Early and Late Onset Alzheimer’s Disease: Two Different Entities?

The first case of Alzheimer’s disease (AD), a 51-year-old woman, was described in 1906. For several years, AD was assumed to be a rare form of presenile dementia. In 1976, Robert Katzman claimed that AD and senile dementia should be considered the same disease, on the basis of the identical cerebral changes, senile plaques and neurofibrillary tangles. Nowadays, pathologic, genetic, and molecular evidence suggest a distinction between type 1 (late onset) and type 2 (early onset) AD.

Presenile AD (EOAD), conventionally beginning before 65, is either familial or sporadic. Familial EOAD, accounting for 30% of cases, is an autosomal dominant disease caused by mutations in three genes, APP, PSEN1, and PSEN 2. The mean age of onset is 45, with atypical presenting symptoms such as epilepsy, myoclonus, paraparesis, and cerebellar ataxia. Sporadic EOAD starts at 60 on average, and the topography of cortical impairment and the corresponding symptoms mainly depends on the apolipoprotein E (ApoE) genotype. Indeed, ApoE4 homozygous cases have mesial temporal hypometabolism and classical amnestic onset. On the other hand, ApoE3/3, that are the majority of sporadic EOAD cases, show fronto-parieto-occipital impairment with alteration of symbolic functions.

EOAD substantially differs from the senile form in that it has a quick course, less frequent amnestic onset, impairment of symbolic functions, ApoE3 genotype, and higher tau pathology. Moreover, EOAD has a greater metabolic impairment than expected for a given amyloid load, as shown by PET amyloid and metabolic measurements. This suggests that factors other than amyloid-β, such as mitochondrial damage, oxidative stress, and inflammation, play a pathogenic role in presenile AD. Alternatively, the explanation may rest on the different ‘strains’ of soluble amyloid-β, not detectable with PET tracers. In familial, autosomal dominant EOAD, N-terminal truncated amyloid-β peptides prevail on the full-length 1-42 isoform. Moreover, in cognitive normal elderly cases, the full-length amyloid-β 1-42 is the major peptide, in comparison with sporadic AD. Thus, the composition of soluble amyloid-β may dictate the toxicity of the molecule and explain the large spectrum of aging/AD.

In conclusion, in LOAD, amyloid deposition is the main pathologic event that is strictly correlated with ApoE genotype. In sporadic EOAD, diffuse tau pathology and severe neuronal atrophy predominate. Whether different conformers of amyloid-β are responsible for the two pathological pictures is still unclear.

Last comment on 19 February 2016 by Gianluigi Zanusso, Medical


As reported by Massimo Tabaton, how Aβ different conformers are responsible for late as well as early AD is still unclear. In particular, how Aβ induces tau alteration and aggregation remain uncertain. A fascinating hypothesis could be represented by a nucleation effect of tau mediated by Aβ intracellular aggregates. Of note, the co-presence of the pathologic prion protein (PrPsc) with neurofibrillary tangles occurs only in some disease phenotypes, independent of the formation of amyloid fibrils. It is predictable that tau pathology depends on specific conformation of PrPsc, which is known to dictate the phenotype of prion diseases. Given the similarity of the conformational transmission of PrPsc with that shown for Aβ species, this could speculate that Aβ aggregates induce the aggregation of tau and that such event occurs only with a specific Aβ peptides or a mixture of peptides. Thus, soluble Aβ species, in the brain, is composed by three major species, 1-42, pyE3-42, and pyE11-42. The prevalence of the N-terminal truncated peptides is proportional to the severity of AD phenotype.

In this blog, Massimo Tabaton suggests, correctly, that EOAD and LOAD simply represent two sides of the same coin, and that the genetic context determines either phenotype. Dr. Tabaton also suggests that—in the hypothesis that AD is an amyloidosis which, in turn, causes tauopathy—it is the ratio between N-terminal truncated and full-length Aβ peptides that may determine the phenotype: EOAD when N-truncated Aβ predominate, LOAD when Aβ1-42/40 are the most abundant forms.

Considering the different ages at onset in familial and sporadic EOAD (45 versus 60 years of age), we must also assume that while in familial EOAD the genetic components (mutations on Presenilins/AβPP or enhanced AβPP levels) override any other effect, in sporadic EOAD and in LOAD, a dose-dependent effect of APOE ε4 on onset age predominates. Indeed, according to Tabaton’s theory, in sporadic EOAD and in LOAD, APOE ε4 could directly influence the phenotype: possibly clearing soluble full-length Aβ faster than shorter N-truncated peptides. In this scenario, however, the lack of clinical beneficial effects of therapies based on antibodies against Aβ peptides or by parallel approaches using β-γ inhibitors is disturbing. This suggests that inhibiting Aβ peptides may not be enough, and that other parallel, toxic events may be linked to their formation early in the neurodegenerative process.

Massimo Tabaton in his post raises an important issue: whether the disease phenotypes of early and late onset Alzheimer’s disease might be related to different conformers of amyloid-β (Aβ).

In the large spectrum of neurodegenerative disorders, emerging evidence indicates that the causative misfolded proteins share prion-like properties. An important characteristic of prion propagation is the ability to replicate distinct strains where the biological information are enciphered within different conformations of protein aggregates.

Although the correlation between protein conformers and strains is a consolidated issue in prion diseases, recent studies indicate that structurally distinct Aβ particles behave as distinct prion-like strains encoding diverse disease phenotypes. Since the identification of Aβ40 and Aβ42, several Aβ fragments had been characterized in AD brains with variable N- and C-terminal ragged ends, which might provide a correlation with distinct disease phenotypes. For instance, it has been shown that rapidly progressive AD variants correlated with distinct structural characteristics of Aβ42 and that the wide spectrum of AD clinical phenotypes might be related to variable Aβ42 conformers behaving as a prion-like strain [1]. To strengthen the existence of Aβ strains, transmission studies showed that susceptible transgenic mice challenged with brain tissue from sporadic and heritable AD propagate with distinct disease phenotypes indicating that Aβ enciphers specific biological information as prion strains [2].

Since in prion diseases, the ability of prion conformers to dictate a specific phenotypic variability is well known including the clinical phenotype, the distribution of brain lesions and pattern of PrP deposition in the brain tissue and transmission properties, it might be not unexpected that in AD the different disease phenotypes might be related to distinct Aβ strains.

Therefore, to provide an answer to the present question and other questions on AD phenotypes and Aβ, a suggested strategy will be a critical transfer of knowledge from prions to other prion-like disorders such as AD.

[1] Watts JC, Condello C, Stöhr J, Oehler A, Lee J, DeArmond SJ, Lannfelt L, Ingelsson M, Giles K, Prusiner SB (2014) Serial propagation of distinct strains of Aβ prions from Alzheimer's disease patients. Proc Natl Acad Sci U S A 111, 10323-10328.
[2] Cohen ML, Kim C, Haldiman T, ElHag M, Mehndiratta P, Pichet T, Lissemore F, Shea M, Cohen Y, Chen W, Blevins J, Appleby BS, Surewicz K, Surewicz WK, Sajatovic M, Tatsuoka C, Zhang S, Mayo P, Butkiewicz M, Haines JL, Lerner AJ, Safar JG (2015) Rapidly progressive Alzheimer's disease features distinct structures of amyloid-β. Brain 138, 1009-1022.