Alzheimer Disease

A number sign (#) is used with this entry because of evidence that familial Alzheimer disease-1 (AD1) is caused by mutation in the gene encoding the amyloid precursor protein (APP; 104760) on chromosome 21q.

A homozygous mutation in the APP gene with a dominant-negative effect on amyloidogenesis was found in a patient with an early-onset progressive dementia and his affected younger sister (104760.0022).

A coding single-nucleotide polymorphism (SNP) in the APP gene (104760.0023) has been shown to have a protective effect against Alzheimer disease.

See also APP-related cerebral amyloid angiopathy (CAA; 605714), which shows overlapping clinical and neuropathologic features.

Description

Alzheimer disease is the most common form of progressive dementia in the elderly. It is a neurodegenerative disorder characterized by the neuropathologic findings of intracellular neurofibrillary tangles (NFT) and extracellular amyloid plaques that accumulate in vulnerable brain regions (Sennvik et al., 2000). Terry and Davies (1980) pointed out that the 'presenile' form, with onset before age 65, is identical to the most common form of late-onset or 'senile' dementia, and suggested the term 'senile dementia of the Alzheimer type' (SDAT).

Haines (1991) reviewed the genetics of AD. Selkoe (1996) reviewed the pathophysiology, chromosomal loci, and pathogenetic mechanisms of Alzheimer disease. Theuns and Van Broeckhoven (2000) reviewed the transcriptional regulation of the genes involved in Alzheimer disease.

Genetic Heterogeneity of Alzheimer Disease

Alzheimer disease is a genetically heterogeneous disorder. See also AD2 (104310), associated with the APOE*4 allele (107741) on chromosome 19; AD3 (607822), caused by mutation in the presenilin-1 gene (PSEN1; 104311) on 14q; and AD4 (606889), caused by mutation in the PSEN2 gene (600759) on 1q31.

There is evidence for additional AD loci on other chromosomes; see AD5 (602096) on 12p11, AD6 (605526) on 10q24, AD7 (606187) on 10p13, AD8 (607116) on 20p, AD9 (608907), associated with variation in the ABCA7 gene (605414) on 19p13, AD10 (609636) on 7q36, AD11 (609790) on 9q22, AD12 (611073) on 8p12-q22, AD13 (611152) on 1q21, AD14 (611154) on 1q25, AD15 (611155) on 3q22-q24, AD16 (300756) on Xq21.3, AD17 (615080) on 6p21.2, and AD18 (615590), associated with variation in the ADAM10 gene (602192) on 15q21.

Evidence also suggests that mitochondrial DNA polymorphisms may be risk factors in Alzheimer disease (502500).

Finally, there have been associations between AD and various polymorphisms in other genes, including alpha-2-macroglobulin (A2M; 103950.0005), low density lipoprotein-related protein-1 (LRP1; 107770), the transferrin gene (TF; 190000), the hemochromatosis gene (HFE; 613609), the NOS3 gene (163729), the vascular endothelial growth factor gene (VEGF; 192240), the ABCA2 gene (600047), and the TNF gene (191160) (see MOLECULAR GENETICS).

Clinical Features

Alzheimer (1907) provided the first report of the disease (see HISTORY).

Schottky (1932) described a familial form of presenile dementia in 4 generations. The diagnosis was confirmed at autopsy in a patient in the fourth generation. Lowenberg and Waggoner (1934) reported a family with unusually early onset of dementia in the father and 4 of 5 children. Postmortem findings in 1 case were consistent with dementia of the Alzheimer type. McMenemey et al. (1939) described 4 affected males in 2 generations with pathologic confirmation in one.

Heston et al. (1966) described a family with 19 affected in 4 generations. Dementia was coupled with conspicuous parkinsonism and long tract signs.

Rice et al. (1980) and Ball (1980) reported a kindred in which members had clinical features of familial AD. Two patients had neuropathologic changes of spongiform encephalopathy of the Creutzfeldt-Jakob type (CJD; 123400) at autopsy, but the long clinical course was unusual for CJD. Corkin et al. (1983) found no difference in parental age of patients with AD compared to controls. Nee et al. (1983) reported an extensively affected kindred, with 51 affected persons in 8 generations. There was no increased incidence of Down syndrome (190685) or hematologic malignancy.

Heyman et al. (1983) found dementia in first-degree relatives of 17 (25%) of 68 probands with AD. These families also demonstrated an increase in the frequency of Down syndrome (3.6 per 1,000 as compared with an expected rate of 1.3 per 1,000). No excess of hematologic malignancy was found in relatives. In a study of the families of 188 Down syndrome children and 185 controls, Berr et al. (1989) found no evidence of an excess of dementia cases suggestive of AD in the families of patients with Down syndrome. In a large multicenter study of first-degree relatives of 118 AD probands and nondemented spouse controls, Silverman et al. (1994) found no association between familial AD and Down syndrome.

Stokin et al. (2005) identified axonal defects in mouse models of Alzheimer disease that preceded known disease-related pathology by more than a year; the authors observed similar axonal defects in the early stages of Alzheimer disease in humans. Axonal defects consisted of swellings that accumulated abnormal amounts of microtubule-associated and molecular motor proteins, organelles, and vesicles. Impairing axonal transport by reducing the dosage of a kinesin molecular motor protein enhanced the frequency of axonal defects and increased amyloid-beta peptide levels and amyloid deposition. Stokin et al. (2005) suggested that reductions in microtubule-dependent transport may stimulate proteolytic processing of beta-amyloid precursor protein (104760), resulting in the development of senile plaques and Alzheimer disease.

Bateman et al. (2012) performed a prospective, longitudinal study analyzing data from 128 subjects at risk for carrying a mutation for autosomal dominant AD. Subjects underwent baseline clinical and cognitive assessments, brain imaging, and cerebrospinal fluid and blood tests. Bateman et al. (2012) used the participant's age at baseline assessment and the parent's age at the onset of symptoms of AD to calculate the estimated years from expected symptom onset (age of the participant minus parent's age at symptom onset). They then conducted cross-sectional analyses of baseline data in relation to estimated years from expected symptom onset in order to determine the relative order and magnitude of pathophysiologic changes. Concentrations of amyloid-beta-42 in the CSF appeared to decline 25 years before expected symptom onset. Amyloid-beta deposition, as measured by positron-emission tomography with the use of Pittsburgh compound B, was detected 15 years before expected symptom onset. Increased concentrations of tau protein in the CSF and an increase in brain atrophy were detected 15 years before expected symptom onset. Cerebral hypometabolism and impaired episodic memory were observed 10 years before expected symptom onset. Global cognitive impairment, as measured by Mini-Mental State Examination and the Clinical Dementia Rating scale, was detected 5 years before expected symptom onset, and patients met diagnostic criteria for dementia at an average of 3 years after expected symptom onset. Bateman et al. (2012) cautioned that their results required confirmation with use of longitudinal data and may not apply to patients with sporadic Alzheimer disease.

Familial Alzheimer Disease 1

Karlinsky et al. (1992) reported a family from Toronto with autosomal dominant inheritance of Alzheimer disease. The disorder was characterized by early onset of memory deficits, decreased speed of cognitive processing, and impaired attention to complex cognitive sets. The family immigrated to Canada from the British Isles in the 18th century. Genetic analysis identified a mutation in the APP gene (V717I; 104760.0002).

Farlow et al. (1994) reviewed the clinical characteristics of the disorder in the AD family reported by Murrell et al. (1991) in which affected members had a mutation in the APP gene (V717F; 104760.0003). The mean age of onset of dementia was 43 years. The earliest cognitive functions affected were recent memory, information-processing speed, sequential tracking, and conceptual reasoning. Language and visuoperceptual skills were largely spared early in the course of the disease. Later, there were progressive cognitive deficits and inability to perform the activities of daily living. Death occurred, on average, 6 years after onset. The family was Romanian, many of its members having migrated to Indiana.

Rossi et al. (2004) reported a family in which at least 6 members spanning 3 generations had Alzheimer disease and strokes associated with a heterozygous mutation in the APP gene (A713T; 104760.0009). At age 52 years, the proband developed progressive cognitive decline with memory loss and visuospatial troubles, as well as stroke-like episodes characterized by monoparesis and language disturbances detectable for a few days. MRI showed T2-weighted signal hyperintensities in subcortical and periventricular white matter without bleeding. Neuropathologic examination showed neurofibrillary tangles and A-beta-40- and A-beta-42-immunoreactive deposits in the neuropil. The vessel walls showed only A-beta-40 deposits, consistent with amyloid angiopathy. There were also multiple white matter infarcts along the long penetrating arteries. Other affected family members had a similar clinical picture. Several unaffected family members carried the mutation, and all but 1 were under 65 years of age.

Edwards-Lee et al. (2005) reported an African American family in which multiple members spanning 3 generations had early-onset AD. The distinctive clinical features in this family were a rapidly progressive dementia starting in the fourth decade, seizures, myoclonus, parkinsonism, and spasticity. Variable features included aggressiveness, visual disturbances, and pathologic laughter. Two sibs who were tested were heterozygous for a mutation in the APP gene (T714I; 104760.0015).

Early-Onset Alzheimer Disease with Cerebral Amyloid Angiopathy

Because Alzheimer disease associated with cerebral amyloid angiopathy (CAA) is also found in Down syndrome, Rovelet-Lecrux et al. (2006) reasoned that the APP locus located on chromosome 21q21 might be affected by gene dosage alterations in a subset of demented individuals. To test this hypothesis, they analyzed APP using quantitative multiplex PCR of short fluorescent fragments, a sensitive method for detecting duplications that is based on the simultaneous amplification of multiple short genomic sequences using dye-labeled primers under quantitative conditions. This analysis was performed in 12 unrelated individuals with autosomal dominant early-onset Alzheimer disease (ADEOAD) in whom a previous mutation screen of PSEN1 (104311), PSEN2 (600759), and APP had been negative; 5 of these individuals belonged to Alzheimer disease-affected families in which the cooccurrence of CAA had been diagnosed according to neuropathologic (Vonsattel et al., 1991) or clinical criteria (intracerebral hemorrhages (ICH) in at least 1 affected individual). In the 5 index cases with the combination of early-onset Alzheimer disease and CAA, they found evidence for a duplication of the APP locus (104760.0020). In the corresponding families, the APP locus duplication was present in affected subjects but not in healthy subjects over the age of 60 years. The phenotypes of the affected subjects in the 5 families were similar. None had mental retardation before the onset of dementia. None had clinical features suggestive of Down syndrome. The most common clinical manifestation was progressive dementia of Alzheimer disease type (mean age of onset 52 +/- 4.4 years) associated, in some cases, with lobar ICH. Neuropathologic examination of the brains of 5 individuals from 3 kindreds showed abundant amyloid deposits, present both as dense-cored plaques and as diffuse deposits, in all regions analyzed. Neurofibrillary tangles were noted in a distribution consistent with the diagnosis of definite Alzheimer disease. However, the most prominent feature was severe CAA. Rovelet-Lecrux et al. (2006) estimated that in their whole cohort of 65 ADEOAD families, the frequency of the APP locus duplication was roughly 8% (5 of 65), which corresponds to half of the contribution of APP missense mutations to ADEOAD.

Other Features

In longitudinal studies using magnetic resonance spectroscopic imaging (MRSI), Adalsteinsson et al. (2000) found that 12 patients with AD had a striking decline in the neuronal marker N-acetyl aspartate, compared to 14 controls. However, there was little decline in underlying gray matter volume in these patients.

In a comparison of 59 unrelated patients with AD and over 1,000 controls, Borenstein Graves et al. (2001) found that a combination of low head circumference and presence of the APOE4 allele strongly predicted earlier onset of AD. The authors suggested that the clinical expression of AD may occur when degeneration in specific brain regions falls below a critical threshold of 'brain reserve,' beyond which normal cognitive function cannot be maintained.

In a study of 461 sibs of 371 probands diagnosed with AD, Sweet et al. (2002) found that AD plus psychosis in probands was associated with a significantly increased risk for AD plus psychosis in family members (odds ratio = 2.4), demonstrating familial aggregation of this phenotype.

In a PET study comparing brain glucose metabolism between 46 patients with sporadic AD and 40 patients with familial AD, Mosconi et al. (2003) found that both groups had reductions in the metabolic rate of glucose in similar regional areas of the brain, particularly the posterior cingulate cortex, the parahippocampal gyrus, and occipital areas, suggesting common neurophysiologic pathways of degeneration. However, patients with familial AD had a more severe reduction in glucose metabolism in all these areas, suggesting that genetic predisposition further strains the degenerative process.

Biochemical Features

Zubenko et al. (1987) described a biophysical alteration of platelet membranes in Alzheimer disease. They concluded that increased platelet membrane fluidity (see 173560) characterized a subgroup of patients with early age of symptomatic onset and rapidly progressive course. Zubenko and Ferrell (1988) described monozygotic twins concordant for probable AD and for increased platelet membrane fluid.

Abraham et al. (1988) identified one of the components of the amyloid deposits seen in AD as the serine protease inhibitor alpha-1-antichymotrypsin (AACT; 107280). Birchall and Chappell (1988) suggested that individual vulnerability of genetic factors influencing intake, transport or excretion of aluminum may be a mechanism for familial AD.

Yan et al. (1996) reported that the RAGE protein (AGER; 600214) is an important receptor for the amyloid beta peptide and that expression of this receptor is increased in AD. They noted that expression of RAGE was particularly increased in neurons close to deposits of amyloid beta peptide and to neurofibrillary tangles.

Cholinergic projection neurons of the basal forebrain nucleus basalis express nerve growth factor (NGF) receptors p75(NTR) (162010) and TrkA (191315), which promote cell survival. These same cells undergo extensive degeneration in AD. Counts et al. (2004) found an approximately 50% average reduction in TrkA levels in 4 cortical brain regions of 15 patients with AD, compared to 18 individuals with no cognitive impairment and 16 with mild/moderate cognitive impairment. By contrast, cortical p75(NTR) levels were stable across the diagnostic groups. Scores on the Mini-Mental State Examination (MMSE) correlated with TrkA levels in the anterior cingulate, superior frontal, and superior temporal cortices. Counts et al. (2004) suggested that reduced TrkA levels may be the cause or result of abnormal cholinergic function in AD.

The Framingham (Massachusetts) Study cohort has been evaluated biennially since 1948. In a sample of 1,092 subjects (mean age, 76 years) from this cohort, Seshadri et al. (2002) analyzed the relation of the plasma total homocysteine level measured at baseline and that measured 8 years earlier to the risk of newly diagnosed dementia on follow-up. They used multivariable proportional-hazards regression to adjust for age, sex, apoE genotype, vascular risk factors other than homocysteine, and plasma levels of folate and vitamins B12 and B6. Over a median follow-up period of 8 years, dementia developed in 111 subjects, including 83 given a diagnosis of Alzheimer disease. The multivariable-adjusted relative risk of dementia was 1.4 for each increase of 1 standard deviation in the log-transformed homocysteine value either at baseline or 8 years earlier. The relative risk of Alzheimer disease was 1.8 per increase of 1 SD at baseline and 1.6 per increase of 1 SD 8 years before baseline. With a plasma homocysteine level greater than 14 micromol per liter, the risk of Alzheimer disease nearly doubled. Seshadri et al. (2002) concluded that an increased plasma homocysteine level is a strong, independent risk factor for the development of dementia and Alzheimer disease.

Among 563 AD patients and 118 controls, Prince et al. (2004) found that presence of the APOE4 allele was strongly associated with reduced CSF levels of beta-amyloid-42 in both patients and controls. In a retrospective study of 443 AD patients, Evans et al. (2004) found that increased serum total cholesterol was associated with more rapid disease progression in patients who did not have the APOE4 allele. The effect was not seen in patients with the APOE4 allele and high cholesterol.

Botella-Lopez et al. (2006) found increased levels of a 180-kD reelin (RELN; 600514) fragment in CSF from 19 patients with AD compared to 11 nondemented controls. Western blot and PCR analysis confirmed increased levels of reelin protein and mRNA in tissue samples from the frontal cortex of AD patients. Reelin was not increased in plasma samples, suggesting distinct cellular origins. The reelin 180-kD fragment was also increased in CSF samples of other neurodegenerative disorders, including frontotemporal dementia (600274), progressive supranuclear palsy (PSP; 601104), and Parkinson disease (PD; 168600).

Tesseur et al. (2006) found significantly decreased levels of TGF-beta receptor type II (TGFBR2; 190182) in human AD brain compared to controls; the decrease was correlated with pathologic hallmarks of the disease. Similar decreases were not seen in brain extracts from patients with other forms of dementia. In a mouse model of AD, reduced neuronal TGFBR2 signaling resulted in accelerated age-dependent neurodegeneration and promoted beta-amyloid accumulation and dendritic loss. Reduced TGFBR2 signaling in neuroblastoma cell cultures resulted in increased levels of secreted beta-amyloid and soluble APP. The findings suggested a role for TGF-beta (TGFB1; 190180) signaling in the pathogenesis of AD.

Counts et al. (2007) found a 60% increase in CHRNA7 (118511) mRNA levels in cholinergic neurons of the nucleus basalis in patients with mild to moderate Alzheimer disease compared to those with mild cognitive impairment or normal controls. Expression levels of CHRNA7 were inversely associated with cognitive test scores. Counts et al. (2007) suggested that upregulation of CHRNA7 receptors may be a compensatory response to maintain basocortical cholinergic activity during disease progression or may act with beta-amyloid in disease pathogenesis.

Pathogenesis

In a study of the families of Alzheimer disease patients, Heston (1977) found an excess of Down syndrome and of myeloproliferative disorders, including lymphoma and leukemia. Neurons of Alzheimer patients show a neurofibrillary tangle that is made up of disordered microtubules. An identical lesion occurs in the neurons of Down syndrome, at an earlier age than in Alzheimer disease. Leukemia and accelerated aging are also features of Down syndrome. Heston (1977) and Heston and Mastri (1977) speculated a disorder of microtubules as a common pathomechanism. Heston and White (1978) further speculated defective organization of microfilaments and microtubules in AD. Using immunoprecipitation techniques, Grundke-Iqbal et al. (1979) showed that neurofibrillary tangles in AD probably originate from neurotubules. Harper et al. (1979) could not confirm a systemic microtubular defect in Alzheimer disease; cultured skin fibroblasts from AD patients showed normal tubulin networks. Nordenson et al. (1980) found an increased frequency of acentric fragments in karyotypes from AD patients, and suggested that this was consistent with defective tubulin protein leading to erratic function of the spindle mechanism.

Gajdusek (1986) suggested that the amyloid in Alzheimer disease and Down syndrome is formed from a precursor synthesized in neurons as well as in microglial cells and brain macrophages. He further suggested that the precursor synthesized in neurons produces intracellular neurofibrillary tangles, and that the precursor synthesized in microglial cells and brain macrophages is exuded from the cell, forming the extracellular amyloid plaques and vascular amyloid deposits. Dying neurons may also contribute to extracellular deposits.

Bergeron et al. (1987) found that cerebral amyloid angiopathy (605714) was present in 86% of AD patients and 40% of age-matched controls. The findings suggested that cerebral amyloid angiopathy is an integral component of AD.

Using immunocytochemistry, Wolozin et al. (1988) identified a 68-kD protein in cerebral cortical neurons from both normal human fetal and neonatal brain and brain tissue from neonates with Down syndrome. The number of reactive neurons decreased sharply after age 2 years, but reappeared in older individuals with Down syndrome and in patients with Alzheimer disease.

Carrell (1988) speculated that plaque formation in AD was a consequence of proteolysis of a precursor protein; self-aggregation of the cleaved A4 peptides explained the precipitated amyloid, while release of a trophic inhibitory domain explained the interwoven neuritic development. Using computer-enhanced imaging of immunocytochemical stains of Alzheimer disease prefrontal cortex, Majocha et al. (1988) described the distribution of amyloid protein deposits exclusive of other senile plaque components. Joachim et al. (1989) presented evidence suggesting that Alzheimer disease is not restricted to the brain but is a widespread systemic disorder with accumulation of amyloid beta protein (104760) in nonneuronal tissues.

Ellis et al. (1996) found that 83% of 117 patients with autopsy-confirmed AD had at least a mild degree of cerebral amyloid angiopathy. Thirty (25.6%) of 117 brains showed moderate to severe CAA affecting the cerebral vessels in one or more cortical regions. These brains also showed a significantly higher frequency of hemorrhages or ischemic lesions compared to those with little or no amyloid angiopathy (43.3% versus 23.0%; odds ratio = 2.6). High CAA scores also correlated with the presence of cerebral arteriosclerosis and with older age at onset of dementia.

In light of the findings of Tomita et al. (1997) concerning PSEN2 mutation and altered metabolism of APP (summarized in 600759.0001), Hardy (1997) reviewed the evidence that Alzheimer disease has many etiologies, but one pathogenesis. Mutations in all known pathogenic genes have in common the fact that they alter processing of APP, thus lending strong support to the amyloid cascade hypothesis. Heintz and Zoghbi (1997) suggested that alpha-synuclein (163890) may provide a link between Parkinson disease (see 168600) and Alzheimer disease and possibly other neurodegenerative diseases.

The neurofibrillary tangle, one of the neuropathologic hallmarks of AD, contains paired helical filaments (PHFs) composed of the microtubule-associated protein tau (MAPT; 157140). Tau is hyperphosphorylated in PHFs, and phosphorylation of tau abolishes its ability to bind microtubules and promote microtubule assembly. Lu et al. (1999) demonstrated that PIN1 (601052) binds hyperphosphorylated tau and copurifies with PHFs, resulting in depletion of soluble PIN1 in the brains of patients with AD. PIN1 can restore the ability of phosphorylated tau to bind microtubules and promote microtubule assembly in vitro. Since depletion of PIN1 induces mitotic arrest and apoptotic cell death, sequestration of PIN1 into PHFs may contribute to neuronal death.

From detailed analysis of pathologic load and spatiotemporal distribution of beta-amyloid deposits and tau pathology in sporadic AD, Delacourte et al. (2002) concluded that there is a synergistic effect of amyloid aggregation in the propagation of tau pathology.

Kayed et al. (2003) produced an antibody that specifically recognized micellar amyloid beta but not soluble, low molecular weight amyloid beta or amyloid beta fibrils. The antibody also specifically recognized soluble oligomers among all other types of amyloidogenic proteins and peptides examined, indicating that they have a common structure and may share a common pathogenic mechanism. Kayed et al. (2003) showed that all of the soluble oligomers tested displayed a common conformation-dependent structure that was unique to soluble oligomers regardless of sequence. The in vitro toxicity of soluble oligomers was inhibited by oligomer-specific antibody. Soluble oligomers have a unique distribution in human Alzheimer disease brain that is distinct from that of fibrillar amyloid. Kayed et al. (2003) concluded that different types of soluble amyloid oligomers have a common structure and suggested that they share a common mechanism of toxicity.

Revesz et al. (2003) reviewed the pathology and genetics of APP-related CAA and discussed the different neuropathologic consequences of different APP mutations. Those that result in increased beta-amyloid-40 tend to result in increased deposition of amyloid in the vessels, consistent with CAA, whereas those that result in increased beta-amyloid-42 tend to result in parenchymal deposition of amyloid and the formation of amyloid plaques. These latter changes are common in classic Alzheimer disease.

To determine whether decreased neprilysin (MME; 120520) levels contribute to the accumulation of amyloid deposits in AD or normal aging, Russo et al. (2005) analyzed MME mRNA and protein levels in cerebral cortex from 10 cognitively normal elderly individuals with amyloid plaques (NA), 10 individuals with AD, and 10 controls who were free of amyloid plaques. They found a significant decrease in MME mRNA levels in both AD and NA individuals compared to controls. Russo et al. (2005) concluded that decreased MME expression correlates with amyloid-beta deposition but not with degeneration and dementia.

Using Western blotting, immunoprecipitation assays, and surface plasmon resonance analysis, Guo et al. (2006) showed that beta-amyloid-40 and -42 formed stable complexes with soluble tau and that prior phosphorylation of MAPT inhibited complex formation. Immunostaining of brain extracts from patients with AD and controls showed that phosphorylated tau and beta-amyloid were present within the same neuron. Guo et al. (2006) postulated that an initial step in AD pathogenesis may be the intracellular binding of soluble beta-amyloid to soluble nonphosphorylated tau.

By neuropathologic examination, Wilkins et al. (2006) found no difference in the presence or degree of neurofibrillary tangles, senile plaques, Lewy bodies, or amyloid angiopathy between 10 African American and 10 white individuals with AD. The findings suggested that race is not a major influence on AD pathology.

In HEK293 cells in vitro, Ni et al. (2006) found that activation of beta-2-adrenergic receptors (ADRB2; 109690) stimulated gamma-secretase activity and beta-amyloid production. Stimulation involved the association of ADRB2 with PSEN1 and required agonist-induced endocytosis of ADRB2. Similar effects were observed after activation of the opioid receptor OPRD1 (165195). In mouse models of AD, chronic treatment with ADRB2 agonists increased cerebral amyloid plaques, and treatment with ADRB2 antagonists reduced cerebral amyloid plaques. Ni et al. (2006) postulated that abnormal activation of ADRB2 receptors may contribute to beta-amyloid accumulation in AD.

Sun et al. (2006) found that hypoxia increased BACE1 (604252) beta-secretase activity and resulted in significantly increased beta-amyloid production in both wildtype human cells and human cells that stably overexpressed an AD-related APP mutation. Studies in transgenic mice with APP mutations showed that hypoxia upregulated Bace1 mRNA and increased deposition of brain beta-A40 and A42 compared to transgenic mice not exposed to hypoxic conditions. The findings suggested that hypoxia can facilitate AD pathogenesis and provided a molecular mechanism that linked vascular factors to AD.

In studies of rodent and human cells, Li et al. (2007) found that overexpression of hyperphosphorylated tau antagonized apoptosis of neuronal cells by stabilizing beta-catenin (CTNNB1; 116806). The findings explained why NFT-bearing neurons survive proapoptotic insults and instead die chronically of degeneration.

Schilling et al. (2008) found that the N-terminal pyroglutamate (pE) formation of amyloid beta (104760) is catalyzed by glutaminyl cyclase (607065) in vivo. Glutaminyl cyclase expression was upregulated in the cortices of individuals with Alzheimer disease and correlated with the appearance of pE-modified amyloid beta. Oral application of a glutaminyl cyclase inhibitor resulted in reduced amyloid beta(3(pE)-42) burden in 2 different transgenic mouse models of Alzheimer disease and in a new Drosophila model. Treatment of mice was accompanied by reductions in amyloid beta(X-40/42), diminished plaque formation and gliosis, and improved performance in context memory and spatial learning tests. Schilling et al. (2008) suggested that their observations were consistent with the hypothesis that amyloid beta(3(pE)-42) acts as a seed for amyloid beta aggregation by self-aggregation and coaggregation with amyloid beta(1-40/42). Therefore, amyloid beta(3(pE)-40/42) peptides seem to represent amyloid beta forms with exceptional potency for disturbing neuronal function. The authors suggested that the reduction of brain pE-modified amyloid beta by inhibition of glutaminyl cyclase offers a new therapeutic option for the treatment of Alzheimer disease and provides implications for other amyloidoses.

In vascular smooth muscle cells isolated from AD patients with CAA, Bell et al. (2009) found an association between beta-amyloid deposition and increased expression of serum response factor (SRF; 600589) and myocardin (MYOCD; 606127) compared to controls. Further studies indicated the MYOCD upregulated SRF and generated a beta-amyloid nonclearing phenotype through transactivation of SREBP2 (600481), which downregulates LRP1, a key beta-amyloid clearance receptor. SRF silencing led to increased beta-amyloid clearance. Hypoxia stimulated SRF/MYOCD expression in human cerebral vascular smooth muscle cells and in animal models of AD. Bell et al. (2009) suggested that SRF and MYOCD function as a transcriptional switch, controlling beta-amyloid cerebrovascular clearance and progression of AD.

Using microarray analysis, followed by RT-PCR of human postmortem hippocampus, Qin et al. (2009) found that decreased expression of the PPARGC1A gene (604517), a regulator of gluconeogenesis, correlated with progression of moderate to severe clinical dementia in patients with AD, as well as increased density of neuritic plaques and beta-amyloid-42. Hyperglycemia was found to attenuate PPARGC1A expression and increase beta-amyloid in the medium of Tg2576 AD neurons; this phenomenon was decreased by exogenous expression of PPARGC1A. Further studies indicated that suppression of PPARGC1A in hyperglycemia resulted in activation of the FOXO3A (602681) transcription factor, which inhibits nonamyloidogenic secretase processing of APP and promotes amyloidogenic processing of APP. The findings provided a molecular mechanism for a link between glucose metabolism and AD.

Mawuenyega et al. (2010) measured amyloid-beta kinetics in the CNS of 12 AD participants and 12 cognitively intact controls. Mawuenyega et al. (2010) found no differences in the rate of production of amyloid-beta-42 or amyloid-beta-40 in AD patients versus controls. However, there was a significant difference in the rate of amyloid-beta-40 and amyloid-beta-42 clearance in the AD subjects versus controls. There was roughly 30% impairment in the clearance of both amyloid-beta-42 and amyloid-beta-40, with a P value of 0.03 and 0.01, respectively. Estimates based on a 30% decrease in amyloid-beta clearance rate suggested that brain amyloid-beta accumulates over about 10 years in AD. The authors pointed out that the limitations of this study included the relatively small number of participants and the inability to prove causality of impaired amyloid-beta clearance for AD.

Israel et al. (2012) reprogrammed primary fibroblasts from 2 patients with familial Alzheimer disease, in both caused by a duplication of the amyloid-beta precursor protein gene (APP; 104760), 2 with sporadic Alzheimer disease, and 2 nondemented control individuals into induced pluripotent stem cell (iPSC) lines. Neurons from differentiated cultures were purified with fluorescence-activated cell sorting and characterized. Purified cultures contained more than 90% neurons, clustered with fetal brain mRNA samples by microarray criteria, and could form functional synaptic contacts. Virtually all cells exhibited normal electrophysiologic activity. Relative to controls, iPSC-derived, purified neurons from the 2 patients with the duplication and 1 sporadic patient exhibited significantly higher levels of the pathologic markers of amyloid-beta(1-40), phospho-tau(thr231), and active glycogen synthase kinase-3-beta (aGSK-3-beta). Neurons from the duplication and the same sporadic patient also accumulated large RAB5 (179512)-positive early endosomes compared to controls. Treatment of purified neurons with beta-secretase inhibitors, but not gamma-secretase inhibitors, caused significant reductions in phospho-tau(thr231) and aGSK-3-beta levels. Israel et al. (2012) concluded that their results suggested a direct relationship between APP proteolytic processing, but not amyloid-beta, in GSK-3-beta activation and tau phosphorylation in human neurons. Additionally, Israel et al. (2012) observed that neurons with the genome of 1 of the sporadic patients exhibited the phenotypes seen in familial Alzheimer disease samples.

Laganowsky et al. (2012) identified a segment of the amyloid-forming protein alpha-B crystallin (123590) that forms an oligomeric complex exhibiting properties of other amyloid oligomers: beta-sheet-rich structure, cytotoxicity, and recognition by an oligomer-specific antibody. The x-ray-derived atomic structure of the oligomer revealed a cylindrical barrel formed from 6 antiparallel protein strands that Laganowsky et al. (2012) termed a cylindrin. The cylindrin structure is compatible with a sequence segment from the beta-amyloid protein of Alzheimer disease. Laganowsky et al. (2012) concluded that cylindrins offer models for the hitherto elusive structures of amyloid oligomers.

Amino-terminally truncated, pyroglutamylated (pE) forms of amyloid-beta are strongly associated with Alzheimer disease, are more toxic than amyloid-beta(1-42) and amyloid-beta(1-40), and have been proposed as initiators of Alzheimer disease pathogenesis. Nussbaum et al. (2012) reported a mechanism by which pE-amyloid-beta may trigger Alzheimer disease. Amyloid-beta-3(pE)-42 co-oligomerizes with excess amyloid-beta(1-42) to form metastable low-n oligomers (LNOs) that are structurally distinct and far more cytotoxic to cultured neurons than comparable LNOs made from amyloid-beta(1-42) alone. Tau (157140) is required for cytotoxicity, and LNOs comprising 5% amyloid-beta-3(pE)-42 plus 95% amyloid-beta(1-42) (5% pE-amyloid-beta) seed new cytotoxic LNOs through multiple serial dilutions into amyloid-beta(1-42) monomers in the absence of additional amyloid-beta-3(pE)-42. LNOs isolated from human Alzheimer disease brain contained amyloid-beta-3(pE)-42, and enhanced amyloid-beta-3(pE)-42 formation in mice triggered neuron loss and gliosis at 3 months, but not in a tau-null background. Nussbaum et al. (2012) concluded that amyloid-beta-3(pE)-42 confers tau-dependent neuronal death and causes template-induced misfolding of amyloid-beta(1-42) into structurally distinct LNOs that propagate by a prion-like mechanism. Nussbaum et al. (2012) concluded that their results raised the possibility that amyloid-beta-3(pE)-42 acts similarly at a primary step in Alzheimer disease pathogenesis.

Raj et al. (2014) performed an expression quantitative trait locus (eQTL) study of purified CD4 (186940)+ T cells and monocytes, representing adaptive and innate immunity, in a multiethnic cohort of 461 healthy individuals. Context-specific cis- and trans-eQTLs were identified, and cross-population mapping allowed, in some cases, putative functional assignment of candidate causal regulatory variants for disease-associated loci. Raj et al. (2014) noted an overrepresentation of monocyte-specific eQTLs among Alzheimer disease and Parkinson disease (168600) variants, and of T cell-specific eQTLs among susceptibility alleles for autoimmune diseases, including rheumatoid arthritis (180300) and multiple sclerosis (126200). Raj et al. (2014) concluded that this polarization implicates specific immune cell types in these diseases and points to the need to identify the cell-autonomous effects of disease susceptibility variants.

Using solid-state nuclear magnetic resonance (ssNMR) measurements on amyloid beta-40 and amyloid beta-42 fibrils prepared by seeded growth from extracts of Alzheimer disease brain cortex, Qiang et al. (2017) investigated correlations between structural variation and Alzheimer disease phenotype. The authors compared 2 atypical Alzheimer disease clinical subtypes, the rapidly progressive form (r-AD) and the posterior cortical atrophy variant (PCA-AD), with a typical prolonged-duration form (t-AD). On the basis of ssNMR data from 37 cortical tissue samples from 18 individuals, Qiang et al. (2017) found that a single amyloid beta-40 fibril structure is most abundant in samples from patients with t-AD and PCA-AD, whereas amyloid beta-40 fibrils from r-AD samples exhibit a significantly greater proportion of additional structures. Data for amyloid beta-42 fibrils indicated structural heterogeneity in most samples from all patient categories, with at least 2 prevalent structures. Qiang et al. (2017) concluded that these results demonstrated the existence of a specific predominant amyloid beta-40 fibril structure in t-AD and PCA-AD, suggested that r-AD may relate to additional fibril structures, and indicated that there is a qualitative difference between amyloid beta-40 and amyloid beta-42 aggregates in the brain tissue of patients with Alzheimer disease.

In patients with Alzheimer disease, deposition of amyloid-beta is accompanied by activation of the innate immune system and involves inflammasome-dependent formation of ASC (606838) specks in microglia. ASC specks released by microglia bind rapidly to amyloid-beta and increase the formation of amyloid-beta oligomers and aggregates, acting as an inflammation-driven cross-seed for amyloid-beta pathology. Venegas et al. (2017) showed that intrahippocampal injection of ASC specks resulted in spreading of amyloid-beta pathology in transgenic double-mutant APP(Swe)PSEN1(dE9) mice. By contrast, homogenates from brains of APP(Swe)PSEN1(dE9) mice failed to induce seeding and spreading of amyloid-beta pathology in ASC-deficient double-mutant mice. Moreover, coapplication of an anti-ASC antibody blocked the increase in amyloid-beta pathology in the double-mutant mice. Venegas et al. (2017) concluded that these findings supported the concept that inflammasome activation is connected to seeding and spreading of amyloid-beta pathology in patients with Alzheimer disease.

In mice, Da Mesquita et al. (2018) demonstrated that meningeal lymphatic vessels drain macromolecules from the CNS (cerebrospinal and interstitial fluids) into the cervical lymph nodes. Impairment of meningeal lymphatic function slowed paravascular influx of macromolecules into the brain and efflux of macromolecules from the interstitial fluid, and induced cognitive impairment in mice. Treatment of aged mice with vascular endothelial growth factor C (VEGFC; 601528) enhanced meningeal lymphatic drainage of macromolecules from the cerebrospinal fluid, improving brain perfusion and learning and memory performance. Disruption of meningeal lymphatic vessels in transgenic mouse models of Alzheimer disease promoted amyloid-beta deposition in the meninges, which resembles human meningeal pathology, and aggravated parenchymal amyloid-beta accumulation. Da Mesquita et al. (2018) suggested that meningeal lymphatic dysfunction may be an aggravating factor in Alzheimer disease pathology and in age-associated cognitive decline.

Inheritance

From an extensive study in Sweden, Sjogren et al. (1952) suggested that Alzheimer disease shows multifactorial inheritance. In a study of 52 families with AD, Masters et al. (1981) concluded that the disorder showed autosomal dominant inheritance without maternal effect.

In 7 of 21 families with AD, Powell and Folstein (1984) found evidence of 3-generation transmission. Breitner and Folstein (1984) suggested that most cases of Alzheimer disease are familial. Fitch et al. (1988) found a familial incidence of 43%, and detected no clinical differences between the familial and sporadic cases. In one-third of the familial cases, the disorder developed after age 70. Breitner et al. (1988) found that the cumulative incidence of AD among relatives was 49% by age 87. The risk was similar among parents and sibs, and did not differ significantly between relatives of those with early or late onset.

In a study of 70 kindreds containing 541 affected and 1,066 unaffected offspring of parents with AD parents, Farrer et al. (1990) identified 2 distinct clinical groups: early onset (less than 58 years) and late onset (greater than 58 years). At-risk offspring in early-onset families had an estimated lifetime risk for dementia of 53%, suggesting autosomal dominant inheritance. The lifetime risk in late-onset families was 86%. Farrer et al. (1990) concluded that late-onset AD may be autosomal dominant in some families.

In a complex segregation analysis on 232 nuclear families ascertained through a single proband who was referred for diagnostic evaluation of memory disorder, Farrer et al. (1991) concluded that susceptibility to AD is determined, in part, by a major autosomal dominant allele with an additional multifactorial component. The frequency of the AD susceptibility allele was estimated to be 0.038, but the major locus was thought to account for only 24% of the 'transmission variance,' indicating a substantial role for other genetic and nongenetic mechanisms.

Silverman et al. (1994) used a standardized family history assessment to study first-degree relatives of Alzheimer disease probands and nondemented spouse controls. First-degree relatives of AD probands had a significantly greater cumulative risk of AD (24.8%) than did the relatives of spouse controls (15.2%). The cumulative risk for the disorder among female relatives of probands was significantly greater than that among male relatives.

Rao et al. (1996) carried out a complex segregation analysis in 636 nuclear families of consecutively ascertained and rigorously diagnosed probands in the Multi-Institutional Research in Alzheimer Genetic Epidemiology study in order to derive models of disease transmission that account for the influences of the APOE genotype of the proband and gender. In the total group of families, models postulating sporadic occurrence, no major gene effect, random environmental transmission, and mendelian inheritance were rejected. Transmission of AD in families of probands with at least 1 APOE4 allele best fitted a dominant model. Moreover, single gene inheritance best explained clustering of the disorder in families of probands lacking APOE4, but a more complex genetic model or multiple genetic models may ultimately account for risk in this group of families. The results suggested to Rao et al. (1996) that susceptibility to AD differs between men and women regardless of the proband's APOE status. Assuming a dominant model, AD appeared to be completely penetrant in women, whereas only 62 to 65% of men with predisposing genotypes developed AD. However, parameter estimates from the arbitrary major gene model suggested that AD is expressed dominantly in women and additively in men. These observations, taken together with epidemiologic data, were considered consistent with the hypothesis of an interaction between genes and other biologic factors affecting disease susceptibility.

In a study of 290 patients with Alzheimer disease in the French Collaborative Group and 1,176 of their first-degree relatives, Martinez et al. (1998) found that familial clustering of Alzheimer disease was largely due to factors other than APOE status.

Silverman et al. (1999) hypothesized that elderly individuals who lived beyond the age of 90 years without dementia had a concentration of genetic protective factors against Alzheimer disease. Although they recognized that testing this hypothesis was complicated, probands carrying genetic protective factors should have relatives with lower illness rates not only for early-onset disease, in which genetic risk factors are a strong contributing factor to the incidence of AD, but also for later-onset disease, when the role of these factors appears to be markedly diminished. AD dementia was assessed through family informants in 6,660 first-degree relatives of 1,049 nondemented probands aged 60 to 102 years. Cumulative survival without AD was significantly greater in the relatives of the oldest proband group (aged 90 to 102 years) than it was in the 2 younger groups. In addition, the reduction in the rate of illness for this group was relatively constant across the entire late life span. The results suggested that genetic factors conferring a lifelong reduced liability to AD may be more highly concentrated among nondemented probands aged 90 or more years and their relatives.

Gatz et al. (2006) evaluated genetic and environmental influences on Alzheimer disease in a population of like- and unlike-sex twin pairs (11,884 twin pairs, 392 with one or both members diagnosed with AD from the Swedish Twin Registry; participants were 65 years of age or older). Participants were divided into 5 quantitative genetic groups; male/female monozygotic twins, male/female dizygotic twins, and unlike-sex twins. On the basis of screening for cognitive dysfunction and environmental variables, estimates on heritability, shared environmental influences, and nonshared environmental influences, adjusted for age, were derived from the twin data. Heritability for AD was estimated to be 58% in the full model and 79% in the best-fitting model with the balance of variation explained by nonshared environmental influences. There were no significant differences between men and women in prevalence or heritability after controlling for age. In pairs concordant for AD, intrapair difference in age at onset was significantly greater in dizygotic than in monozygotic pairs, suggesting genetic influences on timing of the disease.

Autosomal Recessive Inheritance

Bowirrat et al. (2000) presented data they interpreted as suggesting an autosomal recessive form of AD. They screened all 821 elderly residents of an Arab community located in Wadi Ara, northern Israel. An unusually high prevalence of AD was observed (20% of those 65 years old or older; 60.5% of those 85 years old or older). Data on the APOE4 allele suggested that it could not explain the AD prevalence in this population. The APOE4 allele was relatively uncommon in Arabs in Wadi Ara; in fact, Bowirrat et al. (2000) stated that it was the lowest frequency of the allele ever recorded. Because of the high consanguinity rate of Arab marriages in Israel, Bowirrat et al. (2000) speculated that recessive genes for AD exist and are responsible for the high AD prevalence in Wadi Ara. Further information was provided by Bowirrat et al. (2001) and Bowirrat et al. (2002). Bowirrat et al. (2002) reported on vascular dementia among elderly Arabs in the same area.

A form of AD mapped to chromosome 10q24, AD6 (605526), showed some evidence of autosomal recessive inheritance.

Di Fede et al. (2009) identified a homozygous mutation in the APP gene (A673V; 104760.0022) in a patient with early-onset progressive AD beginning at age 36 years. He was noncommunicative and could not walk by age 44. Serial MRI showed progressive cortico-subcortical atrophy, and cerebrospinal fluid analysis showed decreased A-beta-1-42 and increased total and 181T-phosphorylated tau compared to controls and similar to subjects with Alzheimer disease. The mutation was also found in homozygosity in the proband's younger sister, who had multiple domain mild cognitive impairment (MCI), believed to a high risk condition for the development of clinically probable Alzheimer disease (Petersen et al., 2001). In the plasma of both the patient and his homozygous sister, amyloid-beta-1-40 and amyloid-beta-1-42 were higher than in nondemented controls, whereas the A673V heterozygous carriers from the family that were tested had intermediate amounts. None of 6 heterozygous individuals in the family had any evidence of dementia when tested at ages ranging from 21 to 88. The A673V mutation, which corresponds to position 2 of amyloid beta, affected APP processing, resulting in enhanced beta-amyloid production and formation of amyloid fibrils in vitro. Coincubation of mutated and wildtype peptides conferred instability on amyloid beta aggregates and inhibited amyloidogenesis and neurotoxicity. Di Fede et al. (2009) concluded that the interaction between mutant and wildtype amyloid beta, favored by the A-to-V substitution at position 2, interferes with nucleation or nucleation-dependent polymerization or both, hindering amyloidogenesis and neurotoxicity and thus protecting the heterozygous carriers.

Diagnosis

Croes et al. (2000) argued against using genetic testing for Alzheimer disease as a diagnostic tool. They suggested that the contribution of genetic testing to clinical diagnosis is small and does not counterbalance the problems associated either with interpretation or with secondary effects on family members.

Itoh et al. (2001) proposed a CSF analysis of hyperphosphorylated tau protein (phosphorylation at serine 199; tau-199) for the antemortem diagnosis of AD. In over 500 patients with dementia, including 236 believed to have AD, there was a significant increase in the tau-199 levels in the AD group compared to the non-AD group. Itoh et al. (2001) noted that the tau-199 test exceeds both sensitivity and specificity over 85% as a sole biomarker of AD; however, they also noted that many of the non-AD tauopathy and degenerative dementias also showed increased tau-199 levels.

Among 131 patients with AD and 72 healthy controls, Sunderland et al. (2003) found significantly lower levels of beta-amyloid(1-42) and significantly higher levels of tau in the CSF of AD patients than in the CSF of controls. However, the data showed considerable variance, with significant overlap between the groups. Metaanalysis of previous studies comparing these markers demonstrated similar findings. The authors suggested that CSF beta-amyloid and tau are biologic markers of AD pathophysiology and that the measures may have potential clinical utility in the future diagnosis of AD.

Among 78 patients with mild cognitive impairment, 23 of whom developed dementia, Herukka et al. (2005) found that a combination of low CSF beta-amyloid-42 and high CSF tau and phosphorylated tau was associated with the development of dementia. The high positive likelihood ratio indicated that combined biomarker tests were useful in confirming the diagnosis of AD, but the low negative likelihood ratio indicated that a negative test result could not rule out the disease. The sensitivity of beta-amyloid-42 and phosphorylated tau ranged from 60.0 to 66.7%, and specificity ranged from 84.6 to 89.7%. Herukka et al. (2005) concluded that changes in CSF biomarkers occur early in the course of AD in most patients.

In a study of 22 patients with AD, Hampel et al. (2005) found a correlation between levels of CSF phosphorylated tau and hippocampal atrophy, independent of disease duration and severity. The authors suggested that CSF phosphorylated tau levels may reflect neuronal damage in AD.

Iqbal et al. (2005) classified 353 AD patients into at least 5 subgroups based on CSF levels of beta-amyloid-42, tau, and ubiquitin. Each subgroup presented a different clinical profile, and the authors suggested that the subgroups may benefit from different therapeutic drugs.

Among 184 healthy individual with normal cognition aged 21 to 88 years, Peskind et al. (2006) found that the concentration of CSF beta-amyloid-42, but not beta-amyloid-40, decreased with age. Those with an APOE4 allele showed a sharp and significant decline in CSF beta-A-42 beginning in the sixth decade compared to those without the APOE4 allele. The findings were consistent with APOE4-modulated acceleration of pathogenic beta-A-42 deposition starting in late middle age in persons with normal cognition, and suggested that early treatment for AD in susceptible individuals may be necessary in midlife or earlier.

In a study of 211 cognitively normal controls, 98 patients with early symptomatic AD, and 19 individuals with other forms of dementia, Tarawneh et al. (2011) found a significant difference in CSF VILIP1 (600817) levels, with higher levels in AD compared to the other 2 groups. CSF VILIP1 levels correlated with CSF tau and phosphorylated-tau181, and negatively correlated with brain volumes in AD. VILIP1 and VILIP1/beta-amyloid-42 predicted future cognitive impairment in the normal controls over the follow-up period. Importantly, this CSF ratio (VILIP1/beta-amyloid-42) predicted future cognitive impairment at least as well as tau/beta-amyloid-42 and p-tau181/beta-amyloid-42. VILIP1 is abundantly expressed in neurons and has been shown to be a marker of neuronal injury in brain injury models (Laterza et al., 2006). The findings of Tarawneh et al. (2011) suggested that CSF VILIP1 and VILIP1/beta-amyloid-42 may offer diagnostic utility for early AD and can predict future cognitive impairment in cognitively normal individuals.

Clinical Management

Donepezil is a specific piperidine-based inhibitor of acetylcholinesterase (AChE) used for the treatment of mild to moderate Alzheimer disease with variable efficacy. Pilotto et al. (2009) examined a group of 115 white AD patients taking the medication, including 69 (60%) responders and 46 patients (40%) nonresponders. Nonresponders had a significantly higher frequency of the -1584G allele (rs1080985) in the CYP2D6 gene (124030) compared to responders (58.7% vs 34.8%, p = 0.013), with an odds ratio of 3.43 for poor response. The -1584G allele is associated with higher enzymatic activity and more rapid drug metabolism. The findings suggested that the rs1080985 SNP in the CYP2D6 gene may influence the clinical efficacy of donepezil in AD patients.

Salloway et al. (2009) found insufficient evidence to support or refute the benefit of the use of bapineuzumab, an anti-beta-amyloid monoclonal antibody, in a randomized control trial of 234 AD patients. However, there was some evidence to suggest improved cognitive and functional endpoints in APOE E4 noncarriers, which supported further investigation. Vasogenic edema in the brain, which occurred in 9.7% of treated patients and none of untreated patients, was identified as a potential side effect, particularly in APOE E4 carriers.

Mapping

Early Linkage Studies

Wheelan and Race (1959) studied a family in which the mother and 5 of 10 children were affected. Possible linkage with the MNS locus was found.

In the large AD kindred reported by Nee et al. (1983), Weitkamp et al. (1983) concluded that genes in the HLA region of chromosome 6 and perhaps also in the Gm region of chromosome 14 are determinants of susceptibility. The association between immunoglobulins and the amyloid in the senile plaque of AD was thought to be significant in this connection. The peak lod score with Gm was 1.37 (at theta = 0.05). Nerl et al. (1984) reported an increase in the frequency of a complement component-4B allele (C4B; 120820) on chromosome 6p21 in patients with AD, but Eikelenboom et al. (1988) failed to find a significant association between C4*B2 allelic frequency and AD.

Linkage to Chromosome 21q

Delabar et al. (1986) analyzed DNA from 4 patients with a phenotype of trisomy 21 and dementia of the Alzheimer type, but who had normal karyotypes. In all 4 cases, duplication of the ETS2 locus (164740) was found, whereas SOD1 (147450) was normal. Chemical investigations and DNA analyses indicated partial trisomy due to duplication of a short segment of chromosome 21, located at the interface between 21q21 and 21q22.1 and carrying the SOD1 and ETS2 genes.

In 4 extensive kindreds with early-onset AD, St. George-Hyslop et al. (1987) found linkage to DNA markers on the centromeric side of chromosome 21q11.2-21q21. The markers in band 21q22, critical to the development of Down syndrome, showed negative lod scores. There was not tight linkage to the SOD1 gene. Using a RFLP of SOD1 in the study of a large AD family David et al. (1988) also concluded that AD and SOD1 are not closely linked.

By somatic cell hybridization and linkage studies, Tanzi et al. (1987) localized the gene responsible for beta-amyloid deposition in Down syndrome to the same vicinity on chromosome 21 as that responsible for AD.

Haines et al. (1987), who studied 4 large families with FAD, found linkage with 2 DNA markers on chromosome 21 that had previously been shown to be linked to each other at a distance of 8 cM. Pair-wise linkage analysis showed a lod score of 2.37 at theta = 0.08 for one and 2.32 at theta = 0.00 for the other. The use of multipoint analysis provided stronger evidence for linkage with a peak score of 4.25.

Blanquet et al. (1987) found that the APP gene and the ETS2 oncogene are distally located. Surprisingly, 2 hybridization peaks were observed for ETS2 in patients with AD, 1 at the normal site of the oncogene and 1 at the site of the amyloid protein. Blanquet et al. (1987) interpreted these results as indicating that AD is associated with a complex rearrangement within chromosome 21, by which 2 distantly related genes come to lie in the vicinity of each other.

Pulst et al. (1989) used a panel of aneuploid cell lines containing various regions of human chromosome 21 to map the physical order of DNA probes linked to the FAD locus. Van Camp et al. (1989) described the isolation of 35 chromosome 21-specific DNA probes for analysis in Alzheimer disease and Down syndrome. Ross et al. (1989) described the isolation of cDNAs from brain and spinal cord, mapping to chromosome 21, for investigation in Alzheimer disease. Using pulsed field gel electrophoresis to construct a physical map of the region of chromosome 21 around the FAD locus, Owen et al. (1989) suggested the following order: cen--D21S16--D21S48--D21S13--D21S46--(D21S52, D21S4)--(D21S1, D21S11).

Van Broeckhoven et al. (1988) concluded that the gene for early-onset familial AD was located close to the centromere of chromosome 21. In 2 AD families, Van Broeckhoven et al. (1989) found linkage to chromosome 21. Results of 1 family yielded a lod score of 1.52 at marker D21S13. Further studies yielded a peak lod score of 6.24 at D21S16. Using genetic linkage analysis, Goate et al. (1989) found a peak lod score of 3.3 between the familial AD locus and locus D21S16.

St. George-Hyslop et al. (1990), including many members of the FAD collaborative study group, undertook a study of 5 polymorphic chromosome 21 markers in a large unselected series of pedigrees with FAD. The results seemed to indicate that, in many families at least, early-onset AD is due to a mutation on chromosome 21, whereas late-onset AD has other causes.

Lawrence et al. (1992) reviewed the reported data on multiplex Alzheimer pedigrees for which lod scores had been reported; the AD1 locus that mapped to the site of the APP locus on 21q accounted for 63 +/- 11% of these pedigrees. The AD1/APP locus was placed at approximately 27.7 Mb from pter, corresponding to genetic intervals of 10.9 cM in males and 33.9 cM in females, flanked proximally by D21S8 and distally by D21S111. There was no evidence in this analysis for a second locus on chromosome 21.

Olson et al. (2001) reported convincing evidence of a major role for the APP locus in late-onset AD. They used a covariate-based affected-sib-pair linkage method to analyze the chromosome 21 clinical and genetic data obtained on affected sibships by the Alzheimer Disease Genetics Initiative of the National Institute of Mental Health. A lod score of 5.54 (P = 0.000002) was obtained when age at last examination/death was included in the linkage model, and a lod score of 5.63 (P = 0.000006) was obtained when age at onset and disease duration were included. Olson et al. (2001) concluded that the APP locus may predispose to AD in the very elderly.

In further use of a covariate-based linkage method to reanalyze genome scan data, Olson et al. (2002) determined that a region on chromosome 20p (AD8; 607116) showed the same linkage pattern to very-late-onset AD as APP. Two-locus analysis provided evidence of strong epistasis between 20p and the APP region, limited to the oldest age group and to those lacking E4 alleles at the APOE locus. Olson et al. (2002) speculated that high-risk polymorphisms in both regions produce a biologic interaction between these 2 proteins that increases susceptibility to a very-late-onset form of AD.

Genetic Heterogeneity

In several families with AD, Van Broeckhoven et al. (1987), Tanzi et al. (1987), and Pulst et al. (1991) excluded linkage to chromosome 21q, indicating genetic heterogeneity.

Percy et al. (1991) described 2 sisters thought to have late-onset AD who also had an unusual chromosome 22-derived marker with a greatly elongated short arm containing 2 well-separated nucleolus organizer regions. Eleven of 24 of their biologic relatives were also found to have the marker; individuals with the marker were 4 times more likely to develop AD.

Zubenko et al. (1998) performed an association study with 391 simple sequence tandem repeat polymorphisms, comparing DNA from 100 autopsied brains with AD, 50 control brains, and 50 nondemented nonagenarians. The strongest association was seen with marker D19S178, presumably reflecting association with APOE. In addition, weaker associations were seen with 5 other markers, D1S518 (1q31-q32.1), D1S547 (1q44), D10S1423 (10p12-p14), D12S1045 (12q24.3), and DXS1047 (Xq25), suggesting the possibility of other susceptibility genes.

In a study in eastern Finland, Hiltunen et al. (1999) found an association between AD and 2 markers on chromosome 13q12 (D13S787 and D13S292.) The 13q12 locus was associated with female familial AD patients regardless of APOE genotype. The 2 markers were estimated to reside in an 810-kb YAC clone together with 2 ESTs derived from infant brain and the ATP1AL1 (182360) gene.

Blacker et al. (2003) performed a 9-cM genome screen of 437 families with AD, comprising the full National Institute of Mental Health sample. In standard parametric and nonparametric linkage analyses, they observed a 'highly significant' linkage peak by the criteria of Lander and Kruglyak (1995) on chromosome 19q13, which probably represented APOE. Twelve additional locations, 1q23, 3p26, 4q32, 5p14, 6p21, 6q27, 9q22, 10q24, 11q25, 14q22, 15q26, and 21q22, met criteria for 'suggestive' linkage.

Scott et al. (2003) considered age of onset as a covariant in the analysis of data from 336 markers in 437 multiplex white AD families. A statistically significant increase in the nonparametric multipoint lod score was observed on 2q34, with a peak lod score of 3.2 at D2S2944 in 31 families with a minimum age at onset between 50 and 60 years. Lod scores were also significantly increased on 15q22. The results indicated that linkage to regions on 2q34 and 15q22 were linked to early-onset AD and very-late-onset AD, respectively.

Holmans et al. (2005) performed linkage analyses on 28 sib pairs with late-onset AD. Linkage was observed with chromosome 21 for age-at-onset effects (lod = 2.57). This association was strongest in pairs with mean age at onset greater than 80 years. A similar effect was observed on chromosome 2q (maximum lod = 2.73). Suggestive evidence was observed for age at onset on chromosome 19q (maximum lod = 2.33) and in the vicinity of APOE at 12p (maximum lod = 2.22). Mean rate of decline showed suggestive evidence of linkage to chromosome 9q (maximum lod = 2.29). Holmans et al. (2005) observed suggestive evidence of increased identical by descent in APOE4 homozygotes on chromosome 1 (maximum lod = 3.08) and chromosome 9 (maximum lod = 3.34).

Sillen et al. (2006) conducted a genomewide linkage study on 188 individuals with AD from 71 Swedish families, using 365 markers (average intermarker distance 8.97 cM). They performed nonparametric linkage analyses in the total family material as well as stratified the families with respect to the presence or absence of APOE4. The results suggested that the disorder in these families was tightly linked to the APOE region (19q13). The next highest lod score was to chromosome 5q35, and no linkage was found to chromosomes 9, 10, and 12.

Katzov et al. (2004) presented evidence that both single marker alleles and haplotypes of the ABCA1 gene (600046) may contribute to variable cerebrospinal fluid MAPT and APP levels, and brain beta-amyloid load. The results indicated that variants of ABCA1 may affect the risk of AD, providing support for a genetic link between AD and cholesterol metabolism. In 42 individuals with AD, Katzov et al. (2006) found an association between increased CSF cholesterol and beta-amyloid protein levels. In a study of 1,567 Swedish dementia cases, including 1,275 with Alzheimer disease, and 2,203 controls, Reynolds et al. (2009) found an association between rs2230805 in the ABCA1 gene on chromosome 9q22 and dementia risk (odds ratio of 1.39; p = 7.7 x 10 (-8)). The putative risk allele of rs2230805 was also found to be associated with reduced cerebrospinal fluid levels of beta-amyloid.

Rogaeva et al. (2007) reported that inherited variants of the SORL1 (602005) neuronal sorting receptor on chromosome 11q23 are associated with late-onset Alzheimer disease. These variants, which occur in at least 2 different clusters of intronic sequences within the SORL1 gene, may regulate tissue-specific expression of SORL1. Lee et al. (2007) reported associations between various SNPs and haplotypes in the SORL1 gene and AD among a total of 296 AD patients comprising 3 cohorts of African American, Caribbean Hispanic, and non-Hispanic white individuals. The findings suggested extensive allelic heterogeneity in SORL1, with specific SNPs associated with specific groups. Cellini et al. (2009) also reported an association between SNPs in the SORL1 gene (rs661057, rs12364988, and rs641120) and LOAD among 251 Italian patients with LOAD and 358 healthy controls (p = 0.002 to 0.03; odds ratio, 1.27 to 1.47). There was a more significant association in women, suggesting that SORL1 may possibly affect LOAD through a female-specific mechanism. By metaanalysis of previous studies including 12,464 cases and 17,929 controls of white or Asian descent, Reitz et al. (2011) showed that multiple SORL1 alleles in distinct linkage disequilibrium blocks are associated with risk for AD in white and Asian populations, demonstrating intralocus heterogeneity in the associations with this gene. Reitz et al. (2011) concluded that their findings provided confirmatory evidence of the association of multiple SORL1 variants with AD risk.

Harold et al. (2009) undertook a 2-stage genomewide association study of Alzheimer disease involving 16,000 individuals, which they stated was the most powerful AD GWAS to date. They observed genomewide association with a SNP in the intron of the CLU gene (APOJ; 185430) not previously associated with the disease: rs11136000, P = 1.4 x 10(-9). This association was replicated in stage 2 (2,023 cases and 2,340 controls), producing compelling evidence for association with Alzheimer disease in the combined dataset (P = 8.5 and 10(-10), odds ratio = 0.86).

Lambert et al. (2009) conducted a large genomewide association study of 2,032 individuals from France with Alzheimer disease and 5,328 controls. Markers outside APOE with suggestive evidence of association (P less than 10(-5)) were examined in collections from Belgium, Finland, Italy, and Spain totaling 3,978 Alzheimer disease cases and 3,297 controls. Two loci gave replicated evidence of association: one with CLU, encoding clusterin or apolipoprotein J, on chromosome 8 (rs11136000, odds ratio = 0.86, 95% confidence interval 0.81-0.90, P = 7.5 x 10(-9) for combined data) and the other within CR1 (120620), encoding the complement component (3b/4b) receptor 1, on chromosome 1 (rs6656401, odds ratio = 1.21, 95% confidence interval 1.14-1.29, P = 3.7 x 10(-9) for combined data). Lambert et al. (2009) stated that previous biologic studies supported roles of CLU and CR1 in the clearance of beta-amyloid.

Carrasquillo et al. (2010) replicated the findings of Harold et al. (2009) and Lambert et al. (2009). Among 1,829 Caucasian LOAD cases and 2,576 controls, Carrasquillo et al. (2010) found significant associations with CLU (rs11136000; OR of 0.82, p = 8.6 x 10(-5)), CR1 (rs3818361; OR of 1.15, p = 0.014), and PICALM (rs3851179; OR of 0.80; 1.3 x 10(-5)). All associations remained significant even after Bonferroni correction.

By metaanalysis, Jun et al. (2010) also replicated the findings of Harold et al. (2009) and Lambert et al. (2009). Among 7,070 AD cases and 8,169 controls from 12 different studies of different populations, Jun et al. (2010) found significant associations, after adjusting for age, sex, and APOE status, between LOAD and rs11136000 in CLU (OR of 0.92; p = 0.0096), rs3818361 in CR1 (OR of 1.15; p = 0.0002), and rs3851179 in PICALM (OR of 0.93; p = 0.026), but only in whites. No SNP was significantly associated with AD in the other ethnic groups. The association with CLU was only evident among those without the APOE E4 allele, and the association with PICALM was only evident among those with the APOE E4 allele.

In a genomewide association study of 549 Caribbean Hispanic patients with LOAD and 544 controls, Lee et al. (2011) found that none of the SNPs studied showed a significant association of p = 7.97 x 10(-8) or lower. The strongest evidence for association was with rs9945493 (p = 1.7 x 10(-7); OR of 0.33) on chromosome 18q23. Candidate genes implicated included CUGBP2 (602538) on chromosome 10p13 in APOE E4 carriers and DGKB (604070) on chromosome 7p21. Among Caribbean Hispanics, there was an association between rs881146 in CLU and LOAD (p = 0.002) in APOE E4 carriers, but not with rs11136000. There was a marginal association with rs17159904 in PICALM (p = 0.04) in APOE E4 noncarriers, and with rs7561528 in BIN1 (p = 0.0054) in APOE E4 carriers.

Hollingworth et al. (2011) undertook a combined analysis of 4 genomewide association datasets (stage 1) and identified 10 newly associated variants with p = 1 x 10(-5) or less. They tested these variants for association in an independent sample (stage 2). Three SNPs at 2 loci replicated and showed evidence for association in a further sample (stage 3). Metaanalyses of all data provided compelling evidence that ABCA7 (rs3764650, meta p = 4.5 x 10(-17); including the Alzheimer's Disease Genetic Consortium (ADGC) data, meta p = 5.0 x 10(-21)) and the MS4A gene cluster (rs610932, meta p = 1.8 x 10(-14); including ADGC data, meta p = 1.2 x 10(-16)) were novel Alzheimer disease susceptibility loci.

In a longitudinal study of 1,666 individuals, including 404 (24%) who developed AD at some point, Chibnik et al. (2011) found a significant association between each additional risk allele (A) of rs6656401 in the CR1 gene and faster rate of global cognitive decline (p = 0.011). There was also an association between this risk allele and AD-related amyloid plaques on neuropathology (p = 0.025) in those with postmortem brain material available. For the PICALM locus, there was a trend for faster rate of cognitive decline associated with 2 copies of the risk allele (G) of rs7110631 (p = 0.03). No association was observed between rate of cognitive decline and rs11136000 in the CLU gene.

Reynolds et al. (2010) conducted dense linkage disequilibrium (LD) mapping of a series of 25 genes putatively involved in lipid metabolism in 1,567 Swedish dementia cases (including 1,275 with possible or probable Alzheimer disease (AD)) and 2,203 Swedish controls. Two markers near SREBF1 (184756) in a 400-kb linkage disequilibrium (LD) block on chromosome 17p had significant association after multiple testing correction. Secondary analyses of gene expression levels of candidates within the LD region together with an investigation of gene network context highlighted 2 possible susceptibility genes, ATPAF2 (608918) and TOM1L2. Reynolds et al. (2010) identified several markers in strong LD with rs3183702 that were significantly associated with AD risk in other genomewide association studies with similar effect sizes.

Molecular Genetics

Familial Alzheimer Disease 1

In affected members of 2 families with AD1, Goate et al. (1991) identified a mutation in the APP gene (V717I; 104760.0002). The average age of onset in 1 family was 57 +/- 5 years. The same mutation was found by Naruse et al. (1991) in 2 unrelated Japanese cases of familial early-onset AD, and Yoshioka et al. (1991) found it in a third Japanese family with AD.

In affected members of 2 large Swedish families with early-onset familial Alzheimer disease, Mullan et al. (1992) identified a double mutation in exon 16 of the APP gene (104760.0008). The 2 families were found to be linked by genealogy.

Protection Against Alzheimer Disease

Jonsson et al. (2012) searched for low-frequency variants in the amyloid-beta precursor protein gene with a significant effect on the risk of Alzheimer disease by studying coding variants in APP in a set of whole-genome sequence data from 1,795 Icelanders. Jonsson et al. (2012) found a coding mutation (A673T; 104760.0023) in the APP gene that protects against Alzheimer disease and cognitive decline in the elderly without Alzheimer disease. This substitution is adjacent to the aspartyl protease beta-site in APP, and resulted in an approximately 40% reduction in the formation of amyloidogenic peptides in vitro. The strong protective effect of the A673T substitution against Alzheimer disease provided proof of principle for the hypothesis that reducing the beta-cleavage of APP may protect against the disease. Furthermore, as the A673T allele also protects against cognitive decline in the elderly without Alzheimer disease, Jonsson et al. (2012) hypothesized that the 2 may be mediated through the same or similar mechanisms.

Modifier Genes

It is clear that apoE plays an important role in the genetics of late-onset Alzheimer disease (see AD2; 104310); however, estimates of the total contribution of apoE to the variance in onset of AD vary widely. In an oligogenic segregation analysis of 75 families ascertained through members with late-onset AD, Daw et al. (2000) estimated the number of additional quantitative trait loci (QTLs) and their contribution to the variance in age at onset of AD, as well as the contribution of apoE and sex. They found evidence that 4 additional loci make a contribution to the variance in age at onset of late-onset AD similar to or greater in magnitude than that made by apoE, with 1 locus making a contribution several times greater than that of apoE. They confirmed the previous findings of a dosage effect for the apoE epsilon-4 allele, a protective effect for the epsilon-2 allele, evidence for allelic interactions at the apoE locus, and a small protective effect for males. Although Daw et al. (2000) estimated that the apoE genotype can make a difference of as many as 17 years in age at onset of AD, their estimate of the contribution of apoE (7 to 9%) to total variance in onset of AD was somewhat smaller than that previously reported. Their results suggested that several genes not yet localized to that time may play a larger role than does apoE in late-onset AD.

Li et al. (2002) performed a genome screen to identify genes influencing age at onset in 449 families with Alzheimer disease and 174 families with Parkinson disease. Heritabilities between 40% and 60% for age at onset were found in both the AD and the PD data sets. For PD, significant evidence for linkage to age at onset was found on 1p (lod = 3.41); see 606852. For AD, the age at onset effect of APOE (lod = 3.28) was confirmed. In addition, evidence for age at onset linkage on chromosomes 6 and 10 was identified independently in both the AD and PD data sets. Subsequent unified analyses of these regions identified a single peak on 10q between D10S1239 and D10S1237, with a maximum lod score of 2.62. These data suggested that a common gene affects age at onset in these 2 common complex neurodegenerative diseases.

Li et al. (2003) combined gene expression studies on hippocampus obtained from AD patients and controls with their previously reported linkage data to identify 4 candidate genes on chromosome 10q. Allelic association studies for age-at-onset effects in 1,773 AD patients and 1,041 relatives and 635 PD patients and 727 relatives further limited association to GSTO1 (605482) (p = 0.007) and a second transcribed member of the GST omega class, GSTO2 (612314) (p = 0.005), located next to GSTO1. The authors suggested that GSTO1 may be involved in the posttranslational modification of IL1B (147720).

Zareparsi et al. (2002) noted that several studies had found an increased frequency of the HLA-A2 (142800) allele in patients with early-onset AD and that others had found an association between the A2 allele and an earlier age of onset of AD. Among 458 unrelated patients with AD, Zareparsi et al. (2002) found that HLA-A2 homozygotes had onset of AD 5 years earlier, on average, than either A2 heterozygotes or those without A2, reflecting a gene dosage effect. The risk associated with the A2 homozygous genotype was 2.6 times greater in patients with early-onset AD (less than age 60 years) than in those with late-onset AD. These effects were present regardless of gender, familial or sporadic nature of the disease, or presence or absence of the APOE4 allele. The authors suggested that the A2 allele may have a role in regulating an immune response in the pathogenesis of AD or that there may be a responsible gene in close linkage to A2.

The APBB2 gene (602710) encodes a protein that is capable of binding to APP. In a genetic association study of 3 independently collected case-control series totaling approximately 2,000 samples, Li et al. (2005) found that a SNP in the APBB2 gene, located in a region conserved between the human and mouse genomes, showed a significant interaction with age of disease onset. For this marker, Li et al. (2005) reported that the association of late-onset Alzheimer disease was most pronounced in subjects with disease onset before 75 years of age; odds ratio for homozygotes = 2.43 and for heterozygotes = 2.15.

Go et al. (2005) performed linkage analysis on an NIMH Alzheimer disease sample and demonstrated a specific linkage peak for AD with psychosis on chromosome 8p12, which encompasses the NRG1 gene (142445). The authors also demonstrated a significant association between an NRG1 SNP (rs3924999) and AD with psychosis (chi-square = 7.0; P = 0.008). This SNP is part of a 3-SNP haplotype preferentially transmitted to individuals with the phenotype. Go et al. (2005) suggested that NRG1 plays a role in increasing the genetic risk for positive symptoms of psychosis in a proportion of late-onset AD families.

Sweet et al. (2005) conducted a study to determine if genetic variation in the COMT gene (116790) was associated with a risk of psychosis in Alzheimer disease. The study included a case-control sample of 373 individuals diagnosed with AD with or without psychosis. Subjects were characterized for alleles at 3 COMT loci previously associated with schizophrenia (rs737865, rs4680, and rs165599), and for a C/T transition adjacent to an estrogen response element (ERE6) in the COMT P2 promoter region. Single-locus and haplotype tests of association were conducted. Logit models were used to examine independent and interacting effects of alleles at the associated loci and all analyses were stratified by sex. In female subjects, rs4680 demonstrated a modest association with AD plus psychosis; rs737865 demonstrated a trend towards an association. There was a highly significant association of AD plus psychosis with a 4-locus haplotype, which resulted from additive effects of alleles at and ERE6/rs737865 (the latter were in linkage disequilibrium). In male subjects, no single-locus test was significant, although a strong association between AD with psychosis and the 4-locus haplotype was observed. That association appeared to result from interaction of the ERE6/rs737865, rs4680, rs165599 loci. Genetic variation in COMT was associated with AD plus psychosis and thus appears to contribute to psychosis risk across disorders.

Associations with Susceptibility to Alzheimer Disease

McIlroy et al. (2000) reported a case-control study of 175 individuals with late-onset Alzheimer disease and 187 age- and sex-matched controls from Northern Ireland. The presence of the butyrylcholinesterase K variant (BCHE; 177400.0005) was found to be associated with an increased risk of Alzheimer disease (odds ratio = 3.50, 95% CI 2.20-6.07). This risk increased in subjects 75 years or older (odds ratio = 5.50, 95% CI 2.56-11.87). No evidence of synergy between BCHE K and APOE epsilon-4 was found in this population.

In a series of 239 necropsy-confirmed late-onset AD cases and 342 elderly nondemented controls older than 73 years, Narain et al. (2000) found an association between homozygosity for both the ACE I and D allele polymorphisms (106180.0001) and AD. Whereas the APOE epsilon-4 allele was strongly associated with AD risk in their series, Narain et al. (2000) found no evidence for an interaction between the APOE and ACE loci. In addition, no interactions were observed between ACE and gender or age at death of the AD cases. A metaanalysis of all published reports (12 case-control series in total) suggested that both the I/I and I/D ACE genotypes are associated with increased AD risk (odds ratio for I/I vs D/D, 1.36, 95% CI = 1.13-1.63; OR for D/I vs D/D, 1.33, 95% CI = 1.14-1.53, p = 0.0002). In a metaanalysis of 23 independent published studies, Elkins et al. (2004) found that the OR for AD in individuals with the I allele (I/I or I/D genotype) was 1.27 compared to those with the D/D genotype. The risk of AD was higher among Asians (OR, 2.44) and in patients younger than 75 years of age (OR, 1.54). Elkins et al. (2004) concluded that the ACE I allele is associated with an increased risk of late-onset AD, but noted that the risk is very small compared to the effects of other alleles, especially APOE4.

Prince et al. (2001) genotyped 204 Swedish patients with sporadic late-onset Alzheimer disease and 186 Swedish control subjects for polymorphisms within 15 candidate genes previously reported to show significant association in Alzheimer disease. The genes chosen for analysis were LRP1, ACE, A2M, BLMH (602403), DLST (126063), TNFRSF6 (134637), NOS3 (163729), PSEN1, PSEN2, BCHE, APBB1 (602709), ESR1 (133430), CTSD (116840), MTHFR (607093), and IL1A (147760). No strong evidence was found for genetic association among the 15 tested variants, and the authors concluded that with the exception of possession of the APOE4 allele, none of the other investigated single-nucleotide polymorphisms contributed substantially to the development of AD in the studied sample.

In 2 groups of patients with AD, comprising a total of 201 patients, Papassotiropoulos et al. (2003) found that the frequency of a 24-cholesterol hydroxylase (CYP46; 604087) T-C polymorphism, CYP46*TT, was associated with increased risk of AD (OR = 2.16). The OR for the APOE4 allele carriers was 4.38. The OR for the presence of both CYP46*TT and APOE4 was 9.63, suggesting a synergistic effect of the 2 genotypes. Neuropathologic examination of AD patients and controls showed that brain beta-amyloid load, CSF levels of soluble beta-amyloid-42, and CSF levels of phosphorylated tau were significantly higher in subjects with the CYP46*TT genotype. Papassotiropoulos et al. (2003) suggested that functional alterations of cholesterol 24-hydroxylase may modulate cholesterol concentrations in vulnerable neurons, thereby affecting changes in amyloid precursor protein processing and beta-amyloid production leading to the development of AD. See also Wolozin (2003).

Because glucocorticoid excess increases neuronal vulnerability, genetic variations in the glucocorticoid system may be related to the risk for AD. De Quervain et al. (2004) analyzed SNPs in 10 glucocorticoid-related genes in 351 AD patients and 463 unrelated control subjects. A rare haplotype in the 5-prime regulatory region of the HSD11B1 gene (600713) was associated with a 6-fold increased risk for sporadic AD. The HSD11B1 enzyme controls tissue levels of biologically active glucocorticoids and thereby may influence neuronal vulnerability. In human embryonic kidney cells, the risk-associated haplotype reduced HSD11B1 transcription by 20% compared to the common haplotype.

Robson et al. (2004) examined the interaction between the C2 variant of the TF gene (190000.0004) and the cys282-to-tyr allele of the HFE gene (C282Y; 613609.0001), the most common basis of hemochromatosis, as risk factors for developing AD. The results showed that each of the 2 variants was associated with an increased risk of AD only in the presence of the other. Neither allele alone had any effect. Carriers of both variants were at 5 times greater risk of AD compared with all others. Furthermore, carriers of these 2 alleles plus APOE4 were at still higher risk of AD: of the 14 carriers of the 3 variants identified in this study, 12 had AD and 2 had mild cognitive impairment. Robson et al. (2004) concluded that the combination of TF*C2 and HFE C282Y may lead to an excess of redoxactive iron and the induction of oxidative stress in neurons, which is exacerbated in carriers of APOE4. They noted that 4% of northern Europeans carry the 2 iron-related variants and that iron overload is a treatable condition.

In a study of 148 patients from southern Italy with sporadic AD, Zappia et al. (2004) found that having a myeloperoxidase (MPO) polymorphism genotype, -463G/G (606989.0008), conferred an odds ratio of 1.65 for development of the disease. When combined with an alpha-2-macroglobulin polymorphism genotype, 1000val/val (103950.0001), the odds ratio increased to 23.19. The authors suggested that the synergistic effect of the 2 genotypes may represent a facilitation of beta-amyloid deposition or a decrease in amyloid clearance, and noted that MPO produces oxidizing conditions. The findings were independent of APOE4 status.

Bian et al. (2005) found no association of 6 A2M gene (103950) polymorphisms with Alzheimer disease in a study of 216 late-onset AD patients and 200 control subjects from the Han Chinese population. Comparison of allele, genotype, and haplotype frequencies for polymorphisms in A2M revealed no significant differences between patients and control subjects.

Mace et al. (2005) found a significant association between a C-T SNP (rs908832) in exon 14 of the ABCA2 gene (600047) and Alzheimer disease in a large case-control study involving 440 AD patients. Additional analysis showed the strongest association between the SNP and early-onset AD (odds ratio of 3.82 for disease development in carriers of the T allele compared to controls).

In a survey of 138 published studies on genetic association for AD, Blomqvist et al. (2006) found evidence for publication bias for positive associations. The authors analyzed 62 genetic markers for AD risk in 940 Scottish and Swedish individuals with AD and 405 Scottish and Swedish controls and found no significant associations except for APOE. In particular, no association was found with variants in the PLAU gene (191840).

Kamboh et al. (2006) studied the association of polymorphisms in the UBQLN1 gene (605046) on chromosome 9q21 with AD. They examined the association of 3 SNPs in the gene (intron 6 A/C, intron 8 T/C, and intron 9 A/G), all of which are in significant linkage disequilibrium (p less than 0.0001), in up to 978 late-onset Alzheimer disease patients and 808 controls. Modestly significant associations were observed in the single-site regression analysis, but 3-site haplotype analysis revealed significant associations (p less than 0.0001). One common haplotype, called H4, was associated with AD risk, whereas a less common haplotype, called H5, was associated with protection, Kamboh et al. (2006) suggested that genetic variation in the UBQLN1 gene has a modest effect on risk, age at onset, and disease duration of Alzheimer disease and that the presence of additional putative functional variants either in UBQLN1 or nearby genes exist.

In a study of 265 AD patients and 347 controls, Ramos et al. (2006) reported a possible protective effect against AD development associated with a polymorphism in the TNF gene (-863C-A; 191160.0006). The -863A allele was present in 16.9% of controls and 12.6% of patients. Comparison of the 3 genotypes (C/C, C/A, and A/A) suggested a dose-response effect with the A/A genotype conferring an odds ratio of 0.58. The findings supported a role for inflammation in AD.

Reiman et al. (2007) used a genomewide SNP survey to examine 1,411 individuals with late-onset AD and controls, including 644 carriers of the APOE4 allele and 767 noncarriers. The authors found a significant association between AD and 6 SNPs in the GAB2 gene (606203) that are part of a common haplotype block. Maximal significance of the association was at rs2373115 with an odds ratio of 4.06 (uncorrected p value of 9 x 10(-11)). Carriers of the APOE4 alleles had an even higher disease risk when the SNP risk allele was present (odds ratio of 24.64) compared to noncarriers. Neuropathologic studies found that GAB2 was overexpressed in neurons from AD patients and the protein was detected in neurons, tangle-bearing neurons, and dystrophic neurites. In contrast, both Chapuis et al. (2008) and Miyashita et al. (2009) failed to detect an association between the GAB2 SNP rs2373115 and risk of developing AD in Caucasian and Japanese individuals, respectively. Chapuis et al. (2008) studied 3 European Caucasian populations totaling 1,749 AD cases and 1,406 controls, and Miyashita et al. (2009) studied 1,656 Japanese cases and 1,656 Japanese controls; they suggested that GAB2 is, at best, a minor disease susceptibility gene for AD.

See GSK3B (605004) for a discussion of a possible association between risk of AD and epistatic interaction between variants in the GSK3B and MAPT genes (157140).

Lambert et al. (2013) conducted a large, 2-stage metaanalysis of genomewide association studies in individuals of European ancestry for risk of late-onset Alzheimer disease. In stage 1, Lambert et al. (2013) used genotyped and imputed data (7 million SNPs) to perform metaanalysis on 4 previously published genomewide association studies datasets containing 17,008 Alzheimer disease cases and 37,154 controls. In stage 2, Lambert et al. (2013) genotyped 11,632 SNPs and tested them for association in an independent set of 8,572 Alzheimer disease cases and 11,312 controls. In addition to the APOE locus, 19 loci reached genomewide significance (p less than 5 x 10(-8)) in the combined stage 1 and stage 2 analyses, of which 11 are newly associated with Alzheimer disease.

Population Genetics

In a population-based study in the city of Rouen, France (426,710 residents), Campion et al. (1999) estimated the prevalence of early-onset AD and autosomal dominant early-onset AD to be 41.2 and 5.3 per 100,000 persons, respectively. Early-onset AD was defined as onset of disease at age less than 61 years, and autosomal dominant early-onset AD was defined as the occurrence of at least 3 cases in 3 generations. They identified PSEN1 gene mutations in 19 (56%) of 34 families, and APP gene mutations in 5 (15%) families. In the 10 remaining families and in 9 additional autosomal dominant AD families, no PSEN1, PSEN2, or APP mutations were found. These results showed that PSEN1 and APP mutations account for 71% of autosomal dominant early-onset AD, and that nonpenetrance at age less than 61 years is probably infrequent for PSEN1 or APP mutations.

Finckh et al. (2000) investigated the proportion of early-onset dementia attributable to known genes. They screened for mutations in 4 genes, PSEN1, PSEN2, APP, and the prion protein gene PRNP (176640), in patients with early-onset dementia before age 60 years. In 16 patients the family history was positive for dementia, in 17 patients it was negative, and in 3 patients it was unknown. In 12 patients, they found 5 novel mutations and 5 previously reported mutations that were all considered to be disease-causing. Nine of these 12 patients had a positive family history, indicating a detection rate of 56% (9/16) in patients with a positive family history.

Animal Model

For a detailed discussion of animal models of Alzheimer disease, see 104760.

McGowan et al. (2006) provided a detailed review of mouse models of Alzheimer disease.

Cheng et al. (1988) described the comparative mapping of DNA markers in the region of familial Alzheimer disease on human chromosome 21 and mouse chromosome 16. The linkage group shared by mouse chromosome 16 and human chromosome 21 included both APP and markers linked to familial Alzheimer disease. The linkage group of 6 loci extends from anonymous DNA marker D21S52 to ETS2, and spans 39% recombination in man but only 6.4% recombination in the mouse. A break in synteny occurs distal to ETS2, and the homolog of human marker D21S56 maps to mouse chromosome 17.

Alzheimer disease has a substantial inflammatory component, and activated microglia may play a central role in neuronal degeneration. Tan et al. (1999) demonstrated that the CD40 (109535) expression was increased on cultured microglia treated with freshly solubilized amyloid-beta and on microglia from a transgenic murine model of Alzheimer disease (Tg APPsw). Increased TNF-alpha (191160) production and induction of neuronal injury occurred when amyloid-beta-stimulated microglia were treated with CD40 ligand (300386). Microglia from Tg APPsw mice deficient for CD40 ligand had less activation, suggesting that the CD40-CD40 ligand interaction is necessary for amyloid-beta-induced microglial activation. In addition, abnormal tau phosphorylation was reduced in Tg APPsw animals deficient for CD40 ligand, suggesting that the CD40-CD40 ligand interaction is an early event in Alzheimer disease pathogenesis.

Phosphorylation of tau and other proteins on serine or threonine residues preceding a proline seems to precede formation of neurofibrillary tangles and neurodegeneration in AD. These phospho(ser/thr)-pro motifs exist in 2 distinct conformations, whose conversion in some proteins is catalyzed by the Pin1 prolyl isomerase (PIN1; 601052). Pin1 activity can directly restore the conformation and function of phosphorylated tau or it can do so indirectly by promoting its dephosphorylation. Liou et al. (2003) found that mice with targeted deletion of the Pin1 gene developed several age-dependent phenotypes including retinal atrophy. In addition, Pin1-null mice showed progressive age-dependent motor and behavioral deficits which included abnormal limb clasping reflexes, hunched postures, and reduced mobility in eye irritation. Neuropathologic changes included tau hyperphosphorylation, tau filament formation, and neuronal degeneration in brain and spinal cord.

Lesne et al. (2006) found that memory deficits in middle-aged Tg2576 mice are caused by the extracellular accumulation of a 56-kD soluble amyloid-beta assembly, which they termed A-beta-*56. A-beta-*56 purified from the brains of impaired Tg2576 mice disrupted memory when administered to young rats. Lesne et al. (2006) proposed that A-beta-*56 impairs memory independently of plaques or neuronal loss, and may contribute to cognitive deficits associated with Alzheimer disease.

The neurodegeneration observed in Alzheimer disease has been associated with synaptic dismantling and progressive decrease in neuronal activity. Busche et al. (2008) tested this hypothesis in vivo by using 2-photon calcium ion imaging in a mouse model of Alzheimer disease. The mouse model consists of double transgenic mice overexpressing both beta-amyloid precursor protein (APP; 104760) and mutant presenilin-1 (104311). Although a decrease in neuronal activity was seen in 29% of layer 2/3 cortical neurons, 21% of neurons displayed an unexpected increase in the frequency of spontaneous calcium ion transients. These 'hyperactive' neurons were found exclusively near the plaques of amyloid beta-depositing mice. The hyperactivity appeared to be due to a relative decrease in synaptic inhibition. Thus, Busche et al. (2008) suggested that a redistribution of synaptic drive between silent and hyperactive neurons, rather than an overall decrease in synaptic activity, provides a mechanism for the disturbed cortical function in Alzheimer disease.

Nagahara et al. (2009) reported beneficial effects of entorhinal administration of brain-derived neurotrophic factor (BDNF; 113505) in 3 models of AD-related cognitive decline in mouse and nonhuman primates: an App-mutant mouse strain, aged rats, and aged monkeys. BDNF is widely expressed in the entorhinal cortex and undergoes anterograde transport into the hippocampus, where it is implicated in plasticity mechanisms. In App-transgenic mice, lentiviral BDNF gene delivery administered after disease onset reversed synapse loss, partially normalized aberrant gene expression, improved cell signaling, and restored learning and memory. These changes occurred independently of amyloid plaque load. In aged rats, BDNF protein and lentiviral gene infusion, respectively, reversed cognitive decline and improved age-related perturbations in gene expression. In adult rats and primates, lentiviral BDNF gene delivery prevented lesion-induced death of entorhinal cortical neurons. Finally, lentiviral BDNF gene delivery and expression in aged primates reversed neuronal atrophy and ameliorated age-related cognitive impairment. Nagahara et al. (2009) suggested that BDNF exerts substantial protective effects on crucial neuronal circuitry involved in AD, acting through amyloid-independent mechanisms.

Treusch et al. (2011) modeled amyloid-beta toxicity in yeast by directing the peptide to the secretory pathway. A genomewide screen for toxicity modifiers identified the yeast homolog of phosphatidylinositol-binding clathrin assembly protein (PICALM; 603025) and other endocytic factors connected to Alzheimer disease whose relationship to amyloid-beta had been unknown. The factors identified in yeast modified amyloid-beta toxicity in glutamatergic neurons of C. elegans and in primary rat cortical neurons. In yeast, amyloid-beta impaired the endocytic trafficking of a plasma membrane receptor, which was ameliorated by endocytic pathway factors identified in the yeast screen. Treusch et al. (2011) concluded that links between amyloid-beta, endocytosis, and human Alzheimer disease risk factors can be ascertained with yeast as a model system.

By screening a library of about 80,000 chemical compounds, Kounnas et al. (2010) identified a class of gamma-secretase modulators (GSMs), diarylaminothiazoles, or series A GSMs, that could target production of A-beta-42 and A-beta-40 in cell lines and in Tg 2576 transgenic AD mice. Immobilized series A GSMs bound to Pen2 (PSENEN; 607632) and, to a lesser degree, Ps1. Series A GSMs reduced gamma-secretase activity without interfering with related off-target reactions, lowered A-beta-42 levels in both plasma and brain of Tg 2576 mice, and reduced plaque density and amyloid in Tg 2576 hippocampus and cortex. Daily dosing was well tolerated over the 7-month study.

Metabolites in the kynurenine pathway of tryptophan degradation in mammals are thought to play an important role in neurodegenerative disorders, including Alzheimer disease. Kynurenic acid (KYNA) had been shown to reduce neuronal vulnerability in animal models by inhibiting ionotropic excitatory amino acid receptors, and is neuroprotective in animal models of brain ischemia. Zwilling et al. (2011) synthesized a small-molecule prodrug inhibitor of kynurenine 3-monooxygenase (KMO; 603538), termed JM6, and found that oral administration of JM6 to rats increased KYNA levels and reduced extracellular glutamate in the brain. In a transgenic mouse model of Alzheimer disease, JM6 prevented spatial memory deficits, anxiety-related behavior, and synaptic loss. These findings supported a critical link between tryptophan metabolism in the blood and neurodegeneration.

Cramer et al. (2012) found that oral administration of the RXR (see 180245) agonist bexarotene to a mouse model of Alzheimer disease resulted in enhanced clearance of soluble amyloid-beta within hours in an ApoE-dependent manner. Amyloid-beta plaque area was reduced more than 50% within just 72 hours. Furthermore, bexarotene stimulated the rapid reversal of cognitive, social, and olfactory deficits and improved neural circuit function. Thus, Cramer et al. (2012) concluded that RXR activation stimulates physiologic amyloid-beta clearance mechanisms, resulting in the rapid reversal of a broad range of amyloid-beta-induced deficits.

Several groups provided technical comments on the report of Cramer et al. (2012). While Fitz et al. (2013) confirmed that administration of bexarotene reversed memory deficits in APP/PS1-delta-E9 mice expressing human APOE3 or APOE4 to the levels of their nontransgenic controls and significantly decreased interstitial fluid amyloid-beta, they could not confirm the effects on amyloid deposition. Using a nearly identical treatment regimen, Price et al. (2013) were unable to detect any evidence of drug efficacy despite evidence of target engagement. Tesseur et al. (2013) were not able to reproduce the described effects in several animal models. They remarked that drug formulation appeared to be very critical and that their data called for 'extreme caution' when considering this compound for use in AD patients. Veeraraghavalu et al. (2013) found that although bexarotene reduced soluble beta-amyloid-40 levels in 1 of the mouse models, the drug had no impact on plaque burden in 3 strains that exhibit amyloid beta amyloidosis. Landreth et al. (2013) replied that the data of Fitz et al. (2013), Price et al. (2013), Tesseur et al. (2013), and Veeraraghavalu et al. (2013) replicated and validated their central conclusion that bexarotene stimulates the clearance of soluble beta-amyloid peptides and results in the reversal of behavioral deficits in mouse models of AD. They considered the basis of the inability to reproduce the drug-stimulated microglial-mediated reduction in plaque burden to be unexplained. However, they concluded that plaque burden is functionally unrelated to improved cognition and memory elicited by bexarotene.

Ahn et al. (2014) noted that fibrinogen (see 134820) is a cerebrovascular risk factor in AD that specifically binds beta-amyloid, thereby altering fibrin clot structure and delaying clot degradation. Using a high-throughput screen, they identified RU-505 as an inhibitor of the interaction between beta-amyloid and fibrinogen. RU-505 restored beta-amyloid-induced altered fibrin clot formation and degradation in vitro and inhibited vessel occlusion in AD transgenic mice. Long-term treatment with RU-505 significantly reduced vascular amyloid deposition and microgliosis in cortex and improved cognitive impairment in mouse models of AD. Ahn et al. (2014) proposed that inhibitors of the interaction between beta-amyloid and fibrinogen may be useful in AD therapy.

Using mouse models, Hong et al. (2016) showed that complement and microglia mediate synaptic loss early in AD. C1q (see 120550), the initiating protein of the classical complement cascade, was increased and associated with synapses before overt plaque deposition. Inhibition of C1q, C3 (120700), or the microglial complement receptor CR3 (CD11b/CD18; see 600065) reduced the number of phagocytic microglia, as well as the extent of early synapse loss. C1q was necessary for the toxic effects of soluble beta-amyloid (A-beta) oligomers on synapses and hippocampal long-term potentiation. Finally, microglia in adult brains engulfed synaptic material in a CR3-dependent process when exposed to soluble A-beta oligomers. Together, these findings suggested that the complement-dependent pathway and microglia that prune excess synapses in development are inappropriately activated and mediate synapse loss in AD.

BECN1 (604378) is an essential autophagy protein. Rocchi et al. (2017) found that mice with knockin of a Becn1 gene containing a phe121-to-ala (F121A) mutation had significantly reduced interaction of Becn1 with its inhibitor, Bcl2 (151430), leading to constitutive autophagy in multiple tissues, including brain. The Becn1 F121A-mediated hyperactivation of autophagy significantly decreased amyloid accumulation, prevented cognitive decline, and restored survival in AD mouse models. The authors found that amyloid-beta oligomers were autophagic substrates sequestered in autophagosomes in brains of autophagy-hyperactive AD mice. Chemical inducers and exercise induced autophagy through Becn1-dependent protective effects on amyloid-beta removal and memory in AD mice. Rocchi et al. (2017) concluded that genetic mutations, chemical agents, or exercise can hyperactivate autophagy in vivo by disrupting BECN1-BCL2 binding, sequestering amyloid oligomers and preventing AD progression.

History

Bogerts (1993) provided a biographical sketch and photograph of Alois Alzheimer (1864-1915). Alzheimer was a neuropathologist, clinical psychiatrist, and chairman of psychiatry. He always considered himself a psychiatrist. He worked with Nissl in the application of the Nissl staining techniques for the study of the cerebral cortex in psychosis. Alzheimer discovered the disorder that bears his name when he reported on 'a strange disease of the cerebral cortex' in a 51-year-old woman (Auguste D.) with presenile dementia who displayed diffuse cortical atrophy, nerve cell loss, plaques, and tangles (Alzheimer, 1907). He was then working in Munich in the department of Emil Kraepelin, director of the Munich psychiatric clinic, who coined the term 'Alzheimer's disease.'

O'Brien (1996) reported that the file on the case of Auguste D., who at the age of 51 came under the care of Alois Alzheimer, had come to light; it had been missing since 1910. Auguste D. came under the care of Alzheimer at a Frankfurt hospital in 1901. On the basis of the record, some questions of whether Auguste D. had the disorder now called Alzheimer disease were raised; namely, that autopsy findings included arteriosclerosis noted in the smaller cerebral blood vessels. O'Brien (1996) noted that today this is a criterion for exclusion from a diagnosis of AD.

Maurer et al. (1997) announced that the long-sought clinical record of Auguste D. was discovered in Frankfurt only 2 days after the eightieth anniversary of the death of Professor Alzheimer, who died December 19, 1915. A photograph of the patient, dated November 1902, was provided by Maurer et al. (1997), as well as a copy of her handwriting which led Alzheimer to refer to the condition as 'amnestic writing disorder.'

Graeber et al. (1997) did a retrospective analysis on the case of Johann F., the second patient reported by Alois Alzheimer (1911). Johann F. was a 56-year-old male who suffered from presenile dementia and was hospitalized in Kraepelin's clinic for more than 3 years. Postmortem examination of the patient's brain revealed numerous amyloid plaques but no neurofibrillary tangles in the cerebral cortex, corresponding to a less common form of Alzheimer disease which may be referred to as 'plaque only.' Graeber et al. (1997) recovered well-preserved histologic sections of this case and performed mutation screening of exon 17 of the APP gene and genotyping for APOE alleles. The patient was shown to be homozygous for APOE3 and lacked APP mutations at codons 692, 693, 713, and 717. The investigators speculated that the patient may have had mutations in the PS1 or PS2 gene.

Graeber et al. (1998) described the histopathology and APOE genotype of Alois Alzheimer's first patient, Auguste D. As in the case of Johann F., a large number of tissue sections belonging to Alzheimer's laboratory, which was later headed by Spielmeyer (Spielmeyer, 1916), were later found among material kept at the Institute of Neuropathology of the University of Munich. As described by Alzheimer (1907) in his original report, there were numerous neurofibrillary tangles and many amyloid plaques, especially in the upper cortical layers of this patient. However, there was no microscopic evidence for vascular, i.e., arteriosclerotic, lesions. The histologic preparations did not include the hippocampus or entorhinal region. The APOE genotype of this patient was shown to be E3/E3 by PCR-based restriction enzyme analysis.

Yu et al. (2010) demonstrated that a family from Fulda (Hesse), Germany with Alzheimer disease-4 (AD4; 606889) caused by the N141I mutation in the PSEN2 gene (600759.0001) shared the same haplotype as affected Volga German families reported earlier. This finding indicated that the N141I mutation must have occurred prior to the emigration of the Volga Germans from the Hesse region of Germany to Russia in the 1760s during the reign of Catherine the Great. In addition, the original patient with AD reported by Alzheimer (1907) also lived in same Hesse region as the modern family, which raised the possibility that the original patient may have had the N141I mutation.