Are hereditary hemochromatosis mutations involved in Alzheimer disease?
Abstract
Mutations in the class I-like major histocompatibility complex gene called HFE are associated with hereditary hemochromatosis (HHC), a disorder of excessive iron uptake. We screened DNA samples from patients with familial Alzheimer disease (FAD) (n = 26), adults with Down syndrome (DS) (n = 50), and older (n = 41) and younger (n = 52) healthy normal individuals, for two HHC point mutations—C282Y and H63D. Because the apolipoprotein E (ApoE) E4 allele is a risk factor for AD and possibly also for dementia of the AD type in DS, DNA samples were also ApoE genotyped. Chi-squared analyses were interpreted at the 0.05 level of significance without Bonferroni corrections. In the pooled healthy normal individuals, C282Y was negatively associated with ApoE E4, an effect also apparent in individuals with DS but not with FAD. Relative to older normals, ApoE E4 was overrepresented in both males and females with FAD, consistent with ApoE E4 being a risk factor for AD; HFE mutations were overrepresented in males and underrepresented in females with FAD. Strong gender effects on the distribution of HFE mutations were apparent in comparisons among ApoE E4 negative individuals in the FAD and healthy normal groups (P < 0.002). Our findings are consistent with the proposition that among ApoE E4 negative individuals HFE mutations are predisposing to FAD in males but are somewhat protective in females. Further, ApoE E4 effects in our FAD group are strongest in females lacking HFE mutations. Relative to younger normals there was a tendency for ApoE E4 and H63D to be overrepresented in males and underrepresented in females with DS. The possibility that HFE mutations are important new genetic risk factors for AD should be pursued further. Am J. Med. Genet. 93:58–66, 2000. © 2000 Wiley-Liss, Inc.
INTRODUCTION
Hereditary hemochromatosis (HHC) is an autosomal recessive disorder of excessive iron uptake, deposition, and damage to the liver, pancreas, heart, spleen, joints, and endocrine glands [Bothwell and MacPhail, 1998; Jazwinska, 1998; Yang et al., 1999]. Point mutations have been discovered in individuals with HHC in a class I-like major histocompatibility locus gene on chromosome 6 designated as HFE [Wain et al., 2000]. One HFE mutation associated with HHC is the cysteine-to-tyrosine substitution (C282Y), and another is the histidine-to-aspartic acid substitution (H63D) [Feder et al., 1996, 1998]. Several other less frequent HFE mutations associated with HHC also have been described [Barton et al., 1999; Wallace et al., 1999]. Homozygosity for C282Y accounts for about 80% of cases of classical HHC in Caucasians of northern European descent [Merryweather-Clarke et al., 1997; Marshall et al., 1999]. The role of H63D in HHC is not clear. There appears to be a modestly increased risk of HHC associated with homozygosity for H63D and with C282Y/H63D compound heterozygosity [Risch, 1997]. H63D may be associated with nonclassical iron overload [Porto et al., 1998].
The wild-type HFE protein binds to transferrin receptor and reduces its affinity for iron-loaded transferrin at the basic pH of the cell surface [Lebron et al., 1998]. The H63D mutant protein does not have this effect. The C282Y mutation almost completely prevents the association of the mutant HFE protein with transferrin receptor [Feder et al., 1998]. These observations suggest mechanisms of how the C282Y and H63D mutations result in increased iron uptake in cells.
There is a strong rationale for looking at HFE mutations in Alzheimer disease (AD). Many studies have described an association of disregulation of iron metabolism and an association of redox active iron with AD [Kedziora et al., 1978; Grundke-Iqbal et al., 1990; Kuiper et al., 1994; Leveugle et al., 1994; Kennard et al., 1996; Fischer et al., 1997; LeVine, 1997; Smith et al., 1997, 1998; Percy et al., 1998b, 1999]. Furthermore, heterozygosity for HFE mutations is reported to be a risk factor for cardiovascular disease [Roest et al., 1999; Tuomainen et al., 1999] which also is associated with AD [Stewart et al., 1999]. To our knowledge, no studies of HFE mutations in AD are published. Studies of HFE mutations also should be extended to Down syndrome (DS), because of an unusual association of dementia with DS. Adults with DS are at risk of developing dementia of the AD type (DAT) 30–40 years earlier than in the general population [Holland et al., 1998; Oliver et al., 1998; Janicki and Dalton, 1999]. Virtually all with complete trisomy 21 over the age of 35 or 40 develop brain plaques and tangles resembling those in AD, although the extent of the pathology is variable [Mann et al., 1990; Egensperger et al., 1999]. There is also evidence for aberrations of iron metabolism in DS and an association with DAT [Prasher et al., 1998]. To gather insight, we analyzed DNAs from individuals with familial AD (FAD), adults with DS, and healthy normal individuals (younger and older) for the C282Y and H63D mutations. As the apolipoprotein E (ApoE) E4 allele on chromosome 19 is a confirmed risk factor for AD [Strittmatter et al., 1993; Roses, 1998; Tang et al., 1998] and possibly also for DAT in DS [van Gool et al., 1995; Lambert et al., 1996; Schupf et al., 1996; Alexander et al., 1997; Evenhuis, 1997; Farrer et al., 1997; Prasher et al., 1997; Sekijima et al., 1998; Tyrrell et al., 1998; Rubinsztein et al., 1999], DNAs were also ApoE genotyped and considered in our data interpretation.
MATERIALS AND METHODS
Human Subjects and Diagnoses
Research protocols were approved for ethical acceptability by the University of Toronto and the New York State Institute for Basic Research in Developmental Disabilities.
Patients with probable FAD (n = 26) were identified from the University of Toronto FAD Cell Line Registry. In this registry names are listed according to the date blood samples were received. To promote generation of a random sample, a list was made of the first 15 affected females and first 15 affected males in the registry, each from a different family without prior consideration of age at onset of FAD or genotype. Frozen aliquots of immortalized B lymphoblast lines established from these individuals were used as a source of DNA. Only three of the 26 patients, one male and two females, had FAD that was unequivocally early onset. (See the Discussion for a description of their genotypes.)
The control group consisted of healthy normal volunteers (n = 41) from several Ontario community organizations. This group was matched in distribution of age and gender to the FAD group and designated as the older normals (mean age ± SD, 63.0 ± 9.4 years).
The adults with DS (n = 50) were recruited from several group homes and institutions in New York State. Their degree of intellectual impairment ranged from mild to profound and was, on average, moderate. Inclusion criteria for entrance into the study were the presence of clinically confirmed DS and medical stability for 3 months before entry into the study. Exclusion criteria were deficits (hearing, visual) that would interfere with assessment, and failure to pass a match-to-sample pretest of short-term memory function. Subject ages ranged from 26–61 years (mean age ± SD, 44.3 ± 10.8 years).
The control group for the adults with DS consisted of healthy normal volunteers (n = 52) largely from agencies providing services for people with developmental disabilities in New York State. This group was matched in distribution of age and gender to the group with DS and designated as the younger normals.
All participants were Caucasian except for one male in the DS group who was Hispanic.
DNA Extraction
DNA samples were prepared from blood samples collected in heparin or acid-citrate dextrose (solution A), from banked transformed B lymphoblasts, or from freshly collected buccal cells, using QIAamp DNA Mini Kits (QIAGEN, Mississauga, ON, Canada).
Genotyping
ApoE genotyping was done by a radioactive PCR procedure involving analysis of an HhaI-restricted PCR-amplified sequence on mini QuickPoint sequencing gels (Novex, San Diego, CA) [Wenham et al., 1991]. Hemochromatosis DNA analysis for the two mutations (C282Y and H63D) in the HFE gene was performed using the primers described by Feder et al. [1996]. PCR products were digested with RsaI (New England Biolabs, Beverly, MA) for detection of the C282Y mutation and with MboI (Gibco BRL, Rockville, MD) for detection of the H63D mutation. Digested PCR fragments were separated by polyacrylamide gel electrophoresis and visualized by ethidium bromide staining.
Statistical Analysis
Observed genotype distributions were compared with those expected by cross-tabulation and standard Chi-squared tests. For selected comparisons, subjects were classified as HFE mutation positive or negative, and ApoE E4 positive or negative. Males and females were examined separately because females in the general population and in DS are reported to be at increased risk for AD [see Payami et al., 1996; Lai et al., 1999]. Analyses were interpreted at the 0.05 level of significance without Bonferroni corrections.
RESULTS
Data Summaries
HFE mutation and ApoE genotyping data for individuals are given in Table I. Frequencies of individuals in the four study groups with HFE mutations or the ApoE E4 variant are given in Figure 1. The effects of sex on the distribution of HFE mutations among ApoE E4 positive and negative individuals are shown in Figure 2.
Younger healthy normal individuals | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
ApoE | Males (26) | Females (26) | ||||||||||
2,2 | 2,3 | 3,3 | 2,4 | 3,4 | 4,4 | 2,2 | 2,3 | 3,3 | 2,4 | 3,4 | 4,4 | |
HFE | ||||||||||||
–,– | 2 | 3 | 11 | 3 | 1 | 6 | 1 | 3 | ||||
–,C282Y | 3 | 1 | 1 | |||||||||
–,H63D | 2 | 1 | 1 | 2 | 5 | 1 | 3 | |||||
C282Y,C282Y | ||||||||||||
C282Y,H63D | ||||||||||||
H63D,H63D | 2 |
Older healthy normal individuals | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
ApoE | Males (19) | Females (22) | ||||||||||
2,2 | 2,3 | 3,3 | 2,4 | 3,4 | 4,4 | 2,2 | 2,3 | 3,3 | 2,4 | 3,4 | 4,4 | |
HFE | ||||||||||||
–,– | 9 | 5 | 2 | 7 | 4 | 1 | ||||||
–,C282Y | 1 | 1 | ||||||||||
–,H63D | 2 | 2 | 1 | 1 | 2 | |||||||
C282Y,C282Y | ||||||||||||
C282Y,H63D | 2 | |||||||||||
H63D,H63D | 1 |
Down syndrome | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
ApoE | Males (24) | Females (26) | ||||||||||
2,2 | 2,3 | 3,3 | 2,4 | 3,4 | 4,4 | 2,2 | 2,3 | 3,3 | 2,4 | 3,4 | 4,4 | |
HFE | ||||||||||||
–,– | 1 | 1 | 5 | 6 | 1 | 13 | 1 | 3 | ||||
–,C282Y | 1 | 3 | ||||||||||
–,H63D | 1 | 4 | 2 | 2 | 1 | 2 | ||||||
C282Y,C282Y | 1 | |||||||||||
C282Y,H63D | 1 | |||||||||||
H63D,H63D | 1 |
Familial Alzheimer disease | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
ApoE | Males (12) | Females (14) | ||||||||||
2,2 | 2,3 | 3,3 | 2,4 | 3,4 | 4,4 | 2,2 | 2,3 | 3,3 | 2,4 | 3,4 | 4,4 | |
HFE | ||||||||||||
–,– | 2 | 3 | 1 | 2 | 5 | 3 | ||||||
–,C282Y | 1 | 1 | 1 | |||||||||
–,H63D | 1 | 1 | 2 | 2 | ||||||||
C282Y,C282Y | ||||||||||||
C282Y,H63D | 1 | |||||||||||
H63D,H63D |
- * The integers in the table indicate the number of males or females in each category with the given HFE and ApoE genotypes. There are three common ApoE alleles—E2, E3, and E4. the six corresponding ApoE genotypes are shown in the ApoE row. The endashes in the HFE column denote the wild-type HFE gene.

Frequencies of individuals with HFE mutations or the ApoE E4 allele in the four subject groups. F, frequency of individuals in each category; N (hatched bars), younger healthy normal individuals; N (solid bars), older healthy normal individuals; DS, adults with Down syndrome; AD, patients with familial Alzheimer disease. Refer to text for details.

Co-distribution of HFE mutations and the ApoE E4 variant in the four subject groups. Percentages of individuals in the different categories are shown. Percentages were calculated separately for males and females. N (hatched bars), younger healthy normal individuals; N (solid bars), older healthy normal individuals; DS, adults with Down syndrome; AD, patients with familial Alzheimer disease; E4+ and E4-, with or without one or two ApoE E4 alleles; *H+ and *H-, with or without one or two C282Y or H63D (HFE) mutations. Refer to text for details. Percentages that are significantly different from expected at the 95% confidence level or greater are indicated with asterisks. *,1: The percentage of *H-, E4- males with AD is significantly lower than expected in a comparison with older *H-, E4- males (P = 0.0001); *,1: The percentage of *H-,E4+ females is significantly higher than expected in a comparison with older normal *H, E4+ females (P = 0.036); *,2. There is a trend for the percentage of *H+, E4- females with AD to be lower than expected in a comparison with older normal *H+, E4- females (P = 0.055). When H63D and C282Y mutations were considered separately, results were analogous to those obtained for HFE mutations in combination. P values corresponding to *,1 and *,2 for H63D alone were 0.0066 and 0.022, respectively. P values corresponding to *,1 and *,2 for C282Y alone were 0.0050 and 0.047, respectively.
Frequencies of C282Y and H63D in the Four Subject Groups
The frequencies of healthy normal individuals with one or two C282Y mutations ranged from 0.053 (older males) to 0.12 (younger males), and with one or two H63D mutations from 0.15 (younger males) to 0.50 (younger females) (Table I, Fig. 1). These values are similar to those reported for Caucasians [Merryweather-Clarke et al., 1997; Marshall et al., 1999]. Three H63D homozygotes and two compound C282Y/H63D heterozygotes were found among the healthy normal individuals; one compound C282Y/H63D heterozygote and one H63D homozygote were found within the group with DS; one compound C282Y/H63D heterozygote was found within the group with FAD (Table I). HFE mutations in combination were overrepresented in males and underrepresented in females with FAD relative to the older normal individuals. Similar but less pronounced trends were apparent in the group with DS relative to the matched younger normals. These HFE effects were due in large part to H63D mutations. However, C282Y contributed substantially to the total HFE frequency in males with FAD and in females with DS (Table I, Fig. 1).
C282Y Mutation Is Negatively Associated With ApoE E4 in Healthy Normal Individuals and in DS
Inspection of the data on the two groups of healthy normal individuals in Table I indicated that the C282Y mutation was rarely associated with ApoE E4 allele. Seven of eight C282Y mutations occurred in individuals lacking an ApoE E4 allele. In contrast, H63D mutations occurred in both ApoE E4 positive and negative individuals. The observed distribution of HFE mutations among ApoE E4 positive and negative individuals differed significantly from that expected (P = 0.027). In the group with DS, the C282Y mutation was present exclusively in ApoE E4 negative individuals. In the FAD group, three of four at C282Y mutations were found in association with ApoE E4 positive individuals.
Gender Effects on the Distribution of HFE Mutations in ApoE E4 Positive and Negative Patients With FAD
The frequency of ApoE E4 was greatly increased in both male and female patients with FAD relative to the older and younger normal individuals (Table I, Fig. 1). When subjects were classified as ApoE E4 positive or negative, there was a significant difference between the percentages of patients with FAD and of older normals in each category (P = 0.0016 in both comparisons). The percentage of patients with FAD was greater than that of the normals in the ApoE E4 positive category, consistent with previous reports that ApoE E4 predisposed to FAD. The percentage of patients with FAD was less than that for the normals in the ApoE E4 negative category, consistent with the absence of ApoE E4 being protective against FAD (data summaries not shown). In comparisons between ApoE E4 negative FAD patients and individuals in the two normal groups, there were highly significant sex effects on the distribution of HFE mutations (P < 0.002 in each case). Significant sex effects were not present in analyses of ApoE E4 positive individuals. The basis for these sex effects is evident from Figure 2. When subjects were stratified according to gender and the presence or absence of HFE mutations or the ApoE E4 allele, only two comparisons were statistically significant (denoted by *,1 and *,2 in Fig. 2). Among males, the percentage with FAD who were HFE mutation negative and ApoE E4 negative (denoted as *H-, E4-, Fig. 2) was significantly less than the percentage of older normal individuals in this category (P = 0.0001, denoted by *,1). Among females, the percentage with FAD who were HFE mutation negative and ApoE E4 positive (denoted as *H-, E4+, Fig. 2) was significantly greater than the percentage of older normals in this category (P = 0.036, denoted by *,2). Further, among females there was a trend for the percentage of patients with FAD who were HFE mutation positive and ApoE E4 negative (denoted as *H+, E4-, Fig. 2) to be lower than the percentage of normals in this category (P = 0.055). This effect was not as apparent in comparisons of females with FAD who were HFE mutation and ApoE E4 negative (*H-, E4-, Fig. 2) and older normal females in this category. The results of analyses that considered the H63D and C282Y mutations alone were similar (Fig. 2, legend).
Collectively, these observations support the proposition that there is a complex association between HFE mutations and the ApoE E4 allele in FAD that is different in males and females. The associations described raise the possibility that among ApoE E4 negative males the absence of HFE mutations may be protective against FAD and, conversely, that the presence of HFE mutations predisposes to FAD. Among ApoE E4 negative females, the presence of HFE mutations may afford some protection against FAD. Furthermore, the ApoE E4 allele may be most predisposing to FAD among HFE mutation negative females.
HFE Mutations in ApoE E4 Positive and Negative Individuals With Down Syndrome
There were trends for a deficiency of ApoE E4 positive females with DS and an excess of ApoE E4 positive males in comparison to the healthy younger individuals (Table I, Fig. 1). Differences in the distribution of HFE mutations in ApoE E4 positive or negative males with DS relative to younger normal males resembled those in the FAD–older normal male comparisons, but they were less pronounced (Fig. 2). Differences in the distribution of HFE mutations in ApoE E4 positive or negative females with DS relative to younger normal females showed no consistent trend (Fig. 2).
DISCUSSION
This study appears to be the first to determine HFE mutation frequencies relative to ApoE allele variant frequencies in well-characterized Caucasian individuals with FAD or DS relative to healthy normal Caucasian volunteers matched with respect to distribution of age, gender, and geographical location. Although our findings provide evidence for a nonrandom association of certain mutations/polymorphisms in the HFE gene with FAD and possibly DS, it is not clear if they are causal, if they are risk factors like the ApoE E4 allele, or if there is some linkage disequilibrium process involved, or if the results are consequences of other known anomalies that sometimes arise in genetic association studies [Crawford et al., 2000]. Some studies have suggested that a gene in the HLA locus (HLA-A2), or closely linked to it, may lower the age of onset of AD [Small and Matsuyama, 1986; Payami et al., 1991, 1997; Small et al., 1991, 1999; Combarros et al., 1998; Ballerini et al., 1999]. We question if there is any relationship between our observations and the latter findings, since Summers et al. [1989] found linkage disequilibrium between HLA-A2 and B12 in a population with HHC.
The possibility that the observed association of HFE mutations with AD is the result of survival effects also should be considered. In the general population, HHC is reported to be associated with substantial mortality resulting from liver disease, liver neoplasms, cardiomyopathy, and a combination of liver disease and diabetes [Yang et al., 1998]. Heterozygosity for hemochromatosis is reported to be associated with increased risk for cardiovascular problems [Roest et al., 1999; Tuomainen et al., 1999], colorectal neoplasia, diabetes, hematologic malignancy, and gastric cancer [Nelson et al., 1995]. Thus, the different HFE mutation effects for E4 negative males and females (Fig. 2) may, in part, reflect gender differences in iron metabolism and regulation, and possibly also gender differences in HFE mutation-associated complications and mortality. Future studies might consider both demographic data and data on genetic markers in the estimation of hazard rates, relative risks, and survival functions for different genes or genotypes [Yashin et al., 1999].
Our finding that males and females with FAD both have an excess of ApoE E4 has been reported in many other studies [see Roses, 1998]. Our finding that ApoE E4 effects in FAD appear strongest in females lacking HFE mutations (Fig. 2) supports and extends previous findings [Payami et al., 1996; Bretsky et al., 1999]. Our finding of a relative excess of ApoE E4 positive males with DS is in agreement with a report of Schupf et al. [1996]; in our group with DS, there also was a relative deficiency of ApoE E4 positive females. Although the latter and other differences between males and females with DS relative to the matched controls (Figs. 1, 2) may be trisomy 21 effects, we cannot exclude the possibility that they reflect bias of selection due to the exclusion and inclusion criteria for research participation, which possibly might exclude very low functioning individuals with DS.
Aside from the question of whether HFE mutations are risk factors for FAD or DAT in DS, clinicians should be aware that HHC can occur in individuals with FAD and DS as well as in the general population, and that iron overload can be effectively managed by blood-letting [Jazwinski, 1998; Bothwell and MacPhail, 1998]. Moreover, the overrepresentation of HFE mutations in males with FAD (Figs. 1, 2) possibly might contribute to a cluster of comorbid conditions that are associated with mortality in AD in males—cardiac arrhythmia, chronic obstructive pulmonary disease, Parkinson's disease, and cancer [Gambassi et al., 1998]. Similarly, HFE mutations might contribute to certain complications of DS, for example, severe neonatal liver damage [Ruchelli et al., 1991; Cheung et al., 1995]. Conversely, underrepresentation of HFE mutations in females with AD and DAT (Figs. 1, 2) might be associated with fewer comorbid conditions and predispose to a longer life after diagnosis of AD or DAT and possibly contribute to the increased risk for AD and DAT that has been observed in females [Payami et al., 1996; Lai et al., 1999].
Although the present data are limited, studying associations between different genetic markers can expand our knowledge of gene function. For example, the finding of a negative association between C282Y and ApoE E4 in healthy normal individuals implicates ApoE in iron metabolism. The absence of this negative association between C282Y and ApoE E4 in the patients with FAD, suggests that there may be some selective advantages to the ApoE E4 allele and HFE mutations. In this regard, it is of interest to describe the genotypes of the three patients with early onset FAD included in the study. Curiously, the male patient with FAD caused by an APP mutation was heterozygous for C282Y and negative for ApoE E4. The female patient with the PS-1 mutation was HFE mutation and ApoE E4 negative. The third female patient with early onset AD was HFE mutation negative and heterozygous for ApoE E4.
It is tempting to speculate why there may be pros and cons to having HFE mutations and/or an ApoE E 4 allele. Because of their deficiency of cysteine residues relative to the wild-type proteins, C282Y and ApoE E4 proteins have a lowered redox potential. Because they would not buffer as well against oxidative stress than the wild-type proteins, they would be expected to lead to an increased level of reactive oxygen species (ROS). Infection and chronic inflammation both are associated with activation of the respiratory burst apparatus in phagocytes and the production of ROS [Edwards, 1996]. An increased level of ROS might be helpful in infection but result in increased bystander damage in chronic inflammation. Increased iron uptake associated with C282Y and H63D may also have advantages or disadvantages. It has been suggested that the C282Y mutation might help to maintain body iron in young (menstruating) women [Datz et al., 1998]. However, excessive tissue iron deposition ultimately could be harmful, because of the potential of Fe++ to react with H2O2 to form •OH- radicals (the Fenton reaction) and of Fe++ to react with O2 to form ROS [Qian and Buettner, 1999]. HFE mutations are also associated with increased blood levels of lead [Barton et al., 1994] and would be expected to promote increased uptake of aluminum since transferrin is the major serum binding protein for aluminum [Golub et al., 1999]. There is evidence that aluminum amplifies iron-catalyzed lipid peroxidation [Yoshino et al., 1999]. Many studies have found evidence of oxidative damage and increased amounts of metals such as iron and aluminum in the brains of patients suffering from AD [Perl and Good, 1992; DeVoto and Yokel, 1994; Smith and Perry, 1995; Bouras et al., 1997; Smith et al., 1997; Yang et al., 1999; Christen, 2000]. It is possible that HHC mutation homozygosity or heterozygosity may contribute to excessive iron deposition in the brain of patients with AD and in DS-related DAT. All tissues that have excessive iron deposition might be expected to share certain pathogenic features as the result of ROS damage.
Our clinical studies in DS are providing some insight as to how the ApoE E4 allele and HFE mutations might contribute to the clinical diagnosis of dementia of the AD type in DS. We previously reported that for persons with DS having an ApoE E4 allele is associated with a trend for lower neurocognitive and neurobehavioral Dementia Test Battery (DTB) scores than not having an ApoE E4 allele [Percy et al., 1998a]. Moreover, DTB scores in adults with DS decline markedly with subject age over the age of 45 years [Percy et al., 1998a]. In subjects under the age of 45 years, having an HFE mutation is also associated with somewhat lower DTB scores than not having an HFE mutation. In addition, within ApoE E4 and HFE mutation positive and negative categories, there are trends for gender effects (manuscript in preparation). DTB scores for our subjects with DS thus depend on genetic makeup, gender, and especially subject age.
Finally, it may be relevant that the intramuscular injection of the trivalent metal chelator desferrioxamine, which binds to iron, aluminum, chromium, and to a lesser extent copper and certain other transition metals, has been found to retard AD development and also to reduce mortality in AD [Crapper-McLachlan et al., 1991]. From a therapeutic development standpoint, research into preventative strategies to reduce oxidative damage in AD, DS-related DAT, and HHC is highly warranted. Also, the possibility that HFE mutations may be important new genetic risk factors for AD should be followed up.
Acknowledgements
The authors thank the staff, friends, families, and participants associated with Pathfinder Village, Inc., Edmeston, New York, the Columbia Association for Retarded Citizens, particularly in Mellenville and Hudson, New York, the Independent Group Home Living, Inc., East Moriches, Long Island, New York, the Albany chapter of ARC, Albany, New York, and the Eleanor Roosevelt and Q.D. Heck Developmental Disabilities Services Offices, Schenectady, New York, for support and encouragement. The support of Surrey Place Centre, the New York State Office on Mental Retardation and Developmental Disabilities and the New York State Institute for Basic Research is gratefully appreciated. The Familial Alzheimer Registry Cell Line Bank is maintained by MEP with support, in part, from the University of Toronto Centre for Research in Neurodegenerative Diseases (Director: Dr. Peter St. George-Hyslop). Analyses for HFE mutations were done by Kimball Genetics, Denver, CO. We thank Sharon Bauer, Winnie Jeng, Helen Tran, and Linda Wu for expert technical assistance, Shoshana Moalem for helping with the recruitment of the volunteers, Dr. Annette Taylor for helpful discussions, and the reviewers of this article for helpful comments. Doug Biggerstaff, Biologic Biomedical Communications, provided the graphics.