 CAG repeats on one chromosome (after correction for CCG repeats). She was the spouse of an affected person and had no signs or symptoms at the age of 57, and no family history of Huntington's disease in more than three generations. Analysis of her offspring confirmed this finding but also identified an unexpected homozygote for Huntington's disease. At the time of the analysis this person was 25 years old and asymptomatic but had chromosomes with 39 and 42 CAG repeats, from her mother and affected father, respectively.
The calculations of sensitivity and specificity26 were based on the study of 995 patients with clinically evident Huntington's disease and expanded CAG repeats; 12 persons with clinical signs and symptoms compatible with Huntington's disease but normal CAG repeats; and those with another neuropsychiatric disease who all had CAG repeats within the normal range (113 patients). This analysis yielded an estimated sensitivity of 98.8 percent (95 percent confidence interval, 97.7 to 99.4 percent) and a specificity of 100 percent (95 percent confidence interval, 95.2 to 100 percent) for the use of the number of CAG repeats to identify those with the mutation for Huntington's disease.
Discussion
We have shown here that CAG expansion within the Huntington's disease gene is a highly sensitive and specific marker for the inheritance of the mutation for Huntington's disease. A total of 995 of 1007 persons (98.8 percent) from 43 different national and ethnic groups who clearly had a clinical diagnosis of Huntington's disease had substantially expanded CAG repeats. These results support the earlier findings regarding the sensitivity of CAG expansion in a smaller group of patients, presumably from the United Kingdom27. Twelve persons in our cohort in whom Huntington's disease was diagnosed did not have CAG expansion. At least some of these patients seemed to have other diseases23. Even though other mutations in the gene for Huntington's disease are unlikely, this possibility has not been excluded.
The absence of CAG expansion in other neuropsychiatric disorders means that this molecular test can be used to differentiate Huntington's disease from the other illnesses. The determination of CAG size will be especially useful in symptomatic persons with no clear family history of Huntington's disease, a situation in which diagnosis has been difficult. This study shows that CAG expansion is the molecular basis of Huntington's disease worldwide and suggests that such expansion is directly related to the causation of the disease even though the mechanism remains unknown. The high degree of sensitivity and specificity of CAG expansion for the inheritance of Huntington's disease has substantial implications for both the assessment of symptomatic persons and predictive testing programs.
For persons at risk for Huntington's disease, a direct test for inheritance of the mutation will confer substantial advantages. In particular, the test allows more accurate assessment of genetic risk, without the need to obtain DNA from many family members. Another advantage is that privacy and confidentiality can be maintained because of the reduced need for blood samples from relatives. However, since the misdiagnosis of other illnesses as Huntington's disease may occur, the testing of DNA from at least one affected relative is recommended in order to confirm that CAG expansion is present in other affected persons in the family. This finding will allow the correct interpretation of a normal number of CAG repeats in a person at risk.
We and others24,28,29 have shown a strong inverse correlation between the number of CAG repeats and the patient's age at onset, but the identification of the expanded repeat does not itself prove that the person's signs and symptoms are manifestations of Huntington's disease. Persons who have inherited the mutation for Huntington's disease can still have other neuropsychiatric disorders that are unrelated to the expanded number of CAG repeats. Therefore, a careful neurologic evaluation is crucial in symptomatic persons with expanded CAG repeats. An asymptomatic person with an expanded CAG sequence will probably have signs and symptoms of Huntington's disease at some time in the future. However, this conclusion is based on the present study of CAG repeats in affected persons and may not apply to all asymptomatic persons with CAG expansion. It is not known whether Huntington's disease will definitely develop in everyone with an expanded repeat, particularly those with no family history.
Even with the improved accuracy and simplicity of the direct DNA test, the need for appropriate counseling and support remains. Furthermore, some delay between the request for the test and the provision of test results remains important, because during the pretest counseling period one quarter of the initial program participants withdraw from the testing program14,16,30. At present little is known about the psychological impact of learning definitively that Huntington's disease will develop. In view of the demonstrated low risk of adverse events in predictive testing that includes appropriate counseling,15,16 we recommend that direct predictive testing be provided with counseling and support as outlined in established guidelines31.
The molecular test to confirm the clinical suspicion of Huntington's disease should be highly cost effective. In Canada, this laboratory test is currently offered for approximately $300 per person tested, which is substantially below the cost of other specialized tests used in patients thought to have Huntington's disease, such as computed tomography, magnetic resonance imaging, positron-emission tomography, and neuropsychological examination.
Unexpectedly, we identified two homozygotes for Huntington's disease by direct detection of the expanded allele on both chromosomes. The clinical phenotype and pathological findings in one of these affected persons were similar to those of heterozygotes for the mutation associated with Huntington's disease. In the other person, a 25-year-old homozygote, symptoms were absent. These observations support the previously reported findings that the phenotype of the homozygote is not more severe than that of the heterozygote32,33 and are consistent with the idea that CAG expansion confers a gain of function in the Huntington's disease gene product.
Twelve of 1595 CAG alleles (0.75 percent) on control chromosomes had between 30 and 35 repeats. These intermediate alleles were all seen on the normal chromosomes of affected persons. We and others have shown that new mutations for Huntington's disease arise from intermediate alleles34,35. The stability of these alleles on control chromosomes is uncertain, but it is likely that they represent the pool from which new mutations for the disease arise. Transmission through the male germ line and advanced paternal age are factors that increase the susceptibility of intermediate alleles to further expansion34. A more precise estimate of the frequency and stability of such alleles in the general population awaits a study of a larger number of control chromosomes.
The higher frequency of Huntington's disease among whites and its lower prevalence in other populations, including African blacks and the Japanese, have led to the hypothesis that the mutation reponsible for the disease was carried to different parts of the world by immigrant European settlers2. This theory was further supported by the suggestion that the mutation rate in the gene for Huntington's disease is exceedingly low, perhaps the lowest such rate for any human genetic disease36. It is now apparent, however, that the mutation rate for Huntington's disease is higher than was estimated previously and that new mutations may account for up to 3 percent of cases34. Therefore, new mutations, in addition to European migration, may account for the presence of the disease in many different and sometimes isolated communities.
We have shown in this worldwide study that CAG expansion is present in Huntington's disease in persons of many ancestries and racial groups. In addition to being a sensitive indicator of the inheritance of Huntington's disease, CAG expansion is also highly specific, since it is not seen in other neuropsychiatric disorders with which Huntington's disease is frequently confused.
Supported by the Medical Research Council (Canada), the Canadian Genetic Diseases Network, the Neurodegenerative Disorders Centre (University of British Columbia), the Huntington Society of Canada, the National Institute on Aging (grant AGO 5136), and the Department of Veterans Affairs.
We are indebted to our colleagues and the members of the Canadian Huntington Disease Collaborative Group on Predictive Testing for information and DNA samples, and to Dr. J. Stoessl of London, Ontario, and Claire Goldsmith of Ottawa for additional information.
Source Information
From the Department of Medical Genetics, University of British Columbia, Vancouver (B.K., P.G., S.E.A., J.T., H.T., J.Z., F.S., B.L., M.R.H.); the Clarke Institute of Psychiatry, University of Toronto, Toronto (A.B.); the Karolinska Hospital, Stockholm, Sweden (E.A.); and the Veterans Affairs Medical Center and the Department of Medicine, University of Washington, Seattle (T.D.B.). The following institutions and researchers participated in the study: J. Greenberg, University of Cape Town, Cape Town, South Africa; G. Lucotte, Centre Hospitalier Universitaire-Centre Hospitalier Regional de Reims, Reims, France; M. Anvret, Karolinska Hospital, Stockholm, Sweden; J. Kennedy and A. Petronis, Clarke Institute of Psychiatry, University of Toronto, Toronto; S. Bohlega, King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia; G. Campanella, Department of Neurology, University of Naples, Naples, Italy; H. Shinotoh and D.B. Calne, Neurodegenerative Disorders Centre, University of British Columbia, Vancouver; S. Adam and C. Benjamin, Department of Medical Genetics, University of British Columbia, Vancouver; W.G. Honer, Department of Psychiatry, University of British Columbia, Vancouver; E. Ives, Memorial University, St. John's, Newfoundland; and G.D. Schellenberg, Department of Medicine, University of Washington, Seattle.
Address reprint requests to Dr. Hayden at the Department of Medical Genetics, University of British Columbia, 416-2125 East Mall, NCE Bldg., Vancouver, BC V6T 1Z4, Canada.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
