Archives for March 2012

Creativity and Dementia: a Review

Cogn Process. 2012 Aug;13(3):193-209. doi: 10.1007/s10339-012-0439-y. Epub 2012 Mar 22.

Creativity and dementia: a review.

Palmiero MDi Giacomo DPassafiume D.

Department of Internal Medicine and Public Health, University of L’Aquila, Piazzale S. Tommasi n.1, 67010, Coppito L’Aquila, Italy,


In these last years, creativity was found to play an important role for dementia patients in terms of diagnosis and rehabilitation strategies. This led us to explore the relationships between dementia and creativity. At the aim, artistic creativity and divergent thinking are considered both in non-artists and artists affected by different types of dementia.

In general, artistic creativity can be expressed in exceptional cases both in Alzheimer’s diseaseand Frontotemporal dementia, whereas divergent thinking decreases in dementia.

The creation of paintings or music is anyway important for expressing emotions and well-being. Yet, creativity seems to emerge when the right prefrontal cortex, posterior temporal, and parietal areas are relatively intact, whereas it declines when these areas are damaged.

However, enhanced creativity in dementia is not confirmed by controlled studies conducted in non-artists, and whether artists with dementia can show creativity has to be fully addressed. Future research directions are suggested.


Family Stigma and Caregiver Burden in Alzheimer’s Disease

Gerontologist. 2012 Feb;52(1):89-97. doi: 10.1093/geront/gnr117. Epub 2011 Nov 1.

Family stigma and caregiver burden in Alzheimer’s disease.

Werner P1, Mittelman MS, Goldstein D, Heinik J.



The stigma experienced by the family members of an individual with a stigmatized illness is defined by 3 dimensions: caregiver stigma, lay public stigma, and structural stigma. Research in the area of mental illness suggests that caregivers‘ perception of stigma is associated with increased burden. However, the effect of stigma on caregiver burden among those caring for a relative with Alzheimer’s disease (AD) has yet to undergo theoretical and empirical testing. The aim of this study was to examine whether family stigma is a predictor of caregiver burden in the case of Alzheimer’s disease.

Design and Methods

Structured face-to-face interviews were conducted with 185 adult child caregivers (75% female; mean age = 53 years) for persons with Alzheimer’s disease.


Caregiver stigma variables improved the prediction of caregiver burden by adding an additional 18% to the explained variance over and above the other covariates. The 2 caregiver stigma variables of shame and decreased involvement with caregiving were found to be the major contributing factors.


Our findings show that caregiver stigma increases caregiver burden in the case of Alzheimer’s disease. Using this knowledge, psychosocial interventions should target stigmatic beliefs in order to reduce caregiver burden.



Cause of Death in Patients with Dementia Disorders

Eur J Neurol. 2009 Apr;16(4):488-92.

Cause of death in patients with dementia disorders.

Brunnström HREnglund EM.

Department of Pathology, University Hospital, Lund, Sweden.



Investigations on cause of death may provide valuable information about life expectancy and on conditions of terminal dementia care, which perhaps can be ameliorated.


The autopsy reports were studied on all patients (n = 524; 55.3% females; median age 80 years) with a clinically and neuropathologically diagnosed dementia disorder who underwent a complete autopsy at the University Hospital in Lund, Sweden, during 1974-2004.


The two most common causes of death were bronchopneumonia (38.4%) and ischaemic heart disease (23.1%), whilst neoplastic diseases were uncommon (3.8%). In a general population of elderly studied for comparison, bronchopneumonia accounted for 2.8%, ischaemic heart disease for 22.0%, and neoplasm for 21.3% of the deaths. Amongst the demented patients, circulatory and respiratory system diseases were the causes of death in 23.2% and 55.5% of the Alzheimer patients, respectively, whilst the corresponding figures were 54.8% and 33.1% for the patients with vascular dementia.


In patients with dementia, pneumonia as the immediate cause of death may reflect a terminal stage in which patient care and feeding is difficult to manage well. Knowledge about what actually causes death is of value in the terminal care of patients with dementia disorders.


A Hundred Years of Alzheimer’s Disease Research

Neuron. 2006 Oct 5;52(1):3-13.

A hundred years of Alzheimer’s disease research.

Hardy J1.

Main Text

On November 3, 1906, Alois Alzheimer gave a lecture to the Meeting of the Psychiatrists of South West Germany, presenting the neuropathological and clinical description of the features of one of his cases, Auguste D., who had died of a dementing illness at the age of 55, having been brought to his attention by her worried family at the age of 51 (Alzheimer, 1907). In his short description he discussed the clinical features of her case (which, while clearly including a progressive dementia, is actually rather unusual for our present expectation of the disease), and he also described the pathological features he observed in her brain, which included the presence of plaques (miliary foci) and tangles (fibrils). Alzheimer was not the first to describe the clinical features of the disease (which have been documented in the elderly since the ancient Greeks), nor was he the first to describe the plaques (Redlich, 1898). He does, however, seem to have been the first to describe the tangles (Perusini, 1991); these pathological advances were enabled by the development of silver stains (Bielschowsky, 1902). The disease was given his eponym by his senior colleague, Kraepelin (1910), almost certainly, in part, because of his statement that he was describing a new disease entity:

We must not be satisfied to force it into the existing group of well known disease patterns.

This description initiated the slow separation, using clinicopathologic criteria, between Alzheimer’s disease and other causes of presenile dementia, which included Pick’s disease, Creuzfeldt-Jakob disease, and Gerstmann Straussler and Worster Drought syndromes.

The silver stained slides of Auguste D.’s brain have recently been rediscovered, and their examination has confirmed that we currently use the term Alzheimer’s disease to describe the same diagnostic entity ( [Maurer et al., 1997] and [Graeber et al., 1998]). Unfortunately, no family history information is given for Auguste D., and so we do not know whether she had the autosomal dominant form of the disease. It is likely that, as molecular biological techniques are improved, this will be assessed through PCR amplification of her DNA from the silver stained sections of her brain. We already know that Auguste D. was an apolipoprotein E3 homozygote (Graeber et al., 1998).

After Kraepelin’s naming of the disease, the term was usually used to describe cases with a presenile (classically, less than 65 years) onset age, although Kraepelin’s paper discusses this issue, as do other authors’ over the years ( [Kraepelin, 1910] and [Newton, 1948]; see also Beach, 1987). Often, in these early case descriptions (which of course predate molecular staining techniques), it is not clear whether any particular case has what we would now describe as Alzheimer’s disease or one of the other dementias, most usually a variant of prion disease. Late onset “senile dementia” was barely studied and was generally ascribed to hardening of the arteries: a concept, which though long out of favor, may indeed hold an element of truth (Atwood et al., 2002).

The first autosomal dominant cases of Alzheimer’s disease were described by [Schottky, 1932] and [Van Bogaert and Maert, 1940], and Essen-Moller (1946). Although the family described by Schottky was lost to followup, the families described by van Bogaert were used over 50 years later in chromosome 14 linkage studies (Van Broeckhoven et al., 1992), and the family described by Essen-Moller underwent one of the first documented instances of genetic counseling when their disease was found to be caused by a presenilin mutation (Gustafson et al., 1998).

The modern era of Alzheimer’s research started with the realization that the majority of cases with senile dementia actually had Alzheimer’s pathology, changing Alzheimer’s disease from a rare neurological curiosity to a major research priority. In a series of influential papers, Blessed, Tomlinson, and Roth made the observation that the majority of cases of what had been called “senile dementia” had the plaque and tangle pathology of Alzheimer’s disease ( [Blessed et al., 1968], [Tomlinson et al., 1968] and [Tomlinson et al., 1970]). While the separation between “Alzheimer’s disease” (with an age of onset of <65 years) and Senile Dementia of the Alzheimer Type (SDAT) (with an age of onset of >65 years) was maintained for several years, it became clear that understanding the biology of Alzheimer’s disease would help in the development of an understanding of the causes of dementia in the elderly, and the nosological separation was dropped (Katzman, 1976). The formation of Alzheimer’s Associations in the United States in 1979 and later in other countries and increased research funding followed from this realization (Khachaturian, 2006).

Three basic science research tracks were pursued to try to develop an understanding of the disease. The first, inspired by the remarkable success of the neurochemical pathology work of Carlsson and Hornykiewicz on Parkinson’s disease that led to L-dopa therapy, sought to develop an understanding of any possible selective neurotransmitter loss in Alzheimer’s (Hornykiewicz, 2002). This approach has led to the currently available therapies, largely based on cholinergic cell loss. The second track approached the disease through the understanding of its pathognomic lesions, the neuritic plaque and the neurofibrillary tangle, and the third track took a positional cloning strategy in Mendelian forms of Alzheimer’s to find the causative variants behind the disease etiology. Unexpectedly, the positional cloning strategy and the pathology approach led to a convergent outcome, which has led to an integrated approach to trying to develop an understanding of the disease pathogenesis.

The Neurochemical Approach

In 1976, a fairly selective deficit in cortical markers of cholinergic neurons was discovered in a number of patients ( [Davies and Maloney, 1976], [Bowen et al., 1976] and [Perry et al., 1977]). At the time, the source of the origin of cortical innervation was not known, but was soon discovered to be in the basal forebrain (Wenk et al., 1980) where there was shown to be profound neuronal damage (Whitehouse et al., 1982). In parallel with this literature showing loss of cortical innervation from the nucleus basalis in Alzheimer’s disease was work on experimental animals and human volunteers showing the importance of cholinergic transmission in memory formation ( [Davis et al., 1978] and [Bartus, 1979]). This combined literature led to the prevalent view that modulating cholinergic transmission might offer a rational route to the symptomatic treatment of Alzheimer’s disease. The first, flawed study of a cholinesterase inhibitor (Summers et al., 1986) was followed by other therapies based on cholinesterase inhibition. Currently, nearly all approved therapies are based on cholinesterase inhibition, but these are only marginally effective, without affecting the course of the illness (Lleo et al., 2006). There are probably several reasons why cholinesterase inhibitors have only marginal efficacy, but among them is almost certainly the fact that the selectivity of the disease process for cholinergic innervation was greatly overstated. From the earliest days of neurochemical investigation, it was clear that catecholamine innervation to the cortex was also affected (Adolfsson et al., 1979), and the most prominent pathology in Alzheimer’s is neurofibrillary tangles in cortical pyramidal neurons, the majority of which probably use glutamate as a transmitter (Hardy et al., 1987). Thus, the transmitter replacement approach to Alzheimer’s therapy has likely reached the limit of its potential.

Dissecting the Composition of the Lesions

The first stage in the examination of the structure of the pathognomic Alzheimer’s lesions was by electron microscope. Kidd (1963) identified the structure of the tangles as consisting of paired helical filaments, and Terry et al. (1964) detailed the complex structure of the neuritic plaque and its content of amyloid fibers. Miyakawa and Uehara (1979) described the close physical relationship between the amyloid angiopathy in the disease and the neuritic plaques, and implied that the amyloid in these lesions was likely to be chemically similar.

Determining the chemical compositions of the fibrils in both the plaques and the angiopathic vessels, and the chemical compositions of the paired helical filaments of the tangles, took careful biochemical analysis. [Glenner and Wong, 1984a] and [Glenner and Wong, 1984b] derived the partial sequence of Aβ (which Glenner called β-amyloid) from the angiopathic meningeal vessels of both Alzheimer’s disease and Down’s syndrome. The abstract of his paper on Down’s syndrome (Glenner and Wong, 1984b) is remarkable for its prescient suggestion that familial Alzheimer’s disease could be caused by a mutation of the amyloid gene and its prediction that overexpression of the gene would also cause disease. Masters et al. (1985) derived the same sequence (which they called A4 because they thought the pathogenic species was an oligomer, possibly tetrameric) from plaques, and these findings led to the cloning of the APP gene in 1987 and its localization, as predicted, to chromosome 21 ( [Goldgaber et al., 1987] and [Kang et al., 1987]).

Tau had previously been identified as a protein involved in microtubule assembly (Weingarten et al., 1975). Its identification as the central component of the tangles was made through antigenic analysis ( [Brion et al., 1985], [Wood et al., 1986], [Kosik et al., 1986] and [Nukina and Ihara, 1986]), through protein purification methods (Grundke-Iqbal et al., 1986), and eventually, through direct sequencing of the isolated pair helical filament peptides (Goedert et al., 1988). Tau protein in the tangles was shown to be phosphorylated (48), which was widely believed to underlie its abnormal deposition.

Positionally Cloning the Genes Involved in Alzheimer’s Pathogenesis

Although the first genetic linkage using DNA technologies was to Duchenne Muscular Dystrophy (Murray et al., 1982), it was the localization of the Huntington’s gene to chromosome 4 in 1983 which awakened the neuroscience community to the power of the positional cloning strategy in finding genes for disease (Gusella et al., 1983). The application of molecular genetics to the problem of Alzheimer’s disease got off to a false, but prescient, start. Contemporaneously with the cloning of the APP gene described above, the first genetic linkage studies reported that the APP gene and the Alzheimer’s locus were one and the same ( [St George-Hyslop et al., 1987a] and [Tanzi et al., 1987a]), and further reports indicated that the locus was duplicated in some cases of Alzheimer’s disease (Delabar et al., 1987). All these reports were wrong in the families studied (Tanzi et al., 1987b): the families used for the linkage study were later shown to be chromosome 14-linked (St George-Hyslop et al., 1992), and the duplication report seems simply to have been in error (St George-Hyslop et al., 1987b). However, later events proved that the ideas behind these findings had merit (see below).

Genetic analysis of families with Alzheimer’s disease showed that some families did indeed show linkage to chromosome 21 markers (Goate et al., 1989), but that the disease was genetically heterogeneous ( [Schellenberg et al., 1988] and [St George-Hyslop et al., 1990]). At this time, Frangione and colleagues had shown that Aβ was deposited in the blood vessel walls in hereditary cerebral hemorrhage with amyloidosis, Dutch type (HCHWA-D) (van Duinen et al., 1987). Van Broeckhoven and colleagues and Frangione and colleagues ( [Van Broeckhoven et al., 1990] and [Levy et al., 1990]) showed that an APP mutation underpinned this disorder. With this background, focused sequencing of the APP gene in those Alzheimer’s families that showed evidence for genetic linkage to chromosome 21 identified APP mutations (Goate et al., 1991). The precise positions of these mutations showed that their most likely effect was to alter APP processing ( [Goate et al., 1991], [Murrell et al., 1991], [Chartier-Harlin et al., 1991], [Mullan et al., 1992] and [Hardy, 1992]) (see below). More recently, APP duplications in families with a variant of Alzheimer’s disease have indeed been discovered (Rovelet-Lecrux et al., 2006), and this, together with the absence of Alzheimer’s disease in Down’s syndrome patients who were trisomic distal of the APP locus (Prasher et al., 1998), clearly shows that APP overexpression can be a cause of Alzheimer’s. Thus, the two errors of Alzheimer’s linkage to the APP locus and APP duplication published in 1987 turned out to be presciently correct in families other than the ones in which they were reported.

In general, families with APP mutations have a typical Alzheimer’s disease phenotype (Mullan et al., 1993), although some mutations, including the APP duplications, have a phenotype reminiscent of HCHWA-D ( [van Duinen et al., 1987], [Levy et al., 1990], [Rovelet-Lecrux et al., 2006] and [Hendriks et al., 1992]). The age of onset of cases with APP mutations is typically in the late 40s to mid 50s. Both cases with APP mutations and cases with Down’s syndrome can have Lewy bodies as well as tangles ( [Lantos et al., 1994] and [Lippa et al., 1999]).

The majority of families with autosomal dominant Alzheimer’s disease did not have APP mutations. In 1992, Schellenberg and colleagues identified linkage to chromosome 14 (Schellenberg et al., 1992), and in 1995, St. George-Hyslop and colleagues identified presenilin 1 as the major locus underlying the early onset autosomal dominant disease (Sherrington et al., 1995). Genetic analysis had shown that a group of German families from Russia (the Volga Germans) did not show linkage to chromosome 14 markers (Levy-Lahad et al., 1995a), but rather showed linkage to genetic markers on chromosome 1 (75). A homolog of presenilin 1, presenilin 2, mapped to this region and was shown to underlie the remaining few cases of the autosomal dominant disease ( [Levy-Lahad et al., 1995a], [Levy-Lahad et al., 1995b] and [Rogaev et al., 1995]). In general, cases with presenilin 1 mutations also have typical Alzheimer’s disease with an age of onset in the late 30s to mid 40s (Martin et al., 1991) though some cases present with spastic paraparesis and have unusual “cotton wool” plaques (Crook et al., 1998). Cases with presenilin 2 mutations also have features similar to typical Alzheimer’s disease, but the onset ages are more variable, even within a family (Bird et al., 1989). Cases with presenilin mutations also frequently have Lewy body pathology (Lippa et al., 1998).

ApoE as a Risk Factor for Late Onset Alzheimer’s Disease

In 1991, Roses and his colleagues reported genetic linkage to late onset Alzheimer’s disease to chromosome 19 markers (Pericak-Vance et al., 1991) and then identified apolipoprotein E, encoded on this chromosome, as an Aβ binding protein (Strittmatter et al., 1993). Genetic analysis showed that a common variant of this gene, apoE4, was a risk factor for disease (Corder et al., 1993), and that another common variant, apoE2, was negatively associated with disease (Chartier-Harlin et al., 1994). These remarkable findings are among the most replicated findings in biology (Roses, 2006). They appear to account for about a third of the population-attributable risk for developing Alzheimer’s disease (Farrer et al., 1995). However, the mechanism behind this association remains unclear, with most authorities suspecting a role in amyloid deposition (Holtzman, 2004). Despite extensive work, no other genetic risk factors for late onset disease have been validated (Bertram and Tanzi, 2004).

MAPT Mutations and α-Synuclein Mutations

All cases of autosomal dominant, early onset Alzheimer’s disease reported to date have mutations in either the APP or the presenilin genes. In 1994, Wilhelmsen and colleagues described genetic linkage to chromosome 17 makers in a family with a progressive dementia syndrome (Wilhelmsen et al., 1994). The clinical phenotype of this syndrome was very variable and included some cases that were clinically very similar to Alzheimer’s disease (Foster et al., 1997). A key observation was that the majority of these cases had tau pathology, including, in some cases, neurofibrillary tangles (Spillantini et al., 1998). Levy-Lahad and colleagues and Hutton and colleagues ( [Poorkaj et al., 1998] and [Hutton et al., 1998]) showed that all those cases with tau pathology had MAPT mutations. These data showed that tangles could be directly initiated by MAPT mutations and were extremely useful for animal modeling (see below).

Lewy bodies have long been recognized as a frequent part of the pathology of Alzheimer’s disease (Hansen et al., 1993), but are pathognomic of Parkinson’s disease. In 1996, Nussbaum and colleagues showed that mutations in the α-synuclein gene were one cause of the disorder (Polymeropoulos et al., 1997), and subsequently, Spillantini and colleagues showed that α-synuclein was the major protein component of Lewy bodies (Spillantini et al., 1997). These data parallel those for MAPT reviewed above and show that Lewy bodies could be directly initiated by α-synuclein mutations.

Making Transgenic Models of Alzheimer’s Pathology

Finding the genes involved in Alzheimer’s pathogenesis has provided the information necessary to make partial models of the disease process. APP transgenic mice develop plaques, but they don’t develop significant tau pathology (Games et al., 1995), nor do they develop extensive cell loss (Takeuchi et al., 2000). Crossing these mice with presenilin mutant mice potentiates this existing pathology, but does not lead to cell loss or tau pathology ( [Holcomb et al., 1998] and [Borchelt et al., 1997]). Such mice with plaques have memory decrements, and brain slices from them show LTP abnormalities which are Aβ-dependent ( [Hsiao et al., 1996], [Janus et al., 2000], [Morgan et al., 2000] and [Lesne et al., 2006]), although the relationship of these findings to the memory problems in Alzheimer’s disease, in which there is, of course, extensive cell loss, is not clear.

Mutant MAPT transgenic mice develop tangle pathology (Lewis et al., 2000), and these mice have severe cell loss in those brain regions in which tangles develop. Crossing these mice with APP transgenic mice potentiates the tangle pathology and the cell loss, but does not alter the plaque pathology (Lewis et al., 2001) (Figure 1), making the interpretation that APP is upstream of tangle pathology tenable (Hardy et al., 1998). While no α-synuclein transgenic mice develop Lewy bodies, the modest pathology which has been observed to develop (Masliah et al., 2000) is potentiated by crossing α-synuclein transgenic mice with APP mice (Masliah et al., 2001) suggesting that the relationships between Aβ and tau and Aβ and α-synuclein have parallels.

The Effects of Alzheimer-Causing Mutations on APP Processing

Assessment of APP metabolism showed that it underwent rapid turnover in cells (Weidemann et al., 1989) and that Aβ was a normal product of metabolism in vitro and in vivo ( [Haass et al., 1992] and [Shoji et al., 1992]). Analysis of the metabolism of APP with the pathogenic APP mutations showed that all of the pathogenic mutations altered APP metabolism such that more Ab42 was produced ( [Citron et al., 1992], [Cai et al., 1993] and [Suzuki et al., 1994]). These data, together with the demonstration that Aβ42 was the earliest species deposited in Alzheimer’s disease (Iwatsubo et al., 1994) and the suggestion that Aβ was neurotoxic (Yankner et al., 1990), formed the intellectual basis of the amyloid hypothesis for the disorder ( [Selkoe, 1991], [Hardy and Allsop, 1991] and [Hardy and Higgins, 1992]) and highlighted the importance of developing a molecular understanding of the three cleavages around the Aβ sequence: β-secretase, responsible for the N-terminal cut of Aβ from APP (at APP 671); α-secretase, responsible for the alternative cleavage of the shorter fragment p3 at around residue 17 of Aβ (APP687); and γ-secretase, responsible for releasing both Aβ and p3 from the relevant C-terminal stubs of APP and cleaving intramembranously around APP residues 711–713 (Haass and Selkoe, 1993).

What Are the Toxic Species of Aβ and Tau?

A continued area of active research has been aimed at understanding the potential mechanism of Aβ toxicity. While original formulations of the amyloid hypothesis envisaged the deposited peptide as neurotoxic, it has become increasingly clear that smaller species assemblies have both behavioral and electrophysiological effects ( [Lambert et al., 1998], [Walsh et al., 2002] and [Lesne et al., 2006]). A similar discussion has related to the understanding of the mechanism of cell death induced by tau: while clearly tangles are a substantial space-filling lesion (Sumpter et al., 1986), recent data suggests that the cell loss in MAPT transgenic mice has a more complicated pathogenesis than this simple space-filling effect (Santacruz et al., 2005).

APP Processing and Deposition as a Therapeutic Target in Alzheimer’s Disease


The identification of the Swedish mutation at the β-secretase site (Mullan et al., 1992) illustrated the potential value of modulating this cleavage for therapy. In an elegant series of APP mutagenesis experiments, Citron and colleagues showed that only the Swedish variant had a greater flux though the β-secretase pathway, explaining why only that single pathogenic mutation has been found at that site and providing an assay to help identify the enzyme responsible for it (Citron et al., 1995). Cloning of the β-secretase (BACE) gene revealed it to be a membrane-bound aspartyl protease (Vassar et al., 1999). Genetic deletion of BACE1 in mice completely abolishes Aβ production and deposition and rescues cognitive impairments in APP transgenic animals, suggesting that BACE1 is a high-priority therapeutic target for Alzheimer’s disease. However, there are several caveats associated with BACE1 inhibition. While BACE1 knockout mice do not produce Aβ and show no major pathological abnormalities, they do display subtle deficits in explorative activities as well as spatial learning and memory, implying potential side effects in completely abolishing BACE1 activity ( [Ohno et al., 2004] and [Laird et al., 2005]). Additionally, validating it as a pharmacological target for Alzheimer’s treatment (Citron, 2004), the large size of the active site of BACE1 (Hong et al., 2000 has made designing selective inhibitors that can pass through the blood-brain barrier a considerable challenge (Citron, 2004). Since complete inhibition of β- or γ-secretase activity presents considerable side effects, a cocktail of BACE1 and γ-secretase inhibitors that partially suppress these two proteolytic activities may serve as a better therapeutic strategy to block Aβ production.


α-Secretase cleaves at position 16 of the Aβ sequence (APP687) (Esch et al., 1990). The cleavage does not appear to be sequence-specific, but rather appears to rely on the maintenance of an α-helical domain at this point (Sisodia, 1992). It is inhibited by some of the mutations in the middle of the Aβ sequence that lead to Alzheimer’s disease, apparently because they disrupt this α-helical sequence (De Jonghe et al., 1998). Much less work has been carried out on the α-secretase cleavage because, as a treatment strategy, one would want to potentiate cleavage (Hooper and Turner, 2002), and this is perceived to be more difficult than inhibiting cleavage by β-secretase or γ-secretase. However, it seems likely that α-secretase cleavage is mediated by a mix of the metalloproteases ADAM9, ADAM10, and ADAM17 (Asai et al., 2003).


Analysis of tissues from presenilin mutation carriers and transfected cells showed that the presenilin mutations, like the APP mutations, increased the proportion of Aβ42 derived from APP processing (Scheuner et al., 1996). Presenilins were shown to be involved in the Notch pathway (Levitan and Greenwald, 1995), and mice in which the presenilin gene was knocked out suffered embryonic lethality (Wong et al., 1997). Cultured primary neurons from presenilin knockout mice were unable to cleave the C-terminal fragment of APP, showing that presenilins were intimately involved in this cleavage (De Strooper et al., 1998), as well as the analogous cleavage of Notch ( [De Strooper et al., 1999] and [Struhl and Greenwald, 1999]). Focused mutagenesis of two transmembrane aspartates in presenilin strongly suggested that these were the residues responsible for the γ-secretase intramembranous cleavage (Wolfe et al., 1999), and subsequent reconstitution experiments have conclusively shown that presenilin and other components (APH, PEN2, and Nicastrin) make up the core of the γ-secretase complex responsible for the cleavage of APP, Notch, and other transmembrane proteins (Edbauer et al., 2003), with possibly other accessory proteins contributing to the full complex (Chen et al., 2006).

While it is presumed that a lot off effort has been invested by the pharmaceutical industry in developing γ-secretase inhibitors, surprisingly little has been published. As discussed above, presenilin knockouts are embryonically lethal (Wong et al., 1997). However, conditional neuronal knockouts, and mice showing reduced expression of presenilin, are viable, although they have behavioral defects ( [Yu et al., 2001] and [Dewachter et al., 2002]) and harbor a potential for initiating tumorigenesis (Xia et al., 2001). These data were interpreted as suggesting that modest inhibition of γ-secretase may be a viable drug target despite its multifunctional role.

The most extensive work relevant to γ-secretase modulation has been published on the nonsteroidal anti-inflammatory drugs (NSAIDS). Some of these compounds, through a none-cyclo-oxygenase dependent process, modify γ-secretase activity (Weggen et al., 2001). NSAIDS had been proposed as treatments for Alzheimer’s disease based on the occurrence of many markers of inflammation in the pathology of the disease (McGeer and Rogers, 1992). However, these data suggest that some such compounds may also be allosteric modulators of the γ-secretase cleavage.

Aβ Immunization

In 1999, Schenk and colleagues (Schenk et al., 1999) reported that Aβ immunization of APP transgenic mice massively reduced their amyloid pathology. The motivation behind these experiments was not clear, but it seems likely that the intention was to try and induce tangle pathology and cell loss through activation of the immune response (Webster et al., 1997). The observation that Aβ immunization reduced amyloid pathology was multiply confirmed and extended to passive Aβ immunization and the observation that the mice’s memory function improved ( [Morgan et al., 2000] and [Janus et al., 2000]). Neuropathological examination of individuals with Alzheimer’s disease who died after receiving the immunization revealed that it did indeed reduce amyloid, but not tau, pathology (Nicoll et al., 2003). Unfortunately, however, the trial had to be stopped, because in a proportion of the Alzheimer’s cases (though not in the healthy volunteers) the immunization led to a meningoencephalitis (Winblad and Blum, 2003). It remains unclear whether this complication is related to the whole approach of Aβ immunization or to some idiosyncrasy of the trial ( [Schenk, 2004] and [Atwood et al., 2003]). No final analyses of the clinical effects of the trial have been reported (Winblad and Blum, 2003).

Other Treatment Approaches

The production and wide availability of transgenic mice exhibiting many of the pathological features of the disease has enabled (relatively) quick and inexpensive ways of testing therapies, largely against amyloid deposition, but also, more recently, against tau pathology (Noble et al., 2005). Most, but not all, of these approaches have been framed around the amyloid hypothesis of the disorder: besides NSAIDs, they include metal chelators, antioxidants, cholesterol lowering drugs, and tau phosphorylation inhibitors (Duff and Suleman, 2004). A surprising number of these approaches have been successful in mice, and currently, many of these approaches are in various stages of clinical trials.

Defining the Alzheimer’s Prodrome

A corollary of the attempts to develop mechanistic therapy for Alzheimer’s disease, as opposed to the limited palliative therapy now available, is the development of better means of making the diagnosis of the Alzheimer’s process at an early point. In this regard, the recognition of mild cognitive impairment (MCI) as an important aspect of this prodrome has had considerable impact (Petersen et al., 1999). MRI analysis of presenilin and APP mutation carriers has clearly shown that the degenerative process begins several years before clinical manifestations of the disease become evident (Scahill et al., 2002); using an amyloid binding PET ligand, Klunk and colleagues (Klunk et al., 2004) have developed a methodology which may be useful both for identifying people in the Alzheimer’s prodrome and monitoring the efficacy of clinical trials.

Concluding Remarks

As reviewed above, there have been essentially three periods of Alzheimer’s research (see also Table 1). The first was the definition of the disease by clinicians and pathologists. The second was the neurochemical work, which led to the identification of the cholinergic lesion in the disease, and upon which current therapies are based, and the third is the molecular biological and molecular genetic approach to the dissection of the pathogenesis. This last era of Alzheimer’s research has paralleled work on other neurologic and psychiatric diseases and undoubtedly holds great promise both for dissecting disease pathogenesis and for developing mechanism-based therapies. However, it has to be acknowledged that no therapies for any neurologic or psychiatric disease have, as yet, been developed by this approach.

Table 1. A Summary of the History of Research into Alzheimer’s Disease and Related Disorders
Year Related Developments Alzheimer’s Disease
1902 Improved silver stains
1906 Alzheimer’s case history of Auguste D.
1910 Alzheimer’s Disease named
1922 Lewy body described
1932 First hereditary case described
1962 L-dopa therapy in Parkinson’s
1963/4 Ultrastructure of plaque and tangle by electron microscopy
1968 Recognition of prevalence of disease in the elderly
1976 Cholinergic deficit described
1983 Huntington’s genetic linkage Sequence of Aβ from Alzheimer’s amyloid angiopathy
1984 Sequence of Aβ from Down’s syndrome from amyloid angiopathy
1985 Cloning of the prion gene Sequence of Aβ from plaques
1986 Tau as major component of tangles
1987 Cloning of APP and localization to chromosome 21
1989 Mutations in prion gene in CJD/GSS
1990 Mutations in APP cause HCHWA-D; prion mutations cause neurodegeneration in mice Alzheimer’s disease genetically heterogeneous
1991 APP mutations in Alzheimer’s; a descriptive system of cataloguing the neuropathology determined
1993 ApoE4 associated with Alzheimer’s; cholinergic therapy approved for AD
1994 APP mutations increase Aβ42
1995 APP transgenic mice made with plaque pathology; presenilins cloned as loci for early onset Alzheimer’s
1996 Presenilin mutations shown to alter APP processing
1997 Synuclein mutations identified in PD; Synuclein identified as major component of Lewy bodies
1998 Tau mutations identified in FTDP-17 Presenilins identified as γ-secretase
1999 BACE cloned; Aβ immunization in mice reduces amyloid pathology
2000 Mice with tangles made using FTDP-17 mutations
2001 Mice with plaques and tangles made
2003 Aβ vaccine trials halted because of side effects


By far, the prevailing hypothesis for disease pathogenesis based on these molecular data is the amyloid hypothesis, and as we approach the centenary of Alzheimer’s lecture, we are in a research limbo: many trials of agents based on the amyloid hypothesis are either in progress or being planned, but no completed, large-scale, double blind trials have yet been reported. These, of course, are the true test of the hypothesis, and will be the outcome by which it will be judged. If we can only treat the model disease in mice, it will not have been useful.

There remain several large gaps in our knowledge: is Aβ toxicity observed in cell systems relevant to the loss of neurons and to the disruption of memory and cognitive function in the disease? If so, what is the mechanism of Aβ and tau toxicity? What is the relationship, if any, between APP and tau or α-synuclein? What is the function of APP, and what, if any, is the function of Aβ? In addition, we have little understanding as to the reasons for the distinctive neuroanatomy of the disease (Braak and Braak, 1991), and the spatial relationships between the pathognomic lesions remain almost unstudied (Duyckaerts, 2004). We also have very little understanding of the role of apolipoprotein E in disease pathogenesis and, despite an enormous effort, we have not found other convincing genetic or environmental risk factors. The last 100 years, and especially the last 40 years, since the disease was rediscovered by Blessed, Tomlinson, and Roth has brought an enormous amount of scientific progress in terms of understanding the etiology of the disease. It has also brought a lot of progress in social attitudes to the disease in the reduction of its associated stigma. Medical application of this progress has, however, come slower and much remains to be done in the second century of Alzheimer’s research. If Auguste D. were alive today, her sad prognosis would be much the same as in 1906.


Copyright © 2015 Elsevier Inc. except certain content provided by third parties