Segregation of a missense mutation in the amyloid protein precursor gene with familial Alzheimers disease

Alison Goate1

In 1991, little was known about the pathogenesis of Alzheimer's disease (AD). Earlier studies had demonstrated that plaques contain amyloid p (Ap) and that neurofibril-lary tangles were composed of paired-helical filaments of hyperphosphorylated tau (Glenner and Wong 1984a; Masters et al. 1985a; Grundke-Iqbal et al. 1986a). However, a major impediment to a more detailed understanding of AD was the absence of cellular or animal models of disease.

Mutations in APP cause AD and stroke resulting from cerebral hemorrhage

It has been known for more than 50 years that families exist in which AD has an early onset (< 60 years) and is inherited as an autosomal dominant trait (Familial Alzheimer's disease (FAD); Lowenberg and Waggoner 1934), but the techniques of molecular genetics only made analysis of these families feasible in the late 1980s. Initial studies of FAD focused on chromosome 21 because individuals with Down Syndrome all develop AD and Ap is derived from a precursor, p-amyloid protein precursor (APP), that is encoded by a gene on chromosome 21 (Goldgaber et al. 1987b; Kang et al. 1987; Robakis et al. 1987b; Tanzi et al. 1987). However, the APP gene was quickly excluded in several families (Tanzi et al. 1987; Van Broeckhoven et al. 1987). At this time, FAD was assumed to be a homogeneous disorder; therefore, exclusion of APP in one family was thought to exclude the gene in all families.

A turning point in AD genetics was a multi-center investigation that analyzed many families and came to the conclusion that FAD exhibited non-allelic genetic heterogeneity (St George-Hyslop et al. 1990). Shortly thereafter, two papers reported linkage to the APP gene and a mutation in APP in a disorder called hereditary cerebral hemorrhage with amyloidosis, Dutch type (HCHWA-D; Levy et al. 1990; Van Broeckhoven et al. 1990). These papers led our group to re-evaluate the APP gene in our own series of FAD kindreds. We had previously reported linkage to chromosome 21 in these families (Goate et al. 1989). Segregation analysis of multiple markers along chromosome 21 in the largest of these families demonstrated a common disease haplotype in all affected individuals. Information from two unaffected individuals placed the disease gene between D21S1 and D21S17, a region that includes the APP gene. Exons 16 and 17 were sequenced first because these exons encode the Ap peptide and because the mutation

1 Depts. of Psychiatry, Neurology & Genetics, Washington University School of Medicine, 660 S.

Euclid Ave., St. Louis, MO 63110

that causes HCHWA-D is in exon 17. This sequencing revealed a mutation that results in a missense mutation, V717I (Goate et al. 1991). This mutation was present in all affected individuals in the family but none of the unaffected individuals. Furthermore, it was absent from 100 unrelated normal individuals but present in a second, early-onset FAD kindred. The V717I substitution is conservative but its location, close to the C-terminus of the Ap peptide, suggested that it might influence production of Ap.

We made four predictions: 1) other FAD kindreds would be identified with APP mutations; 2) other FAD genes would be identified; 3) Ap deposition is the central event in the pathogenesis of AD; and 4) regulatory variants in APP might lead to late onset AD.

Mutations in APP alter processing or the physico-chemical properties of Ap

Eight months after our original report, we reported a second mutation in APP that caused FAD (Chartier-Harlin et al. 1991). This mutation was also at codon 717 but resulted in a V717G amino acid substitution. Based upon the two mutations, we hypothesized that FAD mutations in APP alter APP processing to enhance Ap production and thus Ap deposition. In the 15 years since the publication of these papers, 23 amino acid substitutions have been described in the APP gene (http//:www.alzforum.), 19 of which have been shown to alter Ap metabolism in vitro or cause age-dependent Ap deposition in vivo (reviewed in Selkoe and Podlisny 2002) (Fig. 1, Table 1). In vitro overexpression of APP FAD mutations has demonstrated that all mutations affect APP processing, leading to changes in the amount of Ap produced, changes in the ratios of the Ap species produced and/or changes in the physico-chemical properties of Ap. The so-called Swedish mutation results in an APP molecule that is a better substrate for p-secretase, resulting in higher levels of Ap (Citron et al. 1992). In contrast, FAD mu-

Swedish Mutation

fi-secretase a-sec reíase y-secretase

Fig. 1. Location of disease-causing mutations in APP. APP is a type 1 transmembrane protein. FAD mutations in APP are located within and flanking the Ap sequence and close to the proteolytic cleavage sites within APP. FAD mutations are shown in red above the normal sequence of the protein. Numbers indicate the amino acid position within the Ap peptide. The locations of the major proteolytic cleavage sites in APP are indicated by arrows below the sequence fi-secretase a-sec reíase y-secretase

Fig. 1. Location of disease-causing mutations in APP. APP is a type 1 transmembrane protein. FAD mutations in APP are located within and flanking the Ap sequence and close to the proteolytic cleavage sites within APP. FAD mutations are shown in red above the normal sequence of the protein. Numbers indicate the amino acid position within the Ap peptide. The locations of the major proteolytic cleavage sites in APP are indicated by arrows below the sequence

Table 1. Pathogenic mutations in the APP gene (AU)


(number of families)


Age of Onset (years)

KM670NL (Swedish) (1)

AD + CAA (need to define?)


D678N (1)



A692G (Flemish) (2)

AD / cerebral hemorrhage


E693G (Artic) (2)



E693K (Italian) (3)



E693Q (Dutch) (4)

E693Q (Dutch) (4)

Cerebral hemorrhage


D694N (Iowa) (2)



cerebral hemorrhage

L705V (1)



T714I (Austrian) (3)



T714A (Iranian) (2)



V715M (French) (1)



V715A (German) (3)



I716V (Florida) (1)



I716T (1)



V717F (3)



V717G (1)



V717I (London) (29)



V717L (Indiana) (3)



L723P (Australian) (1)



Adapted from and com/mut/default.asp

Adapted from and com/mut/default.asp tations located between APP714 and APP723 result in altered cleavage by y-secretase (Suzuki et al. 1994). The effect of these mutations is more complexbecause the amount of Ap and the ratios of the different Ap species (Ap37-Ap43) vary with each mutation (Hecimovic et al. 2004). However, a common feature of all mutations seems to be an increase in Ap42 relative to other Ap species.

Five mutations have been reported within the Ap sequence at residues APP692-694. These mutations are often associated with cerebral hemorrhage rather than AD (Levy et al. 1990; Hendriks et al. 1992; Nilsberth et al. 2001). Although these mutations are located near the a-secretase cleavage site and thus could alter APP processing, they are also within the Ap peptide and thus alter the physico-chemical properties of the peptide, leading to increased protofibril formation (Nilsberth et al. 2001; Stenh et al. 2002).

Several of these mutations have also been used to develop transgenic animals (Games et al. 1995; Sturchler-Pierrat et al. 1997; Hsiao 1998). A consistent property of these animals is an age-dependent Ap deposition. Another, striking observation coming from these mice is that overexpression of Ap42 leads to parenchymal Ap deposition, such as that seen in AD, whereas overexpression of Ap40 leads to Ap deposition primarily in the cerebral vessels (Herzig et al. 2004). Thus APPSwe, which results in higher levels of both Ap40 and Ap42, leads to both pathologies (Fryer et al. 2005), whereas APP717 mutations lead primarily to parenchymal Ap deposition

(Games et al. 1995) and APP692 leads to Ap deposition in the cerebral vessels (Herzig et al. 2004). Recently, duplication of the APP gene has been reported in several families. Consistent with data from transgenic mice, families that overexpress APP but do not have altered Ap ratios have both cerebral hemorrhage and dementia (Rovelet-Lecrux et al. 2006).

Mutations in at least three genes can cause FAD

In 1995, mutations in two homologues now called presenilin 1 (PS1) and presenilin 2 (PS2) were reported in several large FAD kindreds (Sherrington et al. 1995; Levy-Lahad et al. 1995b; Rogaev et al. 1995). In vitro and in vivo studies have demonstrated that FAD mutations in PS1 and PS2 also lead to changes in y-secretase cleavage of APP, resulting in higher Ap42/Ap40 ratios and early Ap deposition (reviewed in Selkoe and Podlisny 2002). All known FAD mutations appear to alter APP processing to produce more Ap, increase the propensity of Ap to form protofibrils or alter the ratio of the Ap species. Most early onset FAD kindreds appear to carry a mutation in either the substrate or the enzyme that generates Ap. It is unclear how many other FAD genes there are because most large FAD kindreds carry a mutation in one of the three known genes.

A major focus of current genetic research is the identification of genetic risk factors for late-onset AD (LOAD). Currently, the only known genetic risk factor for LOAD is APOE4 (Strittmatter et al. 1993a; Corder et al. 1993). However, only 50% of AD cases carry one or more copies of the E4 allele, suggesting that there must be other risk factors.

Is Ap deposition central to the disease process?

The third prediction has proven to be the most controversial. While it is clear that FAD mutations in APP result in increased Ap deposition, it is unclear whether the deposition is itself pathogenic. Several alternative hypotheses have been put forward. Rather than the deposited Ap being neurotoxic, some have suggested that the neurodegeneration observed in AD is caused by either soluble oligomers of Ap, the build-up of C-terminal fragments of APP or abnormal signaling by the intracellular domain of APP (Neve and Robakis 1998; Walsh and Selkoe 2004b).

A key question for many years was whether LOAD also involves an Ap-centric mechanism. Elegant transgenic studies have demonstrated that Apolipoprotein E (APOE) is required for Ap fibrillogenesis (amyloid formation) and that APOE4 promotes Ap deposition and amyloid formation (Holtzman et al. 2000a). The fact that all four known AD genes implicate Ap and that APOE implicates Ap fibrillogenesis directly provides support for the hypothesis that Ap deposition is central to the disease.

Can overexpression of APP lead to AD?

The fourth prediction was that variants in APP that altered the level of APP expression might also result in AD. Suprisingly, this hypothesis has not been rigorously tested.

Genetic linkage studies in LOAD families have reported evidence for linkage on the long arm of chromosome 21 (Myers et al. 2002). Despite these promising results, genetic analyses of the APP gene in LOAD have been very limited and inadequate to test the hypothesis. Thus, more than 15 years after the original report ofa missense mutation in the APP gene causing early onset FAD, the APP gene remains a promising but untested candidate risk factor for LOAD.


Our report of a missense mutation in the APP gene that caused FAD provided an important turning point in AD research. This paper and subsequent papers provided information that has led to the development of cellular and animal models that recapitulate at least part of the AD phenotype. These models have greatly enhanced our understanding of the pathobiology of AD and have led to the identification of drug targets for AD and the development of drugs that are currently in clinical trials. Furthermore, the major predictions of this paper have withstood the test of time remarkably well.

Rudolph E. Tanzi
Unraveling Alzheimers Disease

Unraveling Alzheimers Disease

I leave absolutely nothing out! Everything that I learned about Alzheimer’s I share with you. This is the most comprehensive report on Alzheimer’s you will ever read. No stone is left unturned in this comprehensive report.

Get My Free Ebook

Post a comment