Gene Map Locus: 4p16.3
Shiwach (1994) performed a retrospective study of 110 patients with Huntington disease in 30 families. He found the minimal lifetime prevalence of depression to be 39%. The frequency of symptomatic schizophrenia was 9%, and significant personality change was found in 72% of the sample. The age at onset is highly variable: some showed signs in the first decade and some not until over 60 years of age. The mode is between 30 and 40 years (Chandler et al., 1960). In a study of 196 kindreds, Reed and Neel (1959) found only 8 in which both parents of a single patient with Huntington chorea were 60 years of age or older and normal. The clinical features develop progressively with severe increase in choreic movements and dementia and the disease terminates in death on average 17 years after manifestation of the first symptoms. Lovestone et al. (1996) described an unusual HD family in which all 4 affected members presented first with a severe psychiatric syndrome which in 3 cases was schizophreniform in nature. Two other living members with no apparent signs of motor disorder had received psychiatric treatment, 1 for schizophrenia.
Giordani et al. (1995) performed extensive neuropsychological evaluations on 8 genotype positive individuals comparing them to 8 genotype negative individuals from families with Huntington disease. They found no significant differences between these 2 groups, casting further doubt on earlier reports that suggested cognitive impairments or premonitory signs of the classical neurological syndrome of Huntington disease.
Video clips of a Huntington patient were taken at 2-year intervals and demonstrate the progression in the disease. The was taken when the patient was 60 years old and 6-7 years after the onset of symptoms and shows the characteristic hyperactivity of Huntington patients. The demonstrates a clear degeneration in control of head movements and in particular of the tongue. The documents the continued degeneration. Images kindly donated by John Burn.
Rosenberg et al. (1995) performed a double blind study on 33 persons at risk for HD who had applied for genetic testing. Significantly inferior cognitive functioning was disclosed in gene carriers by a battery of neuropsychological tests covering attentional, visuospatial, learning, memory, and planning functions. Primarily, attentional, learning, and planning functions were affected. Bamford et al. (1995) performed a prospective analysis of neuropsychological performance and CT scans of 60 individuals with Huntington disease. They found that psychomotor skills showed the most significant consistent decline among cognitive functions assessed.
Behan and Bone (1977) reported hereditary chorea without dementia. This probably represented the extreme of variability of Huntington chorea. The oldest affected person in their family was aged 61 years. A family with older members with Huntington chorea and no dementia has been observed by McKusick. Peyser et al. (1995) found no beneficial effect in treatment with d-alpha-tocopherol in a cohort of 73 patients with Huntington disease. However, postop analysis suggested possible beneficial effect on neurologic symptoms for patients early in the course of the disease.
Horton et al. (1995) used serial dilution PCR to demonstrate an 11-fold increase of the common 4977 nucleotide mitochondrial DNA deletion in temporal lobes of Huntington disease patients compared to normal controls. Huntington disease frontal lobes have 5-fold greater levels, whereas occipital lobe and putamen deletion levels were comparable with control levels. The authors hypothesized that the increased rate of mitochondrial DNA deletions could be caused by elevated oxygen radical production by mitochondria in Huntington disease patients. Gu et al. (1996) demonstrated marked deficiency of the mitochondrial respiratory chain in the caudate nucleus but not the platelets from patients with Huntington disease.
Reiner et al. (1988) used immunohistochemical methods to study neurons producing substance P and enkephalin, projecting to the globus pallidus and to the substantia nigra, in brains from 17 patients with Huntington disease in various stages of the disorder. The authors found that in the early and middle stages of HD, the enkephalin-producing neurons with projections to the external portion of the globus pallidus were more affected than substance P-containing neurons projecting to the internal pallidal segment. (This result was confirmed by Sapp et al., 1995). Reiner et al. (1988) also found that substance P-producing neurons projecting to the substantia nigra pars reticulata were more affected than those projecting to the pars compacta. In the advanced stages of the disease, neurons projecting to all striatal areas were depleted. Richfield and Herkenham (1994) found greater loss of cannabinoid receptors on striatal nerve terminals in the lateral globus pallidus compared to the medial pallidum in Huntington disease of all neuropathologic grades, supporting the preferential loss of striatal neurons that project to the lateral globus pallidus.
Hoogeveen et al. (1993) synthesized oligopeptides corresponding to the carboxy-terminal end of the predicted HD gene product. Immunobiochemical studies with polyclonal antibodies directed against this synthetic peptide revealed the presence of a protein (huntingtin) with a molecular mass of approximately 330 kD in lymphoblastoid cells from normal individuals and patients with Huntington disease. Immunocytochemical studies showed a cytoplasmic localization in various cell types, including neurons. In most of the neuronal cells, the protein was also present in the nucleus. No difference in molecular mass or intracellular localization was found between normal and mutant cells.
De Rooij et al. (1996) used affinity-purified antibodies to analyze the subcellular location of huntintin. In mouse embryonic fibroblasts, human skin fibroblasts, and mouse neuroblastoma cells, they detected huntigtin in the cytoplasm and the nucleus.
Dure et al. (1994) examined the in situ hybridization of riboprobes specific for the IT15 gene against normal human fetal and adult brains. In both types of specimen, the autoradiographic signal correlated strongly with cell number except in the germinal matrix and white matter where there is a significant proportion of glial cells. This suggests that IT15 expression is predominantly neuronal. However, there was no predominance of IT15 expression in the striatum of the fetal brain.
Aronin et al. (1995) detected mutant huntingtin protein in cortical synaptosomes isolated from brains of Huntington disease heterozygotes and demonstrated that the mutant species is synthesized and transported with the normal protein to nerve endings. In half of the juvenile cases, huntingtin resolved as a complex of bands after electrophoresis and immunostaining, which confirmed previous DNA evidence for somatic mosaicism. Mutant huntingtin was present in both normal and affected regions.
Wallace and Hall (1972) suggested that in Queensland, Australia, 2 possibly allelic forms may exist, one with early onset and one with late onset. Myers et al. (1982) confirmed the preponderance of inheritance from the father when HD had an early onset. 'Anticipation' was thought to reflect the finding that persons with early onset in prior generations were selectively nonreproductive because of manifestation of the disorder. In 238 patients, Myers et al. (1983) correlated age of onset with whether inheritance was from the father or the mother. More than twice as many of the late-onset cases (age 50 or later) inherited the HD gene from an affected mother than from an affected father. Affected offspring of late-onset females also had late-onset disease while those of late-onset males had significantly earlier ages of onset. The authors interpreted these findings as suggesting a heritable extrachromosomal factor, perhaps mitochondrial. They cited Harding (1981) as suggesting that autosomal dominant late-onset cerebellar ataxia is marked by earlier age of onset and death in offspring of affected males.
Boehnke et al. (1983) tested models to account for the stronger parent-offspring age-of-onset correlation when the mother is the affected parent and the excess of paternal transmission in cases with onset at less than 21 years. They proposed 2 models in which a maternal factor--cytoplasmic (possibly mitochondrial) in one case and autosomal or X-linked in the other--acts to delay onset.
Folstein et al. (1984, 1985) contrasted HD in 2 very large Maryland pedigrees: one was a black family residing in a bayshore, tobacco farming community; the other was a white Lutheran family living in a farming community in the western Maryland foothills and descended from an immigrant from Germany. They differed, respectively, in age of onset (33 years vs 50 years), presence of manic-depressive symptoms (2 vs 75), number of cases of juvenile onset (6 vs 0), mode of onset (abnormal gait vs psychiatric symptoms), and frequency of rigidity or akinesia (5/21 vs 1/15). In the black family, the mean age of onset was 25 years when the father was affected and 41 years when the mother was affected; the corresponding figures in the white family were 49 and 52 years. Allelic mutations were postulated.
In another survey in Maryland, Folstein et al. (1987) found that the prevalence in blacks was unexpectedly high and equal to that in whites. Age of onset was earlier in blacks, and their clinical features at all ages of onset were similar to those seen in juvenile-onset Huntington disease. Blacks had more severe bradykinesia and abnormalities of eye movement and less frequent psychiatric disorder, particularly depression. Went et al. (1984) confirmed the earlier report that early-onset HD is almost always inherited from the father but could not confirm the notion that late-onset disease is more often inherited from the mother.
Wexler et al. (1985, 1987) identified persons homozygous for the Huntington gene by study of branches of the large Venezuelan kindred in which there are instances of both parents being affected. Homozygosity was indicated by homozygosity for the G8 probe. Remarkably, comparison with the usual heterozygotes revealed no difference of phenotype. Wexler et al. (1987) suggested that this is the first human disease in which complete dominance has been demonstrated. Connarty et al. (1996) identified 2 patients in Wessex in the U.K. in whom expansion of the HD triplet repeat was found on both chromosomes. Both were males who presented in middle age with typical clinical features. Unfortunately, no other family members were available for analysis. (Zlotogora (1997) reviewed evidence that myotonic dystrophy (160900) is also a true dominant.)
Ridley et al. (1988) found that while the mean age of onset in offspring of affected mothers did not differ greatly from that in their mothers, the distribution of age of onset in the offspring of affected fathers fell into 2 groups; the larger group showed an age of onset only slightly younger than that in their affected fathers, and a smaller group had, on average, an age of onset 24 years younger than that of their affected fathers. Analysis of the grandparental origin of the Huntington allele suggested that while propensity to anticipation could be inherited for a number of generations through the male line, it originated at the time of differentiation of the germ line of a male who acquired the Huntington allele from his mother. Ridley et al. (1988) suggested that major anticipation indicates an epigenetic change in methylation of the nucleic acid of the genome, which is imposed in the course of 'genomic imprinting,' that is, in the mechanism by which the parental origin of alleles is indicated (Reik et al., 1987; Sapienza et al., 1987).
In South Wales over a 10-year period, Quarrell et al. (1986) found 192 patients with HD in whom there was a positive family history and an additional 37 patients who had clinical features consistent with HD but who had no affected relatives despite detailed inquiries. After review, 22 of the 37 were still thought to have HD on clinical grounds; the diagnosis was considered less likely in 15. Postmortem supported the diagnosis in 6 of 7 cases so studied; a patient labeled HD on the death certificate had Kufs disease (204300) at postmortem. The fact that no evidence of linkage disequilibrium has been found in HD with the G8 marker (Conneally et al., 1989) may suggest that the mutation is ancient and has occurred on very few occasions.
Quarrell et al. (1988) presented data suggesting that there has been a steady decline in births at risk for HD in both North Wales and South Wales in the period between 1973 and 1987. Lanska et al. (1988) determined an overall mortality rate for HD in the U.S. of 2.27 per million population per year. Age-specific mortality rates peaked around age 60. Lanska et al. (1988) suggested from their experience that the risk of suicide may have been overstated.
Adams et al. (1988) found that life-table estimates of age of onset of motor symptoms have produced a median age 5 years older than the observed mean when correction for truncated intervals of observation (censoring) was made. The bias of censoring refers to the variable intervals of observation and loss to observation at different ages. For example, gene carriers lost to follow-up, those deceased before onset of disease, and those who had not yet manifested the disease at the time of data collection were excluded from the observed distribution of age of onset.
Adams et al. (1988) also found that the offspring of affected males had significantly younger onset than did offspring of affected females, and a trend suggested an excess of paternal descent among juvenile-onset cases. Ridley et al. (1991) showed that the age of onset varies between families and between paternal and maternal transmission and that rigidity is associated specifically with very early onset, major anticipation, paternal transmission, and young parental age of onset. Major anticipation was defined as an age of onset of the proband more than 15 years less than that in the affected parent. They proposed that age of onset depends on the state of methylation of the HD locus, which varies as a familial trait, and as a consequence of 'genomic imprinting' determined by parental transmission. They further suggested that young familial age of onset and paternal imprinting occasionally interact to produce a major change in gene expression, that is, the early-onset/rigid variant.
Navarrete et al. (1994) described a family in which a brother and sister had very early onset of Huntington disease. Clinical manifestations were apparent in both sibs at the age of 8 years; the brother died at age 10. The father of these sibs was affected from the age of 29 years.
In 2 families with Huntington disease linked to the short arm of chromosome 4, Sax et al. (1989) demonstrated remarkable intrafamilial variability. In 1 family, affected persons of 3 generations showed a 50-year variation in age of onset. The member with the latest onset (at age 67) died at age 91 with autopsy-confirmed HD. The next generation had hypotonic chorea beginning in the fourth decade with death in the fifth. In the third generation, a rigid patient, inheriting the illness from an affected father, had onset at age 16, while her sibs had chorea beginning in the third decade. In the second family, several members had cerebellar signs as well as chorea and dementia; MRI and CT showed olivopontocerebellar and striatal atrophy. Whether these phenotypes are the result of different allelic genes at the HD locus or of unlinked autosomal modifying loci is unknown. Kerbeshian et al. (1991) described a patient with childhood-onset Tourette syndrome who later developed Huntington disease.
A large Tasmanian family with Huntington disease was first described by Brothers (1949). Pridmore (1990) traced 9 generations, starting with the father of the woman who brought the disease to Tasmania. From that woman, 6 lines had living affected descendants and a total of 765 living descendants at risk. The numbers of affected males and females were equal. The mean age of onset was 48.6 years and the mean age of death, 61.8 years. Affected members were at least as fertile as members of the general population. Pridmore (1990) concluded that late-onset disease (defined as death after 63 years of age) was associated with significantly greater fertility (in men more so than women) compared with that of affected sibs of the same sex. Unaffected sibs produced fewer offspring than in the general population.
Farrer et al. (1993) tested the hypothesis that the normal HD allele or a closely linked gene on the nonmutant chromosome influences age at onset of HD. Analysis of the transmission patterns of genetically linked markers at the D4S10 locus in the normal parent against age at onset in the affected offspring in 21 sibships and 14 kindreds showed a significant tendency for sibs who have similar onset ages to share the same D4S10 allele from the normal parent. Affected sibs who inherited different D4S10 alleles from the normal parent tended to have more variable ages at onset, thus providing support for the hypothesis.
Wolff et al. (1989) reported an isolated case of HD in an extensively studied family. Nonpaternity appeared to be excluded, and DNA markers closely linked to the HD gene indicated several clearly unaffected sibs who shared 1 or the other or both of the patient's haplotypes. The posterior probability of a new mutation to HD in the patient was calculated to exceed 99%, even if an a priori probability of nonpaternity of 10% and a mutation rate of HD of 1 in 10 million gametes were assumed.
Morrison et al. (1995) achieved virtually complete ascertainment of HD in Northern Ireland which, with a population of 1.5 million, showed a 1991 prevalence rate of 6.4/100,000. Estimates of heterozygote frequency gave values between 10 and 11 x 10(-5). The direct and indirect mutation rates were 0.32 x 10(-6) and 1.05 x 10(-6), respectively. Genetic fitness was increased in the affected HD population but decreased in the at-risk population. Fertility in HD was not reduced, but it appeared that at-risk persons had actively limited their family size. Factors responsible for this included, among others, the fear of developing HD and genetic counseling of families.
Farrer and Conneally (1985) postulated that age of onset is governed generally by a set of independently inherited aging genes, but expression of the HD genes may be significantly delayed in persons with a particular maternally transmitted factor. Myers et al. (1985) presented data that suggested a protective effect conferred on the offspring of affected women, who show an older mean age of onset than offspring of affected men, regardless of the onset age in the parent. Pointing out that some repetitive elements in many chromosomes of the mouse are methylated differently in males and females, Erickson (1985) suggested that such differences ('chromosomal imprinting') may be responsible for the greater severity (i.e., juvenile-onset) of Huntington disease in offspring of affected males and greater severity of myotonic dystrophy in offspring of affected females. Reik (1988) also suggested genomic imprinting as an alternative mechanism to maternally inherited extrachromosomal factors to account for the parental origin effect. By imprinting, the gene itself becomes modified in a different way depending on whether it passes through the maternal or the paternal germline. The modification may involve methylation of DNA and could result in earlier or higher level of expression of the gene when it is transmitted by the father.
Reik (1989) reviewed the topic of genomic imprinting in relation to genetic disorders of man, and as possible examples pointed to the earlier onset of spinocerebellar ataxia (164400) with paternal transmission, the increased severity of neurofibromatosis I (162200) with maternal transmission, the earlier onset of neurofibromatosis II (101000) with maternal transmission, and the preferential loss of maternal alleles in sporadic osteosarcoma (180200). Ridley et al. (1988) reviewed extensively the ascertainment bias producing or working against the observation of anticipation.
Hodge et al. (1980) excluded linkage with haptoglobin (reported by others) and found no positive lod scores for any of 14 other markers. The Huntington disease gene was assigned to chromosome 4 by demonstration of close linkage to an arbitrary (random) DNA segment (designated D4S10) which had been mapped to chromosome 4 by somatic cell hybridization (Gusella et al., 1984; Wexler et al., 1984). Gusella et al. (1984) found close linkage of an anonymous DNA segment to Huntington disease in a large Venezuelan kindred and a smaller American kindred. In the initial study, the total lod score was 8.53 at theta = 0.00. No obligatory recombinants were found. The DNA segment was detected by a sequence called G8 by the authors and renamed D4S10 at the seventh Human Gene Mapping Workshop in Los Angeles in August 1983. Linkage was with different haplotypes in the 2 kindreds studied. The upper limit of 99% confidence was set at 10 cM. D4S10 and HD were found to be remote from GC and MNS (known to be on 4q), as indicated by negative lod scores; thus, they may be on 4p. Gusella et al. (1984) identified further restriction enzyme polymorphism of the G8 probe found to be linked to HD; with this, the frequency of identifiable heterozygosity could be raised to about 90%. Folstein et al. (1985) found close linkage of HD and the G8 probe in both of the large Maryland kindreds reported earlier (Folstein et al., 1984).
The G8 locus and presumably the Huntington disease locus are deleted in the Wolf-Hirschhorn (4p-) syndrome (Gusella et al., 1985). This information helped map the HD locus to 4p. Most 4p- syndrome patients do not survive long enough to develop manifestations of HD. McKeown et al. (1987) found that the G8 locus was not deleted in a case of 4p- syndrome. By in situ hybridization (Wang et al., 1985; Magenis et al., 1985; Zabel et al., 1985; Wang et al., 1986), the HD-linked marker, G8, was mapped to 4p16.1. From studies by in situ hybridization to partially deleted chromosomes with known breakpoints, Magenis et al. (1986) concluded that the G8 probe is located in the distal half of band 4p16.1. Wang et al. (1986), also by in situ hybridization in patients with deletions of 4p, mapped G8 to 4p16.1-p16.3. Of their 2 patients, 1 had the typical phenotype of the Wolf-Hirschhorn syndrome (WHS) with a minute deletion of the segment p16.1-p16.3. Wang et al. (1986) concluded that the 4pter region could be excluded as a site.
Landegent et al. (1986) used a nonfluorescent method of in situ hybridization to assign the D4S10 locus to 4p16.3 rather than 4p16.1. The in situ hybridization method involved haptenization of nucleic acids in the probe by chemical attachment of 2-acetylaminofluorene (AAF) groups, marking of the hybridized probe by an indirect immunoperoxidase/diaminobenzidine reaction, and reflection-contrast microscopic visualization of the precipitated dye. Tranebjaerg et al. (1984) concluded that the 'critical segment' in Wolf syndrome is 4p16.3. Froster-Iskenius et al. (1986) described a kindred in which an apparently balanced reciprocal translocation between 4q and 5p was segregating together with Huntington disease in 2 generations. In situ hybridization studies revealed that the linked DNA marker (G8) was located in the region 4p16 of both the normal and translocated chromosome 4. Thus, the association may be a chance occurrence.
Collins et al. (1987) applied the strategy of chromosome jumping to identify new probes from the terminal portion of 4p. Jumping clones were identified that traveled in each direction from G8. In 2 of 3 persons recombinant for G8 and HD who were also informative for the newly identified probes, the jumping clone traveled with HD. Thus, a jump of approximately 200 kb had crossed 2 out of 3 recombination points between G8 and HD. The information defined unequivocally the location of HD distal to G8 and suggested that the physical distance between them may not be as large as previously suspected. Gilliam et al. (1987) presented evidence that the HD gene lies in 4p16.3 between D4S10 proximally and the telomere distally. Multipoint linkage analysis of the 4 loci--HD, D4S10, RAF2 (see 164760), and D4S62--indicated that D4S62 is close to D4S10 and centromeric to it. One particularly informative individual from the large Venezuelan kindred showed recombination between 2 RFLPs within the D4S10 segment. The 2 are located about 33 kb apart. The information at hand indicated the direction of cloning necessary for reaching the HD gene. Hayden et al. (1988) found that the polymorphic marker D4S62 is conserved in rat, mouse, and monkey liver and brain.
Gilliam et al. (1987) described an anonymous DNA segment, D4S43, which is exceedingly tightly linked to HD. Like the disease gene, it is located in the most distal portion of 4p, flanked by D4S10 and the telomere. In 3 extended HD kindreds, no recombination with HD was found, placing it less than 1.5 cM from the genetic defect. Expansion of the region to include 108 kb of cloned DNA led to the identification of 8 RFLPs and at least 2 independent coding segments. These genes might be candidates for the site of the HD defect; however, D4S43 RFLPs did not display linkage disequilibrium with the disease gene as one would expect if such were the case.
Wasmuth et al. (1988) characterized a new RFLP marker, D4S95, a highly polymorphic locus which displayed no recombination with HD in the families tested. Robbins et al. (1989) used genetic linkage analysis to demonstrate that the gene causing Huntington disease is telomeric to D4S95 and D4S90, both markers known to be tightly linked to the HD locus.
Doggett et al. (1989) prepared a physical map that extended from the most distal of the loci linked to HD (but proximal to HD) to the telomere of chromosome 4. The mapping identified at least 2 CpG islands and placed the most likely location of the HD defect remarkably close (within 325 kb) to the telomere. Conneally et al. (1989) pooled linkage data on G8 vs HD from 63 HD families (57 Caucasian, 4 Black American, and 2 Japanese). The combined maximum lod score was 87.69 at theta = 0.04 (99% confidence interval, 0.018-0.071). The maximum frequency of recombination was 0.03 in males and 0.05 in females. The data suggested that there is only 1 HD locus, though a second rare locus could not be ruled out. Kanazawa et al. (1990) presented linkage data in 9 Japanese families supporting the view that the Japanese Huntington disease gene is identical with the 'Western gene,' in spite of the lower prevalence rate in Japan. The linkage relationships appear to be the same as those that have been observed in European families.
Buetow et al. (1991) provided a genetic map of chromosome 4 with extensive information on the mapping of 4p16.3. They presented evidence for linkage heterogeneity in this region and suggested that it might explain the fact that in some families (Doggett et al., 1989; Robbins et al., 1989), HD has been localized to the most distal 325 kb of 4p16.3, telomeric to D4S90, the most distal marker in the map presented by Buetow et al. (1991), whereas in other families (MacDonald et al., 1989; Snell et al., 1989) HD has been localized proximal to D4S90. A microinversion in 4p16.3 in HD patients could provide an explanation. In 10 South African families of black, white, and mixed ancestry, Greenberg et al. (1991) found tight linkage to D4S10 (G8); maximum lod score = 8.14 at theta = 0.00. Because of the diverse ethnic backgrounds, the data provided evidence that there is only a single HD locus.
Bates et al. (1992) characterized a YAC contig spanning the region most likely to contain the HD mutation. Zuo et al. (1992) prepared a set of YAC clones spanning 2.2 Mb at the tip of the short arm of chromosome 4 presumably containing the HD gene. Skraastad et al. (1992) detected highly significant linkage disequilibrium with D4S95 in 45 Dutch families, consistent with studies in other populations. The area of linkage disequilibrium extended from D4S10 proximally to D4S95, covering 1,100 kb. The results confirmed the suggestion that the HD gene maps near D4S95. Using a direct cDNA selection strategy, Goldberg et al. (1993) identified at least 7 transcription units within the 2.2-Mb DNA interval thought to contain the HD gene. Screening with one of the cDNA clones identified an Alu insertion in genomic DNA from 2 persons with HD, which showed complete cosegregation with the disease in these families but was not found in 1,000 control chromosomes. A gene that encodes a 12-kb transcript, which maps in close proximity to the Alu insertion site, was considered a strong candidate for the HD gene.
In an analysis of 78 HD chromosomes with multiallelic markers, MacDonald et al. (1992) found 26 different haplotypes, suggesting a variety of independent HD mutations. The most frequent haplotype, accounting for about one-third of disease chromosomes, suggested that the disease gene is between D4S182 and D4S180. Alternative mechanisms for creating haplotype diversity do not require multiple mutational origin, however.
Stine and Smith (1990) studied the effects of mutation, migration, random drift, and selection on the changes in the frequency of genes associated with HD, porphyria variegata (176200), and lipoid proteinosis (247100) in the Afrikaner population of South Africa. By limiting analyses to pedigrees descendant from founding families, it was possible to exclude migration and new mutation as major sources of change. Calculations which overestimated the possible effect of random drift demonstrated that drift did not account for the changes. Therefore, these changes must have been caused by natural selection, and a coefficient of selection was estimated for each trait. A value of 0.34 was obtained for the coefficient of selection demonstrated by the HD gene, indicating a selective disadvantage rather than advantage suggested by some other studies.
Myers et al. (1989) performed molecular genetic studies in 4 offspring of 3 different affected x affected matings for possible homozygosity. One of the 4 was found to have a 95% likelihood of being an HD homozygote. The individual's age at onset and symptoms were similar to those in affected HD heterozygous relatives. Thus, the findings from the New England Huntington Disease Research Center corroborated the finding of Wexler et al. (1987). Harper et al. (1985) stated that the polymorphism with 4 enzymes (HindIII, EcoRI, NciI, and BstI) applied to the G8 locus shows that over 80% of subjects are heterozygous. They further stated that the latest estimate of the interval between the G8 and the HD loci was 5 cM. In 16 British kindreds, Youngman et al. (1986) found 2 recombinants yielding a maximum lod score of 17.6 at theta 0.02. This is more evidence against multilocus heterogeneity in Huntington disease.
Sabl and Laird (1992) suggested that dominant position-effect variegation (PEV) may be involved in HD. PEV is the variable but clonally stable inactivation of a euchromatic gene that has been placed adjacent to heterochromatic sequences. In an example in Drosophila melanogaster, a fully dominant mutant phenotype (and HD is such) results from stable epigenetic inactivation of an allele adjacent to the structural alteration (cis-inactivation) combined with a complementary inactivation of the homologous normal allele (trans-inactivation). Sabl and Laird (1992) proposed that the trans-inactivation of the normal allele may occasionally persist through meiosis. This so-called epigene conversion occurring at the HD locus in a few percent of meioses could account for anomalies in the region's genetic map.
On the basis of a review of the epidemiology of Huntington disease, Harper (1992) predicted that molecular studies in the future will show the following: more than one mutation has occurred at the HD locus. A very small number of mutations, possibly a single common one, will be found to account for most HD cases in populations of European origin. Any predominant mutation will probably have an extremely ancient origin, possibly dating back millenia. No single focus in northern Europe will be found as the point of origin of such a principal mutation. Phenotype will correlate poorly with specific mutations.
McLaughlin et al. (1996) reported that cytoplasmic protein extracts from several rat brain regions, including striatum and cortex (sites of neuronal degeneration in HD), contain a 63 kD RNA-binding protein that interacts specifically with CAG repeat sequences. They noted that the protein RNA interactions are dependent upon the length of the CAG repeat, and that longer repeats bind substantially more protein. McLaughlin et al. (1996) identified 2 CAG binding proteins in human cortex and striatum, one of 63 kD and another of 49 kD. They concluded that these data suggest mechanisms by which RNA binding proteins may be involved in the pathological course of trinucleotide-associated neurological diseases.
Pyrimidine oligodeoxyribonucleotides bind in the major groove of DNA parallel to the purine Watson-Crick strand through formation of specific Hoogsteen hydrogen bonds to the purine Watson-Crick base. Specificity is derived from thymine (T) recognition of adenine/thymine (AT) basepairs (TAT triplets); and N3-protonated cytosine (C+) recognition of guanine/cytosine (GC) basepairs (C + GC triplets). By combining oligonucleotide-directed recognition with enzymatic cleavage, near quantitative cleavage at a single target site can be achieved. Strobel et al. (1991) used this approach to 'liberate' the tip of 4p that contains the entire candidate region for the HD gene. A 16-base pyrimidine oligodeoxyribonucleotide was used with success.
The existence of many genes in the telomeric region of 4p is indicated by the work of Saccone et al. (1992). By chromosomal in situ hybridization, they determined the localization of the G+C-richest fraction of human DNA. Bernardi (1989) pointed out that the human genome is a mosaic of isochores, i.e., large DNA regions (more than 300 kb, on the average) that are compositionally homogeneous (above a size of 3 kb) and belong to a small number of families characterized by different G+C levels. The G+C-richest fraction of DNA has the highest gene concentration, the highest concentration of CpG islands, the highest transcriptional and recombinational activity, and a distinct chromatin structure. The in situ hybridization results showed a concentration of this isochore family, called H3, in telomeric bands and in chromomycin A3-positive/4-prime,6-diamidino-2-phenylindole-negative bands. Mouchiroud et al. (1991) found that the gene density in the GC-richest 3% of the genome is about 16 times higher than in the GC-poorest 62%. Figure 2 of Saccone et al. (1992) showed dramatically the concentration of G+C-rich DNA in the telomeric band of 4p as well as regions on other chromosomes that have been found to be rich in genes by mapping studies, e.g., distal 1p and much of chromosomes 19 and 22.
The Huntington's Disease Collaborative Research Group (1993) found that a 'new' gene, designated IT15 (important transcript 15) and later called huntingtin, which was isolated using cloned trapped exons from the target area, contains a polymorphic trinucleotide repeat that is expanded and unstable on HD chromosomes. A (CAG)n repeat longer than the normal range was observed on HD chromosomes from all 75 disease families examined. The families came from a variety of ethnic backgrounds and demonstrated a variety of 4p16.3 haplotypes. The (CAG)n repeat appeared to be located within the coding sequence of a predicted protein of about 348 kD that is widely expressed but unrelated to any known gene. Thus it turned out that the HD mutation involves an unstable DNA segment similar to those previously observed in several disorders, including the fragile X syndrome (309550), Kennedy syndrome (313200), and myotonic dystrophy. The fact that the phenotype of HD is completely dominant suggests that the disorder results from a gain-of-function mutation in which either the mRNA product or the protein product of the disease allele has some new property or is expressed inappropriately (Myers et al., 1989). (According to the tabulation of Parrish and Nelson (1993), HD was the 21st genetic disorder of previously unknown basic biochemical defect in which the gene was isolated by positional cloning. They reviewed the methods for finding genes and tabulated the methods used in each of the 21 disorders.)
MacDonald et al. (1993) found that unlike the similar CCG repeat in the fragile X syndrome, the expanded HD repeat shows no evidence of somatic instability in a comparison of blood, lymphoblast, and brain DNA from the same persons. Furthermore, 4 pairs of monozygotic HD twins displayed identical CAG repeat lengths, suggesting that repeat size is determined in gametogenesis. However, in contrast to the fragile X syndrome and with HD somatic tissue, mosaicism was readily detected as a diffuse spread of repeat lengths in DNA from HD sperm samples. Thus, the developmental timing of repeat instability appears to differ between HD and fragile X syndrome, indicating perhaps that the fundamental mechanisms leading to repeat expansion are distinct. Goldberg et al. (1993) reported findings in 3 families in which a new mutation for HD had arisen. In all 3 families, a parental intermediate allele (with expansion to 30-38 CAG repeats, greater than that seen in the population but below the range seen in patients with HD) had expanded in more than 1 offspring. In one of the families, 2 sibs with the expanded CAG repeat were clinically affected with HD, thus presenting a pseudorecessive pattern of inheritance.
Zuhlke et al. (1993) studied the length variation of the repeat in 513 non-HD chromosomes from normal individuals and HD patients; the group comprised 23 alleles with 11 to 33 repeats. In an analysis of the inheritance of the (CAG)n stretch, they found meiotic instability for HD alleles, (CAG)40 to (CAG)75, with a mutation frequency of approximately 70%; following the HD allele in 38 pedigrees during 54 meioses, they found a ratio of stable to altered copy number of 15:39. On the other hand, in 431 meioses of normal alleles, only 2 expansions were identified. They found that the risk of expansion during spermatogenesis was enhanced compared to oogenesis, explaining juvenile onset by transmission from affected fathers. No mosaicism or differences in repeat lengths were observed in the DNA from different tissues, including brain and lymphocytes of 2 HD patients, indicating mitotic stability of the mutation. Thus, the determination of the repeat number in the DNA of blood lymphocytes is probably representative of all tissues in a patient.
Duyao et al. (1993), Snell et al. (1993), and Andrew et al. (1993) analyzed the number of CAG repeats in a total of about 1,200 HD genes and in over 2,000 normal controls. Read (1993) summarized and collated the results. In all 3 studies the normal range of repeat numbers was 9-11 at the low and 34-37 at the high end, with a mean ranging from 18.29 to 19.71. Duyao et al. (1993) found a range of 37-86 in HD patients with a mean of 46.42. Snell et al. (1993) found a negative correlation between the number of repeats on the normal paternal allele and the age of onset in individuals with maternally transmitted disease. They interpreted this as suggesting that normal gene function varies because of the size of the repeat in the normal range and a sex-specific modifying effect. Read (1993) commented that this was not seen by the other groups and 'is hard to square with the reported normal age of onset in homozygotes.' In a study of the HD mutation and the characteristics of its transmission in 36 HD families, Trottier et al. (1994) found that instability of the CAG repeats was more frequent and stronger upon transmission from a male than from a female, with a clear tendency toward increased size. They found a significant inverse correlation (p = 0.0001) between the age of onset and the CAG repeat length. The observed scatter would, however, not allow an accurate individual prediction of age of onset. An HD mutation of paternal origin was found in 3 juvenile-onset cases analyzed. In at least 2 of these cases, a large expansion of the HD allele upon paternal transmission may explain the major anticipation observed. Illarioshkin et al. (1994) found significant positive correlation between the rate of progression of clinical symptoms and CAG repeat length in a group of 28 Russian patients with Huntington disease. Ranen et al. (1995) found that the change in repeat length with paternal transmission was significantly correlated with the change in age at onset between the father and offspring. They confirmed an inverse relationship between repeat length and age at onset, the higher frequency of juvenile-onset cases arising from paternal transmission, anticipation as a phenomenon of paternal transmission, and greater expansion of the trinucleotide repeat with paternal transmission.
Andrew et al. (1994) found that 30 of 1,022 persons with HD (2.9%) did not have an expanded CAG repeat in the disease range. They showed that most of these individuals with normal sized alleles, namely 18, represented misdiagnosis, sample mix-up, or clerical error. The remaining 12 patients represented possible phenocopies for HD. In at least 4 cases, family studies of these phenocopies excluded 4p16.3 as the region responsible for the phenotype. Mutations in the HD gene other than CAG expansion have not been excluded for the remaining 8 cases; however, in as many as 7 of these patients, retrospective review of their clinical features identified characteristics not typical for HD. Andrew et al. (1994) concluded that on rare occasions mutations in other, as-yet-undefined genes can present with a clinical phenotype very similar to that of HD.
Rubinsztein et al. (1996) studied a large cohort of individuals who carried between 30 and 40 CAG repeats in the IT15 gene. They used a PCR method that allowed the examination of CAG repeats only, thereby excluding the CCG repeats, which represent a polymorphism, as a confounding factor. No individual with 35 or fewer CAG repeats had clinical manifestations of HD. Most individuals with 36 to 39 CAG repeats were clinically affected, but 10 persons (aged 67-95 years) had no apparent symptoms of HD. The authors concluded that the HD mutation is not fully penetrant in individuals with a borderline number of CAG repeats.
Ambrose et al. (1994) found that the HD locus spans 180 kb and consists of 67 exons ranging in size from 48 bp to 341 bp with an average of 138 bp. A codon loss polymorphism in linkage disequilibrium with the disorder revealed that both normal and HD alleles are represented in the mRNA population in HD heterozygotes, indicating that the defect does not eliminate transcription. The gene is ubiquitously expressed as 2 alternatively polyadenylated forms displaying different relative abundance in various fetal and adult tissues, suggesting the operation of interacting factors in determining specificity of cell loss. In a female carrying a balanced translocation with a breakpoint between exons 40 and 41, the HD gene was disrupted but the phenotype was normal, arguing against simple inactivation of the gene as the mechanism by which the expanded trinucleotide repeat causes HD. The observation suggested that the dominant HD mutation either confers a new property on the mRNA or, more likely, alters an interaction at the protein level.
The wide expression of the HD transcript does not correlate with the pattern of neuropathology in the disease. To study the HD gene product (huntingtin), Trottier et al. (1995) generated monoclonal antibodies against 4 different regions of the protein. On Western blots, these monoclonals detected the huntingtin protein of approximately 350 kD in various human cell lines and in neural and nonneural rodent tissues. In cell lines from HD patients, a doublet protein was detected corresponding to the mutant and normal huntingtin. Immunohistochemical studies in the human brain, using 2 of these antibodies, detected huntingtin in perikarya of some neurons, neuropiles, and varicosities. Huntingtin was also visualized as punctate staining likely to represent nerve endings.
Leeflang et al. (1995) amplified the CAG triplet repeat region of the HD gene in 923 single sperm from 3 affected and 2 normal individuals. Average-sized alleles (15-18 repeats) showed only 3 contraction mutations among 475 sperm (0.6%). A 30-repeat normal allele showed an 11% mutation frequency. The mutation frequency of a 36-repeat intermediate allele was 53% with 8% of all gametes having expansions that brought the allele size into the HD disease range (38 repeats or more). Disease alleles (38-51 repeats) showed a very high mutation frequency (92-99%). As repeat number increased, the authors found a marked elevation in the frequency of expansions, in the mean number of repeats added per expansion, and in the size of the largest observed expansion. Contraction frequencies also appeared to increase with allele size but decreased as repeat number exceeded 36. Since the sperm typing data were of a discrete nature rather than consisting of smears of PCR products from pooled sperm, Leeflang et al. (1995) could compare the observed mutation frequency spectra to the distribution calculated using discrete stochastic models based on current molecular ideas of the expansion process. An excellent fit was found when the model specified that a random number of repeats are added during the progression of the DNA polymerase through the repeated region.
Gutekunst et al. (1995) used both polyclonal and monoclonal anti-fusion protein antibodies to identify native huntingtin in rat, monkey, and human. Western blots revealed a protein with the expected molecular weight that is present in the soluble fraction of rat and monkey brain tissues and lymphoblastoid cell lines from control cases. In lymphoblastoid cell lines from juvenile-onset heterozygote HD cases, both normal and mutant huntingtin was expressed, and increasing repeat expansion leads to lower levels of the mutant protein. Immunocytochemistry indicated that huntingtin is located in neurons throughout the brain, with the highest levels evident in larger neurons. In the human striatum, huntingtin was enriched in a patch-like distribution, potentially corresponding to the first areas affected in HD. Subcellular localization of huntingtin was consistent with a cytosolic protein primarily found in somatodendritic regions. Huntingtin appears to be associated particularly with microtubules, although some is also associated with synaptic vesicles. On the basis of the localization of huntingtin in association with microtubules, Gutekunst et al. (1995) speculated that the mutation impairs the cytoskeletal anchoring or transport of mitochondria, vesicles, or other organelles or molecules.
Li et al. (1995) described a huntingtin associated protein (HAP1; 600947) which is enriched in brain. The authors found that binding of HAP1 to huntingtin is enhanced by an expanded polyglutamate repeat.
Burke et al. (1996) described the isolation of a protein present in brain homogenates that bound to a synthetic 60-glutamine peptide (such as that found huntingtin). Eighteen amino acids of this protein were found to be identical to the amino terminus of glyceraldehyde-3-phosphate dehydrogenase (GAPD; 138400). GAPD was also found to bind to another protein with a polyglutamine tract, namely the DRPLA protein (125370). Burke et al. (1996) demonstrated that synthetic polyglutamine peptides, DRPLA protein, and huntingtin from unaffected individuals with normal-sized polyglutamine tracts bind to GAPD. GAPD had also been shown to bind to RNA, ATP, calcyclin (114110), actin (see 102610), tubulin (see 191130) and amyloid precursor protein (104760). On the basis of their findings, the authors postulated that the diseases characterized by the presence of an expanded CAG repeat, which share a common mode of heritability, may also share a common metabolic pathogenesis involving GAPD as a functional component. Both Roses (1996) and Barinaga (1996) reviewed these findings.
Portera-Cailliau et al. (1995) among others presented evidence that apoptosis is a mode of cell death in Huntington disease. Apopain (600636), a human counterpart of the nematode cysteine protease death-gene product (CED-3) has a key role in proteolytic events leading to apoptosis. Goldberg et al. (1996) showed that apoptotic extracts, and apopain itself, specifically, cleave huntingtin. The rate of cleavage increased with the length of the huntingtin polyglutamine tract, providing an explanation for the gain-of-function associated with CAG expansion. The results suggested to the investigators that HD may be a disorder of inappropriate apoptosis.
As outlined earlier, all mutations for Huntington disease (amplification of the CAG trinucleotide repeat in the coding region of the gene) arise from so-called intermediate alleles (IAs) containing between 29 and 35 CAG repeats. The CAG repeats expand on transmission through the paternal germline to 36 or more repeats. Intermediate alleles are present on approximately 1% of normal chromosomes of Caucasian descent. Affected individuals have an expanded allele of between 36 to 121 CAGs but incomplete penetrance has been found for repeat lengths of 36 to 40 CAGs. Using single sperm analysis, Chong et al. (1997) assessed CAG mutation frequencies of 4 IAs in families with sporadic HD and IAs ascertained from the general population by analyzing 1161 single sperm from 3 persons. They showed that the intermediate alleles of the former group were more unstable than those in the general population with identical size and sequence. Furthermore, comparison of different sized IAs and IAs with different sequences between the CAG and the adjacent CCG tracts indicated that DNA sequence is a major influence on CAG stability. These studies provided estimates of the likelihood of expansion to 36 or more CAG repeats for individuals in the 2 groups. For an IA with (CAG)35 in the family with sporadic HD, the likelihood for sibs to inherit a recurrent mutation equal to or more than (CAG)36 was approximately 10%. For intermediate alleles of a similar size in the general population, the risk of inheriting an expanded allele of 36 or more CAGs through the paternal germline was approximately 6%.
Harper and Sarfarazi (1985) pointed out that predictive testing can be done in prenatal diagnosis without determining the status of the at-risk parent. For example, if the affected grandparent of the fetus is deceased, the other grandparent is genotype BB, and the parent at risk is AB married to a CC individual, the fetus is unlikely to have inherited HD if it is BC, while the risk is 50% if the fetus is AC. The likelihood of the BC fetus being affected is a function of recombination. Bloch and Hayden (1987) pointed out that this 'no news' or 'good news' option has some important consequences. The 'no news' outcome increases the risk of the fetus's having inherited the gene for HD from 25% to about 50%; thus, persons given this information may need longterm support. Also, the implication of linking the status of an at-risk child to that of the at-risk parent may be more serious than realized. Millan et al. (1989) pointed out the importance of not acquiring more information than necessary to exclude or include the diagnosis of HD in a fetus. In a family they studied, the probability of the fetus being affected, approaching 50%, could be deduced from the genotype of the fetus, the 2 parents, and the unaffected paternal grandfather of the conceptus. Genotyping of the unaffected maternal grandmother of the father refined downward somewhat (from 47 to 42%) the risk of HD in the conceptus; however, it ran the risk of making the diagnosis of HD in the father and the information was really unnecessary for genetic counseling. Information about the prenatal exclusion test for HD was given to an unselected series of couples who attended a genetic counseling clinic in Glasgow from 1986 onwards. Ten couples underwent 13 prenatal tests during this period with expressed intention of stopping a pregnancy if the results indicated a high risk (almost 50%) that the fetus carried the HD gene. Although 9 fetuses at nearly 50% risk of carrying the HD gene were identified, only 6 such pregnancies were terminated. In each of the 3 high-risk pregnancies that continued, the mother made a 'final hour' decision not to undergo the scheduled, first-trimester termination. In the experience of Tolmie et al. (1995), who reported these results, late reversal of a previous decision to undergo first-trimester pregnancy termination for a genetic indication was frequent among couples who had undergone the prenatal exclusion test for HD.
Bloch and Hayden (1990) opposed the testing of children at risk for Huntington disease and questioned the usefulness of DNA tests to support a diagnosis of HD in either adulthood or childhood. They opposed testing in adoption cases because of the negative effects on the child's upbringing and education as well as the necessity to adhere to the principle of autonomy on the part of the individual tested. Prenatal testing was undertaken in their practice only if the parents were prepared to make a decision about continuing the pregnancy on the basis of the outcome of the prenatal testing. The parents were given to understand that prenatal testing is similar to testing a minor child. In the program of Bloch and Hayden (1990), 8 exclusion prenatal tests had been performed, with 5 resulting in an increased risk for the fetus. In 4 of these, the parents decided to terminate the pregnancy.
Wiggins et al. (1992) reported on the psychological consequences of predictive testing for HD on the basis of observations in 135 participants in the Canadian program of genetic testing. The participants were in 3 groups according to their test results: the increased-risk group (37 persons); the decreased-risk group (58 persons); and the group with no change in risk (40 persons). They showed that predictive testing had benefits for the psychological health of persons who received results that indicated either an increase or a decrease in the risk of inheriting the gene. In an accompanying editorial, Catherine V. Hayes (1992), president of the Huntington's Disease Society of America, described what it meant to grow up as an 'at-risk' person and to have genetic testing.
Quarrell et al. (1987) suggested the usefulness of the G8 marker in exclusion testing for HD. They cited studies of 52 families from various parts of the world, indicating a maximum total lod score of 75.3 at a recombination fraction of about 5 cM. The 95% confidence intervals were 2.4 and 6.5 cM, with no evidence of multilocus heterogeneity. The marker could be applied either for presymptomatic predictive testing or for exclusion testing in pregnancy, where the estimated risk to the parent is not altered. The requirements for family structure were much less stringent in the case of exclusion testing. In South Wales they found that nearly 90% of couples have the minimum structure required for an exclusion test, whereas for a presymptomatic predictive test only 15% have the ideal 3-generation family structure and only 10% have a suitably extended 2-generation family. The distribution of G8 haplotypes presented the same difficulty whichever test was being considered; only about two-thirds of couples would be informative. If the fetus acquired the G8 haplotype of the affected grandparent, then the risk to the fetus was the same as that of the parent, i.e., 50%. If the fetus has the G8 haplotype of the unaffected grandparent, then the risk to the fetus became 2.5%. If termination of pregnancy was unacceptable despite an adverse result of the test and HD subsequently developed in the parent in generation 2, it would be immediately known that HD would also be likely to arise in the offspring since their risks are the same (apart from the possibility of recombination). To prevent this complication, Quarrell et al. (1987) told couples that if termination of pregnancy was unacceptable for whatever reason, then an exclusion test would be inappropriate.
Tyler et al. (1990) reported on extensive experience with exclusion testing. They pointed out that exclusion testing is often possible even though the risk-status of the parent cannot be ascertained. The problems with exclusion testing seem rather large. The experience with this approach, as with presymptomatic diagnosis in general, will be valuable background for planning practice when completely foolproof methods of diagnosis are available. Early results of predictive testing using D4S10 RFLPs were reported by Meissen et al. (1988). MacDonald et al. (1989) characterized genetically 5 highly informative multiallele RFLPs of value in the presymptomatic diagnosis of HD. Morris et al. (1989) and Craufurd et al. (1989) outlined problems associated with programs for presymptomatic predictive testing for HD. Morris et al. (1989) and Craufurd et al. (1989) outlined problems associated with programs for presymptomatic predictive testing for HD.
Read (1993) commented that the problems arising in connection with HD testing resembled those of HIV testing. The 10 years during which testing for HD required family studies have given clinical geneticists an opportunity to work out proper procedures. A great deal of effort has gone into ensuring that presymptomatic testing is always voluntary and is undertaken only after due consideration by fully informed patients. Testing of children has been firmly discouraged. It is vital that these practices should be continued.
Kremer et al. (1994) reported a worldwide study assessing the sensitivity and specificity of the CAG expansion as a diagnostic test. The study covered 565 families from 43 national and ethnic groups containing 1,007 patients with signs and symptoms compatible with the diagnosis of HD. Of these, 995 had an expanded CAG repeat that included from 36 to 121 repeats; sensitivity = 98.8%, with 95% confidence limits = 97.7-99.4. Included among those contributing to the sensitivity estimate were 12 patients with previously diagnosed HD in whom the number of CAG repeats was in the normal range. Reevaluation of these established that 11 had clinical features atypical of HD. In 1,581 of 1,595 control chromosomes (99.1%), the number of CAG repeats ranged from 10 to 29. The remaining 14 control chromosomes had 30 or more repeats, with 2 of these chromosomes having expansions of 37 and 39 repeats. An estimate of specificity was made from 113 subjects with other neuropsychiatric disorders with which HD is frequently confused. The number of repeats found in these disorders was similar to the number found on normal human chromosomes and showed no overlap with HD; specificity = 100%, with 95% CI = 95.5-100. The study confirmed that CAG expansion is the molecular basis of HD worldwide. Warner et al. (1994) searched for possible missed cases of Huntington disease in a set of 368 patients with psychiatric disorders, including schizophrenia, presenile dementia, and senile dementia. One schizophrenic patient, who died at age 88, had a CAG repeat size of 36; a 68-year-old patient, who died of presenile dementia of Alzheimer disease type, had a CAG repeat size of 34. Neither patient had neuropathologic or clinical evidence of Huntington disease.
Gellera et al. (1996) reported that ideally a series of 3 PCR reactions should be performed to rule out Huntington disease. They reviewed the evidence that the huntingtin gene contains an unstable polyglutamine-encoding (CAG)n repeat which is located in the N-terminal portion of the protein beginning 18 codons downstream of the first ATG codon (143100.0001). The unstable (CAG)n repeat lies immediately upstream from a moderately polymorphic polyproline encoding (CCG)n repeat. Gellera et al. (1996) noted further that a number of reports in the literature indicated that in normal subjects the number of (CAG)n polyglutamine repeats ranges from 10 to 36, while in HD patients it ranges from 37 to 100. The (CCG)n polyproline repeat may vary in size between 7 and 12 repeats in both affected and normal individuals. They reported the occurrence of a CAA trinucleotide deletion (nucleotides 433-435) in HD chromosomes in 2 families that, because of its position within the conventional antisense primer hd447, hampered HD mutation detection if only the (CAG)n tract were amplified. Therefore, Gellera et al. (1996) stressed the importance of using a series of 3 diagnostic PCR reactions: one which amplified the (CAG)n tract alone, one which amplified the (CCG)n tract alone, and one which amplified the whole region.
New mutations are probably rare. Bundey (1983) concluded 'that it is incorrect to say that new mutations for Huntington's chorea occur in less than 0.1% of sufferers. I believe the evidence shows that the true figure is nearer 10%. I therefore consider that the absence of a known affected relative should not deter a neurologist from diagnosing Huntington's chorea in a patient who shows the characteristic clinical features of the disease.' She based her conclusion particularly on estimates of fitness and the Haldane formula for estimating proportion of new mutation cases. However, Mastromauro et al. (1989) could find no evidence of difference in fitness of HD-affected persons from their unaffected sibs or from the general population of Massachusetts.
Palo et al. (1987) estimated the frequency of HD in Finland to be 5 cases per million as contrasted with frequencies of 30 to 70 per million in most Western countries. The lowest frequencies have been found in South African blacks (0.6), in Japan (3.8), and in North American blacks (15). The findings in Finland are consistent with almost all cases having originated from a single source and illustrate founder effect, which is shown by so many other diseases in that country. For example, PKU (261600) has been found in only 5 cases over all time, whereas aspartylglycosaminuria (208400) has been identified in almost 200 living cases in a population of 4.9 million. The part of Finland that is an exception to the above statement is the Aland archipelago where the frequency of HD is high, but this is an exception that proves the rule: the islands have been exposed to other populations (including the British) for centuries. In Finland, Ikonen et al. (1992) reported further studies by RFLP haplotype analysis in combination with genealogic study of all the Finnish HD families. They found that a high percentage (28%) of the families had foreign ancestors. Furthermore, most of the Finnish ancestors were localized to border regions or trade centers of the country, following the old postal routes. The observed high-risk haplotypes formed with markers from the D4S10 and D4S43 loci were evenly distributed among the HD families in different geographic locations. Ikonen et al. (1992) concluded that the HD gene(s) probably arrived in Finland on several occasions via foreign immigrants.
Almqvist et al. (1994) constructed haplotypes for 23 different HD families, 10% of the 233 known HD families in the Swedish Huntington disease register. Ten different haplotypes were observed. Analysis of 2 polymorphic markers within the HD gene indicated that there are at least 3 origins of the HD mutation in Sweden. One of the haplotypes accounted for 89% of the families, suggesting descent from a single ancestor.
Leung et al. (1992) stated that the prevalence of HD in Hong Kong Chinese for the period 1984-1991 was 3.7 per million. They traced the ancestral origin of the patients mainly to the coastal provinces and proposed that Chinese HD had a European origin. They found a male preponderance: 63 males to 26 females. They made no comment on the provinces of origin of the Hong Kong Chinese population generally.
Rubinsztein et al. (1994) investigated the evolution of HD by typing CAG alleles from 5 different human populations and 10 different species of primates. Using computer simulations, they found that human alleles have expanded from a shorter primate ancestor and exhibit unusual asymmetric length distributions. Suggesting that the key element in HD evolution is a simple length-dependent mutational bias toward longer alleles, they predicted that, in the absence of interference, expansion of trinucleotide repeats will continue and accelerate, leading to an ever-increasing incidence of HD. Masuda et al. (1995) demonstrated that the size of the CAG repeat in Japanese HD patients range from 37 to 95 repeats, as compared with a range from 7 to 29 in normal controls. Whereas HD chromosomes in the west are strongly associated with the (CCG)7 repeat, immediately 3-prime adjacent to the CAG repeat, Japanese HD chromosomes were found to be in strong linkage disequilibrium with the (CCG)10 repeat. The frequency of HD in Japan is less than one-tenth of the prevalence in western countries. It had been suggested that the low frequency reflected western European origin with spread to Japan by immigration. The haplotype findings concerning the association of the CAG repeat the the CCG repeat suggests a separate origin with founder effect in the Japanese cases.
Scrimgeour et al. (1995) described a case of apparently typical HD in a 40-year-old Sudanese man from Khartoum, in whom the HD gene showed 51 CAG repeats. It was suspected that his mother and his deceased 16-year-old son were also affected.
Grosson et al. (1994) localized the mouse homologs of the HD gene and 17 other human chromosome 4 loci, including 6 previously unmapped genes, by use of an interspecific cross. All loci mapped in a continuous linkage group on mouse chromosome 5, distal to En2 (engrailed-2; 131310) and Il6 (interleukin-6; 147620), the human counterparts of which are located on chromosome 7. The relative order of the loci on human chromosome 4 and mouse chromosome 5 was maintained for the most part. Grosson et al. (1994) knew of no phenotypic correspondence between human and mouse mutations mapping to this region of synteny conservation. The gene that is mutant in achondroplasia (100800), namely, fibroblast growth factor receptor-3 (FGFR3; 134934), was not among the genes mapped. Are there skeletal dysplasias in the mouse that map to that region of chromosome 5?
Lin et al. (1995) cloned the mouse homolog and showed that it maps to mouse chromosome 5 within a region of conserved synteny with human 4p16.3. They also presented a detailed comparison of the sequence of the putative promoter and the organization of the 5-prime genomic region encompassing the first 5 exons of the mouse Hdh and human HD genes. They found 2 dinucleotide (CT) and 1 trinucleotide intronic polymorphism in Hdh and an intronic CA polymorphism in the HD gene. A comparison of 940-bp sequence 5-prime to the putative translation start site revealed a highly conserved region (78.8% nucleotide identity) between Hdh and the HD gene from nucleotide -56 to -206 (of Hdh).
To understand the normal function of the HD gene, Nasir et al. (1995) created a targeted disruption in exon 5 of Hdh, the murine homolog of the HD gene, using homologous recombination. They found that homozygotes died before embryonic day 8.5, initiate gastrulation, but do not proceed to the formation of somites or to organogenesis. Mice heterozygous for the mutation displayed increased motor activity and cognitive deficits. Neuropathologic assessment of 2 heterozygous mice showed a significant neuronal loss in the subthalamic nucleus. These studies showed that the HD gene is essential for postimplantation development and that it may play an important role in normal functioning of the basal ganglia.
To distinguish between 'loss of function' and 'gain of function' models of HD, Duyao et al. (1995) inactivated the mouse Hdh by gene targeting. Mice heterozygous for Hdh inactivation were phenotypically normal, whereas homozygosity resulted in embryonic death. Homozygotes displayed abnormal gastrulation at embryonic day 7.5 and were resorbing by day 8.5. Thus, they concluded that huntingtin is critical early in embryonic development, before the emergence of the nervous system. That Hdh inactivation does not mimic adult HD neuropathology suggested to the authors that the human disease involves a gain of function.
Zeitlin et al. (1995) also used targeted gene disruption of Hdh and found that 'knockout' mice nullizygous for the Hdh gene showed developmental retardation and disorganization as embryos and died between days 8.5 and 10.5 of gestation. Based on the observation that the level of the regionalized apoptotic cell death in the embryonic ectoderm, a layer expressing the Hdh gene, was much higher than normal in the null mutants, Zeitlin et al. (1995) proposed that huntingtin is involved in processes counterbalancing the operation of an apoptotic pathway.
To identify the functionally important domains of the HD protein, Baxendale et al. (1995) cloned and sequenced the homolog of the HD gene in the pufferfish, Fugu rubripes. The Fugu HD gene spans only 23 kb of genomic DNA, compared to the 170-kb human gene, and yet all 67 exons are conserved. The first exon, the site of the disease-causing triplet repeat in the human, is highly conserved. However, the glutamine repeat in Fugu consists of only 4 residues. Baxendale et al. (1995) also showed that synteny may be conserved over longer stretches of the 2 genomes. The work described a detailed example of sequence comparison between human and Fugu and illustrated the power of the pufferfish genome as a model system in the analysis of human genes.
To investigate the role of the CAG expansion in the first exon of the HD gene in the pathogenesis of the disorder, Goldberg et al. (1996) produced transgenic mice containing the full-length human HD cDNA with 44 CAG repeats. By 1 year, these mice had no behavioral abnormalities; morphometric analysis at 6 months in 1 animal and at 9 months in 2 animals revealed no changes. Despite high levels of mRNA expression, there was no evidence of the HD gene product in any of these transgenic mice. In vitro transfection studies indicated that the inclusion of 120 bp of the 5-prime untranslated region into the cDNA construct and the presence of a frameshift mutation at nucleotide 2349 prevented expression of the HD cDNA. These findings suggested to Goldberg et al. (1996) that the pathogenesis of HD is not mediated through DNA-protein interaction and that presence of the RNA transcript with an expanded CAG repeat is insufficient to cause the disease. Rather, translation of the CAG is crucial for the pathogenesis of HD. In contrast to the situation in humans, the CAG repeat in these mice was remarkably stable in 97 meioses. This suggested to them that other genomic sequences may play a critical role in influencing repeat instability.
Mangiarini et al. (1996) generated mice transgenic for the 5-prime end of the human HD gene, including promoter sequences and exon 1 carrying (CAG)n expansions of approximately 130 residues. In 3 mouse lines, the transgene was ubiquitously expressed at both the mRNA and protein levels. Transgenic mice exhibited a progressive neurologic phenotype that exhibited many of the features of HD, including choreiform movements, involuntary stereotypic movements, tremor, and epileptic seizures, as well as nonmovement disorder components.
Mangiarini et al. (1997) examined the behavior of the CAG repeat in mice transgenic for the HD mutation. They noted that the trinucleotide repeat is unstable during transmission and somatogenesis. Similar studies of intergenerational and somatic cell instability were found with the myotonic dystrophy (DM) CTG repeat in transgenic mice (160900). In studies of both of these repeats, the mutability of the repeats was high, although the instability (in terms of repeat length increases) was modest, showing fluctuations of only a few repeats. The somatic instability of the repeats increased with the age of the mice and appeared to occur in different tissues (perhaps correlating with the level of expression of the transgene in particular tissues or cells). Both expansions and deletions were seen in transgenic repeats, with a tendency toward expansion upon male transmission and contraction upon female transmission.
Hodgson et al. (1996) reported results of their studies designed to rescue the embryonic lethality phenotype that results from targeted disruption of the murine HD gene. They generated viable offspring that were homozygous for the disrupted murine HD gene and that expressed human huntingtin derived from a YAC transgene. These results indicated that the YAC transgene was expressed prior to 7.5 days' gestation and that the human huntingtin protein was functional in a murine background.
Durbach and Hayden (1993) published a personal account of George Huntington based on unpublished sources and communications from several of his descendants. Their account provides insight into his role as a general practitioner, literally a 'horse-and-buggy doctor' as demonstrated by one of the figures, as well as indicating his avocations of sketching, hunting, and fishing.
Van der Weiden (1989) gave a biographical account of George Huntington (1850-1916) and of the American anatomist George Sumner Huntington (1861-1927), and pointed out that biographical data on the 2 have been confused repeatedly.
EDITOR'S NOTE
Huntington disease represents a classic ethical dilemma created by the human genome project, i.e., that of the widened gap between what we know how to diagnose and what we know how to do anything about. Wexler (1992) referred to the dilemma as the Tiresias complex. The blind sear Tiresias confronted Oedipus with the dilemma: 'It is but sorrow to be wise when wisdom profits not' (from Oedipus the King by Sophocles). Wexler (1992) stated the questions as follows: 'Do you want to know how and when you are going to die, especially if you have no power to change the outcome? Should such knowledge be made freely available? How does a person choose to learn this momentous information? How does one cope with the answer?'
Gellera et al. (1996) noted that the unstable (CAG)n repeat lies immediately upstream from a moderately polymorphic polyproline-encoding (CCG)n repeat. They noted further that a number of reports in the literature indicated that in normal subjects the number of (CAG)n repeats ranges from 10 to 36, while in HD patients it ranges from 37 to 100. The downstream (CCG)n repeat may vary in size between 7 and 12 repeats in both affected and normal individuals. They reported the occurrence of a CAA trinucleotide deletion (nucleotides 433-435) in HD chromosomes in 2 families that, because of its position within the conventional antisense primer hd447, hampered HD mutation detection if only the (CAG)n tract were amplified. Therefore, Gellera et al. (1996) stressed the importance of using a series of 3 diagnostic PCR reactions: one which amplified the (CAG)n tract alone, one which amplified the (CCG)n tract alone, and one which amplified the whole region.