Familial Alzheimer's disease (FAD) refers to a subset of Alzheimer's disease (AD) cases characterized by a clear autosomal dominant inheritance pattern, typically with onset before 65 years of age. While sporadic AD accounts for the vast majority of cases, representing greater than 95% of all diagnoses, familial forms account for less than 5% of total AD prevalence. However, FAD provides critical insights into the underlying pathophysiology of Alzheimer's disease through the identification of causative genetic mutations that disrupt fundamental biological pathways involved in amyloid-beta (Aβ) production and processing. [1]
The inheritance pattern of FAD follows an autosomal dominant model, meaning that a single copy of the mutated gene from either parent is sufficient to cause the disease. Offspring of an affected individual have a 50% chance of inheriting the mutation. Three major genes have been definitively implicated in FAD: the amyloid precursor protein gene (APP) on chromosome 21, and the presenilin genes PSEN1 (presenilin 1) on chromosome 14 and PSEN2 (presenilin 2) on chromosome 1. Mutations in these genes lead to increased production or aggregation of amyloid-beta peptides, particularly the more neurotoxic Aβ42 isoform, initiating a cascade of events that ultimately result in progressive neurodegeneration and cognitive decline. [2]
The identification of these causal genes has been instrumental in establishing the amyloid cascade hypothesis, which posits that the accumulation of Aβ peptides in the brain represents the primary driving force behind disease development. While the exact mechanisms remain incompletely understood, genetic discoveries in FAD have provided foundational knowledge that informs current therapeutic development strategies and our understanding of both familial and sporadic AD. [3]
The APP gene encodes the amyloid precursor protein, a transmembrane protein expressed ubiquitously in tissues throughout the body, with particularly high expression in the brain. APP undergoes proteolytic processing through two primary pathways: the non-amyloidogenic pathway, in which alpha-secretase cleaves APP within the Aβ sequence, preventing Aβ formation, and the amyloidogenic pathway, in which beta-secretase (BACE1) and gamma-secretase sequentially cleave APP to generate Aβ peptides of varying lengths. [4]
Over 50 pathogenic mutations have been identified in the APP gene, most of which are located within or near the Aβ coding region. The Swedish mutation (KM670/671NL), the first FAD mutation identified, involves a double amino acid substitution at the beta-secretase cleavage site, resulting in significantly enhanced beta-secretase processing and dramatically increased production of all Aβ isoforms. This mutation, originally described in a large Swedish family, leads to approximately 3-8 fold increase in total Aβ production. [5]
The London mutation (V717I) substitutes isoleucine for valine at position 717, immediately adjacent to the gamma-secretase cleavage site. This mutation specifically increases the ratio of Aβ42 to Aβ40, promoting amyloid fibril formation and plaque deposition. The V717I mutation was the first mutation shown to alter the processing of APP in a manner that preferentially increases the more aggregatable Aβ42 species, supporting the hypothesis that changes in the Aβ42/Aβ40 ratio may be particularly important in disease pathogenesis. [6]
Notably, a protective mutation (A673T) has been identified in the APP gene, specifically at the beta-secretase cleavage site. This substitution reduces beta-secretase cleavage of APP by approximately 40%, resulting in decreased Aβ production. Individuals carrying this mutation demonstrate significantly reduced risk of developing AD and cognitive decline in normal aging, providing genetic evidence supporting the amyloid hypothesis and indicating that reducing Aβ production may be a viable therapeutic strategy. [7]
Other APP mutations include the Flemish mutation (A692G), Arctic mutation (E693G), and various mutations affecting the APP promoter region. The APP gene also exhibits copy number variations, with trisomy 21 (Down syndrome) leading to increased APP expression and universal development of AD neuropathology by age 60-70, further supporting the central role of APP and Aβ in disease pathogenesis. [8]
The PSEN1 gene encodes presenilin 1, the catalytic subunit of the gamma-secretase complex, which performs the final proteolytic cleavage of APP to generate Aβ peptides. PSEN1 is essential for gamma-secretase activity, and knockout of the PSEN1 gene in mice results in complete loss of Aβ production. The presenilin 1 protein forms a heterodimeric complex with three other cofactors (nicastrin, APH-1, and PEN-2) to create a functional aspartyl protease that cleaves multiple substrate proteins, including Notch. [9]
PSEN1 mutations represent the most common cause of familial Alzheimer's disease, accounting for approximately 70-80% of all FAD cases with identified mutations. Over 200 pathogenic PSEN1 mutations have been described to date, spanning the entire gene and involving various missense, nonsense, and splice site alterations. The majority of PSEN1 mutations are missense mutations affecting residues within the transmembrane domains that comprise the catalytic core of the gamma-secretase complex. [10]
Families with PSEN1 mutations typically demonstrate earlier age of onset compared to other FAD genotypes, with mean onset occurring between 30 and 50 years of age, though considerable variation exists. Some mutations, such as the A246E mutation, are associated with particularly early onset, with affected individuals developing symptoms in their mid-20s. The range of age of onset for a given PSEN1 mutation can span 10-20 years, suggesting that modifier genes, environmental factors, or stochastic processes influence disease expression even in the presence of a fully penetrant mutation. [11]
PSEN1 mutations demonstrate significant phenotype variability, even within the same mutation or family. While the core clinical features remain those of typical AD, including progressive episodic memory loss, visuospatial dysfunction, and executive impairment, some PSEN1 mutations are associated with atypical presentations. These may include spastic paraparesis, cerebellar ataxia, seizures, or prominent language difficulties. Certain mutations, such as those at position 280 or 291, appear to be specifically associated with the spastic paraparesis phenotype, potentially reflecting differential effects on gamma-secretase substrate processing beyond APP. [12]
Neuropathological examination in PSEN1 cases typically reveals abundant amyloid plaques and neurofibrillary tangles, consistent with AD pathology. However, some mutations may be associated with distinctive features such as "cotton wool" plaques, which are large, rounded plaques without a dense core, or unusual distribution of neurofibrillary pathology. [13]
The PSEN2 gene encodes presenilin 2, a paralog of PSEN1 that can also function as the catalytic subunit of gamma-secretase. Presenilin 2 shares significant structural and functional homology with presenilin 1, though its expression pattern differs, with higher levels in peripheral tissues and a more restricted distribution in the brain. Like PSEN1, PSEN2 mutations cause FAD through alterations in gamma-secretase activity and resultant changes in Aβ metabolism. [14]
The first PSEN2 mutations were identified in the Volga German kindred, a large family cluster of Alzheimer's disease originating from German immigrants who settled in the Volga River region of Russia. This family demonstrated an unusual pattern of disease inheritance, with variable age of onset and incomplete penetrance, features that distinguished PSEN2-linked FAD from PSEN1-linked disease. Several PSEN2 mutations have since been identified in families worldwide, though PSEN2 mutations are considerably less common than PSEN1 mutations, accounting for fewer than 5% of FAD cases. [15]
A hallmark feature of PSEN2 mutations is variable penetrance, meaning that not all individuals who inherit the mutation develop clinically evident Alzheimer's disease during their lifetime. Age of onset in PSEN2 mutation carriers typically ranges from 40 to 85 years, with some mutation carriers remaining cognitively normal into extreme old age. The N141I mutation, originally described in the Volga German kindred, is associated with mean age of onset of approximately 55 years, though range extends from 40 to 70 years. [16]
This incomplete penetrance suggests that PSEN2 mutations may have a weaker pathogenic effect than PSEN1 mutations, or that protective factors may modify disease expression. The mechanism underlying the variable penetrance remains incompletely understood but may involve genetic modifiers, environmental factors, or differences in the efficiency of alternate splicing that generates catalytically active presenilin 2 isoforms. [17]
Clinical and neuropathological features of PSEN2-associated FAD generally resemble those of other FAD forms, though some studies suggest that disease progression may be somewhat slower and that spastic paraparesis may be less common than in PSEN1 cases. [18]
While mutations in APP, PSEN1, and PSEN2 cause Mendelian forms of familial AD, several other genetic variants have been identified that modify risk for both familial and sporadic AD. These risk genes typically have modest effect sizes and contribute to disease risk in an additive, polygenic manner rather than causing deterministic disease. [19]
APOE (apolipoprotein E, also known as ApoE protein represents the strongest known genetic risk factor for AD. The APOE gene encodes a lipid transport protein expressed in astrocytes and microglia in the brain. Three common alleles exist: APOE ε2, ε3, and ε4. The APOE ε4 allele is associated with increased AD risk and earlier age of onset in a dose-dependent manner, with heterozygotes having approximately 3-4 fold increased risk and homozygotes having approximately 10-15 fold increased risk compared to non-carriers. Conversely, the APOE ε2 allele appears to be protective. APOE influences AD risk through multiple mechanisms, including effects on Aβ aggregation, clearance, neuroinflammation, and lipid metabolism. However, APOE is a risk modifier rather than a causal gene, as many APOE ε4 carriers never develop AD, and many AD cases occur in APOE ε4 non-carriers. [20]
TREM2 (triggering receptor expressed on myeloid cells 2) encodes a receptor expressed on microglia in the brain. Rare variants in TREM2 were initially identified as causing a rare syndrome resembling FAD, and subsequent studies identified common variants that increase AD risk by approximately 2-3 fold. TREM2 plays critical roles in microglial phagocytosis and inflammatory responses to Aβ deposition, suggesting that microglial dysfunction may contribute to AD pathogenesis.
ABCA7 (ATP-binding cassette transporter A7) encodes a lipid transporter expressed in microglia and neurons. Common variants in ABCA7 are associated with modestly increased AD risk. ABCA7 appears to influence Aβ processing and clearance, though the exact mechanisms remain under investigation.
CLU (clusterin, also known as APOJ) encodes a secreted chaperone protein that binds Aβ and may facilitate its clearance from the brain. Common CLU variants are associated with approximately 10-15% increased AD risk. Clusterin is highly expressed in AD brain and is found in association with amyloid plaques, suggesting it may play a role in Aβ aggregation or clearance.
Additional AD risk genes include PICALM, BIN1, CR1, and MS4A cluster genes, among others. Together, these genetic discoveries indicate that multiple biological pathways, including amyloid processing, lipid metabolism, immune function, and endosomal trafficking, contribute to AD pathogenesis.
The study of genotype-phenotype correlations in FAD provides valuable insights into how different genetic alterations modify disease expression. These correlations inform our understanding of disease mechanisms and may eventually guide clinical management and therapeutic development.
Age of onset demonstrates significant genetic modification by the causal gene and specific mutation. PSEN1 mutations generally cause the earliest onset, with mean onset in the mid-40s, followed by APP mutations (mean onset in the mid-50s), and PSEN2 mutations (mean onset in the late 50s to early 60s). Within each gene, specific mutations are associated with characteristic age ranges. For example, PSEN1 A246E mutation leads to onset typically before age 35, while PSEN1 M146L mutation often presents after age 50. The mechanisms underlying these differences likely relate to the magnitude of effects on Aβ production or the Aβ42/Aβ40 ratio.
Clinical features may vary by genotype in ways that provide diagnostic clues. While all FAD cases demonstrate progressive cognitive decline characteristic of AD, certain mutations are associated with distinctive phenotypic features. PSEN1 mutations more commonly present with atypical features including spastic paraparesis, seizures, or prominent language impairment. PSEN2 mutations may present with a more amnestic phenotype and slower progression. APP mutations at the beta-secretase cleavage site (Swedish mutation) typically cause a classical AD phenotype, while other APP mutations may present with atypical features.
Neuropathology also demonstrates genotype-specific patterns. PSEN1 mutations often demonstrate abundant diffuse and neuritic plaques with variable degrees of cerebral amyloid angiopathy. Some PSEN1 mutations, such as PSEN1 deltaE9 (exon 9 deletion), are associated with unusual "cotton wool" plaques and prominent hippocampal pathology. PSEN2 mutations may demonstrate less severe amyloid pathology relative to the degree of cognitive impairment, suggesting that other pathological processes may contribute to disease expression.
The presence of the APOE ε4 allele modifies phenotype in FAD cases, typically resulting in earlier age of onset, particularly in heterozygous mutation carriers. APOE ε4 may also influence the clinical presentation, with some studies suggesting increased prevalence of seizures or more prominent hippocampal atrophy.
Genetic testing for FAD genes is clinically available and may be indicated in specific clinical scenarios. Testing requires careful pre-test genetic counseling to ensure informed decision-making and post-test interpretation and support.
Indications for genetic testing include early-onset Alzheimer's disease (typically defined as onset before 65 years), especially with a family history consistent with autosomal dominant inheritance, or when a pathogenic mutation has been identified in a family member. Testing may also be considered in cases of early-onset AD without a clear family history, as de novo mutations can occur. The appropriateness of testing should be assessed on a case-by-case basis in consultation with a genetic counselor or geneticist experienced in neurodegenerative diseases.
Genetic counseling is essential before and after testing. Pre-test counseling should address the inheritance pattern and implications of positive results, the possibility of uncertain or variants of uncertain significance, the limitations of predictive testing in asymptomatic individuals, and the potential psychological, insurance, and employment implications of genetic information. Written informed consent is typically required before testing.
Testing may include analysis of the three FAD genes (APP, PSEN1, PSEN2) or may be more targeted based on clinical presentation and family history. Next-generation sequencing panels that simultaneously analyze multiple AD-associated genes are increasingly common. Variant interpretation should be performed by laboratories with expertise in AD genetics, and variants of uncertain significance should be reported with appropriate caveats.
Predictive testing in asymptomatic at-risk individuals remains controversial and is generally not recommended outside of research protocols. The early-onset, fully penetrant nature of most FAD mutations differs from late-onset AD, where predictive testing has limited utility due to the modest effect of risk alleles. However, some individuals at risk for FAD may request predictive testing to inform life planning, reproductive decisions, or participation in clinical trials. If predictive testing is pursued, it should be conducted only after extensive genetic counseling and psychological assessment.
The identification of FAD genes has fundamentally shaped therapeutic development strategies for Alzheimer's disease. Understanding the genetic basis of FAD has provided molecular targets for drug development and informed therapeutic approaches.
Targeting amyloid production represents the primary therapeutic strategy derived from FAD genetics. Gamma-secretase inhibitors were developed based on the knowledge that PSEN1 and PSEN2 mutations increase Aβ42 production. However, clinical trials of gamma-secretase inhibitors failed due to unacceptable side effects, particularly related to Notch signaling interference, demonstrating the challenge of targeting an enzyme with multiple physiological substrates.
BACE inhibitors target the beta-secretase cleavage of APP and were developed based on knowledge that the Swedish APP mutation increases beta-secretase cleavage. Several BACE inhibitors entered clinical trials, but development was halted due to adverse effects potentially related to BACE's role in normal neuronal function. These failures highlight the complexity of targeting proteases with multiple substrates and physiological functions.
Immunotherapies targeting Aβ, including both active vaccines and monoclonal antibodies, have undergone extensive clinical testing. The recent approval of lecanemab and donanemab, which selectively target Aβ aggregates, represents a milestone in AD therapeutics, though their clinical benefits are modest. These therapies were developed based on the amyloid hypothesis derived from FAD genetics.
Gene-specific approaches represent a promising future direction. Antisense oligonucleotides or RNA interference technologies could theoretically reduce production of mutant APP or presenilin proteins in mutation carriers. Gene editing approaches using CRISPR-Cas9 technology could potentially correct pathogenic mutations, though significant technical challenges remain. These precision medicine approaches would be tailored to specific mutations, potentially offering greater efficacy than broad-spectrum therapies.
Additionally, knowledge of FAD genetics enables identification of at-risk individuals for prevention trials. Clinical trials in presymptomatic mutation carriers, such as the Dominantly Inherited Alzheimer Network (DIAN) trials, test therapies before symptom onset, potentially offering greater opportunity to modify disease course.
Sleegers K, Brouwers N, Gijselinck I, et al. APP duplication is sufficient to cause early onset Alzheimer's dementia with cerebral amyloid angiopathy. Brain. 2006. ↩︎
Cruts M, Van Broeckhoven C. Molecular genetics of Alzheimer's disease. Ann Med. 1998. ↩︎
Lanoiselee HM, Nicolas G, Wallon D, et al. 'APP, PSEN1, and PSEN2 mutations in early-onset Alzheimer disease: A genetic database study'. JAMA Neurol. 2017. ↩︎
Menendez M. The genetics of Alzheimer's disease. Acta Neurol Scand. 1996. ↩︎
Ikeda M, Yoneda M, Shinosaki K, et al. A novel presenilin-2 gene mutation (Phe141Ile) in a Japanese family with early onset Alzheimer's disease. Psychiatry Clin Neurosci. 1999. ↩︎
Bird TD, Lampe TH, Nemens EJ, et al. 'Familial Alzheimer''s disease in American descendants of the Volga Germans: probable genetic founder and estimator of gene frequency'. Ann Neurol. 1988. ↩︎
Marlowe M,dat Galen D, Bee A, et al. Presenilin-2 mutations and atypical dementia syndromes. Neurology. 2003. ↩︎
Strittmatter WJ, Roses AD. Apolipoprotein E and Alzheimer's disease. Annu Rev Neurosci. 1996. ↩︎
Guerreiro R, Wojtas A, Bras J, et al. TREM2 variants in Alzheimer's disease. N Engl J Med. 2013. ↩︎
Steinberg S, Stefansson H, Jonsson T, et al. Loss-of-function variants in ABCA7 confer risk of Alzheimer's disease. Nat Genet. 2015. ↩︎
Lambert JC, Heath S, Even G, et al. Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer's disease. Nat Genet. 2009. ↩︎
Cruchaga C, Chakraverty S, Mayo K, et al. Rare variants in APP, PSEN1 and PSEN2 increase risk for AD in an age-dependent manner. Neurology. 2012. ↩︎
Mann DM, Iwatsubo T, Cairns NJ, et al. 'Amyloid (Abeta) deposition in chromosome 1-linked Alzheimer''s disease: the CERAD experience'. Neurology. 1996. ↩︎
van Duijn CM, Hendriksen L, Hofman A, et al. Phenotype, genotype, and clinical features of familial Alzheimer's disease in a Dutch kindred. Neuroepidemiology. 1998. ↩︎
Goldman JS, Hahn SE, Williamson J, et al. 'Genetic testing for Alzheimer''s disease: practical considerations and recommendations'. Genet Med. 2006. ↩︎
Ringman JM, Younkin SG, Yuede C, et al. Biochemical markers in persons with preclinical familial Alzheimer disease. Neurology. 2008. ↩︎
Bateman RJ, Xiong C, Benzinger TL, et al. Clinical and biomarker changes in dominantly inherited Alzheimer's disease. N Engl J Med. 2012. ↩︎
Salloway S, Farlow M, McDade E, et al. A study of the drug rember in patients with moderate Alzheimer's disease. N Engl J Med. 2014. ↩︎
van Dyck CH, Swanson CJ, Aisen P, et al. Lecanemab in early Alzheimer's disease. N Engl J Med. 2023. ↩︎
Sims JR, Zimmer JA, Evans CD, et al. 'Donanemab in early symptomatic Alzheimer disease: the TRAILBLAZER-ALZ 2 randomized clinical trial'. JAMA. 2023. ↩︎