Publish in



Please download to get full document.

View again

of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  Review Community Genet 2001;4:197–203  Alzheimer’s Disease: Genes,Pathogenesis and Risk Prediction K.SleegersC.M.vanDuijn Department of Epidemiology and Biostatistics, Erasmus Medical Centre Rotterdam, The Netherlands C.M. van Duijn, PhDDepartment of Epidemiology and Biostatistics, Erasmus Medical Centre RotterdamPO BoxNL–1738 Rotterdam (The Netherlands)Tel. +31 10 408 7394, Fax +31 10 408 9406, E-Mail vanduijn@epib.fgg.eur.nl ABC Fax +41 61 306 12 34E-Mail karger@karger.chwww.karger.com© 2002 S. Karger AG, Basel1422–2795/01/0044–0197$17.50/0Accessible online at:www.karger.com/journals/cmg Key Words Alzheimer’s disease W Genetic epidemiology W Pathogenesis W Risk prediction W Genetic counselling W ß -Amyloid  Abstract With the aging of western society the contribution tomorbidity of diseases of the elderly, such as dementia,will increase exponentially. Thorough preventative andcurative strategies are needed to constrain the increas-ing prevalence of these disabling diseases. Better under-standing of the pathogenesis of disease will enabledevelopment of therapy, prevention and the identifica-tion of high-risk groups in the population. Here, wereview the genetic epidemiology of Alzheimer’s disease,the most common cause of dementia in the westernworld. The search for genetic risk factors, though farfrom completed, has been of major importance for un-derstanding the pathogenesis of Alzheimer’s disease.Although effective therapy is still awaited, these findingshave led to new avenues for the development of drugs. Copyright © 2002 S. Karger AG, Basel Introduction Although in the past century major progress has beenmade in unravelling the genetics of Alzheimer’s disease(AD), many questions remain to be answered. The de-bates about its pathogenesis, diagnosis, therapy and pre-vention have not been settled yet. It is clear that AD is acomplex multifactorial disorder. A great number of possi-ble genetic risk factors have been investigated, but formost of these no clear association has been found [1]. Thesearch for genetic risk factors has yielded three genes(amyloid precursor protein [2–5], presenilin-1 [6–10] andpresenilin-2 genes [11, 12]) in which mutations werefound which result in rare autosomal dominant forms of AD. One susceptibility gene (apolipoprotein E gene) hasbeen identified which is a risk factor in the general popu-lation.In this paper the prevalence and risk factors for AD arereviewed. The emphasis will be on the genes known to beinvolved in AD, their role in understanding the develop-ment of the disease, and their implications for diagnosisand clinical counselling. Clinical Epidemiological Aspects of AD Diagnosis and Prognosis AD is the most common cause of dementia in the west-ern world. The disease is clinically characterized by insid-ious onset and slow progression of cognitive decline. Mostfrequently, loss of short-term memory and impaired im-printing of new information are the presenting symptomsof AD. During the course of the disease, symptoms mayfurther include disturbance of speech, poor judgment,personality change and deterioration of visuospatialskills, with preserved level of consciousness. Patientsgradually lose the ability to be self-supportive, and even-  198 Community Genet 2001;4:197–203 Sleegers/vanDuijn tually they will become bedridden. Death usually occursdue to complications of immobility and malnutrition[13]. The average duration of AD is 8–10 years, althoughat old age survival may be shorter. Available therapeuticstrategies (cholinesterase inhibitors) are not curative, butmay halt the process of decline for half a year to 2 years inthe early stages of the disease in a small number of patients [14].The diagnosis is based on clinical examination andneuropsychological testing. These should yield no cluesfor systemic or other brain diseases capable of causingdementia, such as vascular dementia and subcortical de-mentia (NINCDS-ADRDA) [15]. Although the precisionof the diagnosis has improved considerably with improve-ment of neuropsychological tests and neuroimaging, dur-ing life only a probable diagnosis can be made, with anaccuracy of 80–90%. The definite diagnosis of AD isalways based on histopathological findings in the brain[16]: neuritic plaques, neurofibrillary tangles and loss of neurons in hippocampus and cerebral cortex.The neuritic plaques are composed of aggregations of  ß -amyloid, which are surrounded by dystrophic den-drites, microglia and astrocytes. These plaques are locatedpreferentially in limbic and association cortices of thebrain, areas important for memory and cognition.Neurofibrillary tangles consist of intraneuronal aggre-gations of hyperphosphorylated tau. Tau is a protein thatis normally present in adult human brain, where it exertsits function through stabilizing microtubules, which areessential for cell shape and support and intraneuronaltransport. Hyperphosphorylated tau destabilizes the mi-crotubule network within neurons. Due to subsequentneuronal dysfunction and deficits in neurotransmitters,normal brain function is impaired [17].  Prevalence of AD The number of patients affected with AD (prevalenceof disease) is remarkably stable in western society. Themajor determinant of the prevalence of disease is age.Less than 1% of the people aged 70 years or younger isaffected with AD. But with each 5 years’ increase in agethe prevalence of AD doubles, until by age 90 years up to30% is affected [18]. A large European follow-up studyhas shown that, especially at older age, women are moreoften affected than men [19, 20].Although AD is considered to be a disease of the elder-ly, there are patients in whom first symptoms of AD maybe present as early as at age 35 years. Frequently, adistinction is made between ‘early-onset Alzheimer’sdisease’ (EOAD) and ‘late-onset Alzheimer’s disease’(LOAD). The distinction is arbitrary, since clinical andpathological features are very similar in both groups. Agecriteria for EOAD vary widely, but usually, when the ageat onset of the disease is before 65 years of age, a patientwill be diagnosed with EOAD [1].Given its strong association with age, AD will be anincreasing health care problem in the next decades. Withthe aging of western society, the number of patients isexpected to increase exponentially. By the year 2025, over22 million patients with dementia are expected aroundthe world [21]. Risk Factors AD has a complex etiology. Research in the past centu-ry has focused on many putative environmental factorsthat may either increase or decrease the risk of AD. Theseincluded age, smoking, maternal age at birth, head trau-ma, depression, thyroid disease, anti-inflammatory drugs,estrogen replacement therapy, alcohol, occupational ex-posure, aluminum, education and diet [22]. Findingsregarding these risk factors have been inconsistent. Onlyincreasing age and genetic predisposition are consistentlycorrelated with the disease [1].Perhaps most interesting from an epidemiological per-spective is the finding that studies on vascular risk factorssuch as hypertension, diabetes mellitus, atherosclerosis,and high cholesterol have yielded promising results,showing an up to 2 times increased risk for AD [23–32].The mechanism through which these vascular factors areassociated with AD remains to be elucidated. It has beenargued that these factors may be a primary cause of ADpathology [33]. An alternative explanation may be thatvascular pathology is not a primary cause of AD, but rath-er that it accelerates the primary neurodegenerative pro-cess.The findings of the relationship between vascular pa-thology and AD are in line with cross-cultural observa-tions by Hendrie et al. [34]. They recently publishedresults on differences in age-standardized annual inci-dence rates of AD in an industrialized versus a nonindus-trialized country. A possible explanation for the decreasedrate in the nonindustrialized country is the lower preva-lence of cardiovascular disease in the nonindustrializedpopulation. However, also differences in genetic makeupbetween populations may partly explain these findings.As discussed in the next chapter, genetic susceptibility is,in addition to increased age, the most important determi-nant of AD.  Alzheimer’s Disease: Genes, Pathogenesisand Risk Prediction Community Genet 2001;4:197–203 199 Genetics of AD  Familial Aggregation As opposed to the difficulties encountered in findingenvironmental risk factors for AD, the genetic componentof the disease has long been evident. Epidemiologicalstudies have clearly shown that AD aggregates withinfamilies [35]. First-degree relatives of AD patients have a3.5 times increased risk of developing AD. The relativerisk increases with a decrease in the age at onset of theaffected proband. In relatives of patients with an onsetbefore age 70 years the risk of having AD is increased over4 times [35]. Concordance rates of up to 80% have beenfound in monozygotic twins. In dizygotic twins concor-dance rates were 35% [36].In few families an autosomal dominant pattern of inheritance can be recognized. A segregation analysis sug-gested an autosomal dominant model in less than 1% of 198 families with EOAD [37]. In LOAD, it is difficult tomake a distinction between a dominant, recessive or addi-tive model of inheritance [38]. In the majority of patientsthe etiology appears to fit a multifactorial model in whichmultiple genes and environmental factors interact [1]. Genes Involved in AD Research on genetic determinants initially focused onfamilies with an autosomal dominant pattern of inheri-tance. The first dominant mutation was found in the geneencoding the amyloid precursor protein (APP) on chro-mosome 21 [2–5]. Up until now, 32 families are knownaround the world with EOAD due to a dominant  APP  mutation (http://molgen-www.uia.ac.be/ADMutations/).Besides  APP, two homologous genes were identified, presenilin-1 (PSEN-1) at chromosome 14 and presenilin-2 (PSEN-2) at chromosome 1q31-q42, which account forfamilies segregating AD as an autosomal dominant traitas well. So far, more than 80 mutations of the  PSEN-1 gene have been identified [6–10] (http://molgen-www.uia.ac.be/ADMutations/). Six mutations in  PSEN-2 are described [11, 12] (http://molgen-www.uia.ac.be/AD-Mutations/).Frequency estimates of these mutations in EOAD pa-tients are highly variable, ranging from less than 1 to 50%[e.g. 39–41]. Differences might be due to more or lessstringent diagnostic criteria in the population under study(e.g. probable vs. autopsy-confirmed AD), different maxi-mum age when considering early onset, and selection of study populations. A study population derived from ahighly specialized neurological center is more likely tohave an overrepresentation of cases with high familialaggregation. However, in a population-based sample [41]  APP mutations were found in only 0.5% of all EOADpatients and accounted for only 0.005% of AD in the gen-eral population. Although mutations in  PSEN-1 are morecommon, they still only accounted for 6.5% of all EOADpatients (i.e. 0.065% of AD in the general population).Mutations in  PSEN-2 were seen in less than 1% of allEOAD patients, and less than 0.01% of the general popu-lation. Together, dominant mutations in  APP  ,  PSEN-1 and  PSEN-2 occurred in only 0.075% of AD patients atthe population level [41, 42]. Although these genes have aminor impact in the general population, for the individualcarrier the risk is extremely high. Almost all carriers of these mutations express the disease. As EOAD is rare, riskestimates for carriers of these mutations approximateinfinity.In addition to the three autosomal dominant genes, afourth gene (apolipoprotein E,  APOE  ) was identifiedwhich is localized on chromosome 19 and has three com-mon alleles coding for three different isoforms of the pro-tein. The allele frequencies of this gene (APOE) are 0.08for  APOE*2, 0.77 for  APOE*3 and 0.15 for  APOE*4 inpopulations of European ancestry [42].  APOE*4 is strong-ly associated with LOAD [43, 44] and EOAD [45]. Sub- jects homozygous for  APOE*4 have an almost 15 timesincreased risk of developing AD, but 50% will not developthe disease [46]. Subjects with only one  APOE*4 allelehave a moderately increased risk (around 3 times) [47].Although risks are moderately increased for  APOE*4 forthe individual carrier, due to the fact that the allele iscommon  APOE*4 may explain 17% of the occurrence of AD in the general population [42]. Homozygosity for  APOE*4 contributes less than 2% to AD because of lowprevalence of this genotype (0.0225). It is suggested that  APOE*4 regulates when rather than if the disease occurs[47, 48]. Due to a relatively earlier onset of LOAD inthose homozygous for  APOE*4, the influence of compet-ing morbidity and mortality will be less, thereby enhanc-ing the association between  APOE*4 homozygosity andLOAD. Genes and the Pathogenesis of AD The discovery of mutations in the genes involved inAD has been of great importance for the understanding of the biological mechanisms underlying AD. All causal mu-tations affect the normal metabolism of ß -amyloid, sug-gesting that ß -amyloid constitutes a central event in thepathogenesis of AD. ß -Amyloid, or A ß , is a peptidepresent under physiological circumstances in healthy sub- jects. Due to a mutation in any of the known AD genes,  200 Community Genet 2001;4:197–203 Sleegers/vanDuijn the equilibrium between production and clearance of A ß gets disturbed, resulting in accumulation of A ß in thebrain. Amyloid fibrils are formed and subsequently de-posited into plaques. At present, the most likely hypothe-sis is that at first diffuse plaques are formed. Theseplaques can also be seen in healthy subjects. In ADpatients several of these plaques may evolve into ‘mature’neuritic plaques containing fibrillar aggregates, damagedneurons and activated glial cells in a cascade of pathologi-cal processes, eventually leading to profuse neuronal loss[49–57].APP is a transmembrane protein that is widely ex-pressed on the cell surface. Its functional properties arenot clearly defined, but range from repair of vascular inju-ry to mediation of growth and adhesion of neural andnonneural cells [53]. Recently it has been suggested thatAPP has a function in the regulation of nuclear transcrip-tion [58]. APP is cleaved into A ß . The different mutationsthat have been found so far in the gene coding for APP arelocated at or near cleavage sites [40]. By abnormal cleav-age of APP larger amounts of a longer version of A ß areproduced, called A ß 42 [59–62]. This longer version ismore amyloidogenic and therefore aggregates more easilyinto plaques [63]. There is increasing evidence that A ß 42may play a crucial role in the pathogenesis of AD.The function of the presenilin proteins is unclear, but ithas been shown that mutations in  PSEN also lead toaltered APP processing. A ß 42 levels are raised in brain,plasma and fibroblasts [64] of carriers of a  PSEN muta-tion. Furthermore, in experiments with  PSEN  -transgenicmice and transfected cells higher A ß 42 levels are found aswell [65–68].In sporadic LOAD the pattern of A ß accumulation isless evident. It has been suggested that accumulation israther the result of impaired clearance of A ß than of increased synthesis [57]. In carriers of the  APOE*4 allele,the predominant genetic risk factor in sporadic AD,APOE has increased affinity for A ß and A ß aggregates.Although the precise mechanism has not yet been eluci-dated,  APOE*4 is suggested to facilitate A ß aggregation orto inhibit the elimination of the fibrillar aggregates [44,69, 70].Although a body of genetic evidence supports the A ß cascade hypothesis and although it is the most compre-hensive theory so far, the debates on this hypothesis havenot been settled yet [50, 57, 71, 72]. Disagreement rangesfrom details within the A ß hypothesis (e.g. that not thetotal amount of A ß is important but rather the relativeproportion of A ß 42) to the reverse hypothesis that A ß accumulation is a compensatory mechanism to aging [73].A finding difficult to explain has been that amyloid depo-sition does not seem to correlate very well with cognitivedecline [50]. However, recent findings may have settledthis argument by showing that A ß plasma levels are ele-vated early in the course of the disease and are stronglyrelated to cognitive decline, a finding in favor of the A ß cascade hypothesis [74, 75].While the A ß hypothesis is being refined other patho-genic models are considered as plausible. These includehypotheses on the involvement of tau [76], on neuroplas-ticity [77], aging [72], oxidative stress [78], impaired cere-bromicrovascular perfusion [79], inflammation [80] andlipid homeostasis [70]. Genetic Counselling and Risk Prediction The identification of genes involved in AD has been amajor breakthrough. Yet there is ongoing debate on theiruse in clinical counselling.Given the fact that  APP, PSEN-1 and  PSEN-2 muta-tions have a virtually complete penetrance and that thesemutations are not found in healthy age-matched subjects,one might argue that these mutations are useful for riskprediction and genetic counselling. But the known muta-tions are only present in a minority of cases. Althoughmutations can be found frequently in patient populationsfrom highly specialized centers due to selection bias [81],these mutations are rare in the general population. Thus,for risk prediction and counselling, the absence of aknown mutation should not be conclusive. Even if theunderlying mutation is known in a family with an autoso-mal dominant form of AD, there is a strong argumentagainst screening relatives at risk, because curative thera-py and prevention are not yet at hand. An argument infavor of screening relatives at risk and those that alreadyhave dementia without a definite diagnosis might be totake away incertitude and allow for future plans, butscreening should always be preceded by thorough coun-selling, taking into account ethical considerations.Because of their low frequency, the mutations are notuseful as a diagnostic tool for patients with early onsetsymptoms of dementia [39].Also the use of  APOE is limited in the clinical practice.  APOE*4 increases the susceptibility to AD, but the in-crease in risk is modest, especially in the heterozygous car-riers of the  APOE*4 allele. As only 50% of the AD patientscarry  APOE*4 and a substantial number of patients withother dementias show similar frequencies,  APOE*4 is notsuitable for diagnostic purposes [82]. Even for subjectshomozygous for the  APOE*4 allele there is still a 50%chance not to develop the disease [46]. Thus despite the

Previous Document


Related Documents
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks