VARIANT ALZHEIMER’S DISEASE WITH SPASTIC PARAPARESIS: CLINICAL ...ethesis.helsinki.fi/julkaisut/laa/kliin/vk/verkkoniemi/varianta.pdf · Department of Clinical Neurocience Helsinki

Departments of PathologyUniversity of Helsinki and

Helsinki University Central Hospital

Department of Clinical NeurocienceHelsinki University Central Hospital

Helsinki, Finland

VARIANT ALZHEIMER’S DISEASE

WITH SPASTIC PARAPARESIS:

CLINICAL, NEUROPATHOLOGICAL AND

MOLECULAR GENETIC CHARACTERIZATION

Auli Verkkoniemi

Academic dissertation

To be presented, by the permission of the Medical faculty of the University of

Helsinki, for public examination in the Auditorium 3 at Biomedicum, on

November 2nd, 2001, at 12 noon.

2

Supervisors

Professor Matti HaltiaDepartments of PathologyUniversity of Helsinkiand Helsinki University Central HospitalDocent Mirja SomerDepartment of Medical Genetics

University of Helsinki andThe Family Federation of Finland

Reviewers

Professor Hilkka SoininenDepartment of NeurologyUniversity of KuopioKuopio University Hospital

Professor Anna-Elina LehesjokiFolkhälsan Institute of Genetics andDepartment of Medical GeneticsUniversity of Helsinki

Opponent

Professor Philip ScheltensFree University of AmsterdamISBN 952-91-3980-2 (painettu)

952-10-0173-9 (pdf)

3

To Camilla and Jyrki

4

5

CONTENTS Page

LIST OF ORIGINAL PUBLICATIONS 8

ABBREVIATIONS 9

ABSTRACT 11

INTRODUCTION 13

REVIEW OF THE LITERATURE 15

1. Sporadic Alzheimer’s disease 15

1.1. Clinical features of sporadic AD 15

1.1.1. Symptoms and signs 15

1.1.2. Neuropsychology 16

1.1.3. Neurophysiology 16

1.1.4. Neuroimaging 17

1.2. Main neuropathological features 17

1.2.1. Plaques 17

1.2.2. Neurofibrillary changes 19

1.2.3. Amyloid angiopathy 19

1.3. Genetic risk factors 19

1.3.1. Apolipoprotein E 19

1.3.2. Others 20

2. Autosomal dominant Alzheimer’s disease 21

2.1. Genetics 21

2.1.1. Amyloid precursor protein (APP) gene mutations 21

2.1.2. Presenilin 1 (PS-1) gene mutations 22

2.1.3. Presenilin 2 (PS-2) gene mutations 23

2.1.4. Autosomal dominant Alzheimer’s disease and ApoE 24

2.2. Clinical features 24

2.3. Neuropsychological features 25

2.4. Neuropathological features 26

3. Hypotheses for the pathogenesis of Alzheimer’s disease 27

3.1. Amyloid cascade hypothesis 27

3.2. Other hypotheses 28

3.3. Function and dysfunction of the presenilins 28

4. Familial diseases causing dementia and spastic paraparesis 29

4.1. Prion diseases 29

4.1.1. Familial Creutzfeldt-Jakob disease (FCJD) 29

4.2.1. Gerstmann-Sträussler-Scheinker disease 29

4.2. Hereditary spastic paraplegia (HSP) 29

6

4.3. Familial British dementia (FBD) 30

4.4. Autosomal dominant Alzheimer’s disease and spastic paraparesis 31

AIMS OF THE PRESENT STUDY 33

PATIENTS AND METHODS 34

1. Patients 34

2. Ethical issues 34

3. Clinical methods 35

3.1. Neurological examination 35

3.2. Neuropsychological examination 35

3.3. Psychiatric evaluation 36

3.4 Cerebrospinal fluid (CSF) analysis 36

3.5. Neurophysiologic studies 36

3.6. Neuroimaging 36

4. Neuropathological methods 36

4.1. Histology and immunohistochemistry 37

4.2. Confocal microscopy 37

4.3. Electron microscopy 37

4.4. Biochemical analyses 37

5. Molecular genetic methods 38

5.1. Linkage analysis 38

5.2. Analysis of cDNA 38

5.3. Defining the mutation 38

5.4. ApoE genotyping 39

6. Statistical analysis 39

RESULTS 39

1. Clinical findings 39

1.1. Pedigree analysis 39

1.2. Clinical course 39

1.3. Neurological signs 40

1.3.1. Dementia 40

1.3.2. Spastic paraparesis 40

1.3.3. Clumsiness of hands 41

1.3.4. Other clinical findings 41

1.4. Neuropsychological findings 41

1.5. Psychiatric signs 42

1.6. Cerebrospinal fluid (CSF) findings 42

1.7. Neurophysiologic findings 42

7

1.7.1. EEG 42

1.7.2. EMG 42

1.7.3. Somatosensory evoked potentials 42

1.8. Findings in imaging studies 42

1.8.1. PET imaging and neuropsychological follow-up 43

2. Neuropathological findings 43

2.1. Macroscopic observations 43

2.2. Histologic findings 44

2.2.1. General description 44

2.2.2. Plaques 44

2.2.3. Neurofibrillary pathology 45

2.2.4. Amyloid angiopathy 45

2.2.5. Cerebellar pathology 46

2.2.6. Brain stem and spinal pathology 46

2.3. Electronmicroscopic findings 46

2.4. Biochemical findings 46

3. Molecular genetic findings 46

3.1. Linkage analysis 46

3.2. Analysis of cDNA 47

3.3. Defining the mutation 47

3.4. ApoE genotypes 47

DISCUSSION 47

1. Clinical phenotype 47

2. Neuropathologic phenotype 49

3. Genotype phenotype correlations 51

4. Pathogenetic implications 52

CONCLUSIONS 53

ACKNOWLEDGMENTS 54

REFERENCES 59

ORIGINAL ARTICLES 75

8

LIST OF ORIGINAL PUBLICATIONS

This thesis is based on the following articles which are referred to in the text bytheir Roman numerals:

I Verkkoniemi A, Somer M, Rinne J.O, Myllykangas L, Crook R, Hardy J, Vii-tanen M, Kalimo H, Haltia M. Variant Alzheimer’s disease with spastic para-

paresis: Clinical characterization. Neurology 2000;54:1103-1109.

II Verkkoniemi A, Ylikoski R, Rinne J.O, Somer M, Hietaharju A, ErkinjunttiT, Viitanen M, Kalimo H, Haltia M. Neuropsychological functions in variantAlzheimer’s disease with spastic paraparesis. Submitted

III Verkkoniemi A, Kalimo H, Paetau A, Somer M, Iwatsubo T, Hardy J, HaltiaM. Variant Alzheimer disease with spastic paraparesis: Neuropathologicalphenotype. J Neuropathol Exp Neurol 2001;60:483-492.

IV Crook R, Verkkoniemi A, Perez-Tur J, Mehta N, Baker M, Houlden H, FarrerM, Hutton M, Lincoln S, Hardy J, Gwinn K, Somer M, Paetau A, Kalimo H,Ylikoski R, Pöyhönen M, Kucera S, Haltia M. A variant of Alzheimer’sdisease with spastic paraparesis and unusual plaques due to deletion on exon 9of presenilin 1. Nature Medicine 1998;4:452-455.

V Prihar G, Verkkoniemi A, Perez-Tur J, Crook R, Lincoln S, Houlden H,Somer M, Paetau A, Kalimo H, Grover A, Myllykangas L, Hutton M, HardyJ, Haltia M. Alzheimer disease PS-1 exon 9 deletion defined. NatureMedicine 1999;5 (10) 1090.

9

ABBREVIATIONS

Aβ β-amyloid peptideABri A 4K protein subunit present in plaques of FBDAD Alzheimer’s diseaseALS Amyotrophic lateral sclerosisApoE Apolipoprotein EAPP Amyloid precursor proteinADRDA The Alzheimer’s Disease and Related Disorders AssociationcDNA Complementary or copy deoxyribonucleic acidCERAD Consortium to Establish a Registry of Alzheimer’s DiseaseCJD Creutzfeldt-Jakob diseasecM CentiMorganCNS Central nervous systemCT Computed tomographyCWP Cotton wool plaqueDNA Deoxyribonucleic acidEEG ElectroencephalographyEMG ElectromyographyFAD Familial Alzheimer’s diseaseFBD Familial British dementiaFDG 18F-fluorodeoxyglucoseGM Gray matterGSS Gerstmann-Sträussler-Scheinker syndromeHCHWA-D Hereditary cerebral hemorrhage with amyloidosis of Dutch typeHSP Hereditary spastic paraparesiscHSP Complicated hereditary spastic paraparesispHSP Pure hereditary spastic paraparesiskb KilobasekDa KilodaltonLFB Luxol Fast BlueMMSE Mini-Mental State ExaminationMRI Magnetic resonance imagingmRNA Messenger ribonucleic acidNFT Neurofibrillary tangleSD Standard deviationSEP Somatosensory evoked potentialsSPECT Single photon emission computed tomographyPCR Polymerase chain reactionPET Positron emission tomographyPHF Paired helical filamentsPS-1 Presenilin 1PS-2 Presenilin 2varAD Variant Alzheimer’s disease with spastic paraparesisVEP Visual evoked potentialsWM White matter

10

11

ABSTRACT

A novel variant of autosomal dominant Alzheimer’s disease (AD) with additionalearly neurological signs and interesting neuropathological features occurred in alarge Finnish family with 23 affected individuals (16 men and 7 women) in foursuccessive generations. The six surviving patients were personally examined, andneurological records of nine deceased patients were evaluated. In addition, eightcases occurred in the early twentieth-century generations, for whom the medicaldocuments were lacking. Electrophysiological investigations were available in eightcases. Neuroimaging had been performed on eleven patients, and in four patientsbrain glucose metabolism was examined by positron emission tomography (PET).The neuropsychological evaluation was based on tests of verbal and visual memory,abstract thinking, and visuoconstructive and spatial functions. Four survivingpatients were prospectively tested, and retrospective neuropsychological data werecollected from an additional four deceased patients. Their possible psychiatricsymptoms were also evaluated. Mean age at onset was 50.9±5.0 years (range 40 to61).

Apart from dementia, the most striking clinical feature was gait disorder due tospasticity of the lower extremities. Spastic paraparesis appeared simultaneously orwas preceded by memory impairment. It is noteworthy that spastic paraparesis waspresent in many but not all patients. Memory impairment was present in all pa-tients, and besides that, neuropsychological analysis showed visuoconstructional defi-cits which were an early and prominent feature. Psychiatric symptoms were presentin six patients, two of whom had euphoria and four depression. In addition, somepatients showed impaired fine coordination of hands and dysarthria that suggestedcerebellar involvement. EEG showed intermittent generalized delta-theta activity.MRI of the head showed temporal and hippocampal atrophy, whereas PET showedbilateral temporo-parietal hypometabolism. In addition, variable patterns of hy-pometabolism in PET (hemispherical asymmetry, occipital accentuation) were relatedto individual deficits of cognitive performance.

Molecular genetic analysis was unexpectedly troublesome because sequencing theexons and the immediate flanking intronic regions from genomic DNA amplified byPCR did not disclose the mutation, which was first identified as a deletion of exon 9through sequencing of PCR-amplified cDNA obtained from lymphoblastoidmRNA. By the aid of intronic primers flanking exon 9, the mutation was eventuallydefined as a 4,555kb genomic deletion between exons 8 and 10.

Neuropathological analysis of five affected family members revealed an unusualneuropathological phenotype. The primary and association cortices and hippocam-pus showed a profusion of eosinophilic, roundish structures with distinct borders,termed “cotton wool” plaques (CWPs). Predominant CWPs, a novel type of plaques,were immunoreactive for Aβ42/43, but weakly or not at all for Aβ40 isoforms of theamyloid β peptide (Aβ). CWPs were devoid of a congophilic core, and only occa-

12

sional structures resembling fibrillar amyloid could be identified by electron micros-copy. In addition to the predominant CWPs, variable numbers of diffuse and coredplaques were found in the cerebral cortex.

Degeneration of the lateral corticospinal tracts observed at the level of the me-dulla oblongata and spinal cord may be correlated with spastic paraparesis. Neu-ropathological examination showed that CWPs were particularly numerous in themedial motor cortex representing the lower extremities. Most of the affected indi-viduals showed clumsiness of the hands, dysdiadochokinesia, and several had inten-tion tremor. These findings are indicators of cerebellar damage. Diffuse and non-neuritic cored amyloid plaques but no CWPs occurred in the cerebellum in the neu-ropathological analysis.

CWPs were frequent in the entorhinal cortex and hippocampus; they were espe-cially abundant in the subiculum. The expansive CWPs displaced neurons and eveninterrupted and distorted the granular cell layer of the dentate gyrus as well as thepyramidal cell band of the hippocampus. Confocal microscopy showed reduced den-sity of axons within individual CWPs and few CWP-related dystrophic neurites.The axons seemed either to wind around the CWP or to degenerate immediatelyupon entering into the CWP. It is likely that this wide disruption of cortical cytoar-chitecture is associated with functional disturbances.

In conclusion, a novel phenotype of presenile dementia and spastic paraparesiswas identified. Besides the clinical presentation, also the neuropathological featureswere unusual. However, the plaques were immunoreactive for Aβ, indicating thatthe disease falls under the rubric of the Alzheimer syndrome. We have designatedthis clinical syndrome variant Alzheimer’s disease with spastic paraparesis (varAD).

Our observations in varAD constitute a strong argument against a simple read-ing of the classical amyloid cascade hypothesis. The presence of CWPs without amy-loid cores supports the view that the extracellular, poorly soluble, congophilic amy-loid plaques are not the only initiators of neurodegeneration in AD. However, be-cause the exceptionally high production of Aβ42/43 in varAD is likely to be ofpathogenetic significance, we as well as others therefore suggest that Aβ is neurotox-ic at a stage prior to amyloid fibril formation.

13

INTRODUCTION

Alzheimer’s disease is the most frequent neurodegenerative disease affecting individ-uals of all races and ethnic backgrounds. The first patient was reported in 1907 bythe Bavarian neuropsychiatrist and neuropathologist Alois Alzheimer. He describedthe clinical and neuropathological manifestations in a 51-year-old woman whose firstsymptom was paranoid jealousy. Her other symptoms included deficits in memory,paranoid delusions, disorientation, and abnormal behavior. The disease was progres-sive and led to dementia. At neuropathological examination, profuse senile plaquesand neurofibrillary tangles were found in the cerebral cortex (Alzheimer 1907).

Today, AD is emerging as a major health problem because prevalence of AD in-creases with age and the proportion of elderly population is increasing, particularlyin developed societies. Worldwide, about 20 million people suffer from AD (Haasset De Strooper 1999). In the United States 4 million people have Alzheimer’s dis-ease (Evans et al. 1989). In Finland it has been estimated that 95 000 patients willbe suffering from moderate to severe dementia by the year 2030 (Sulkava et al.1986). Approximately 50 to 70% of such patients are diagnosed as having Alzheim-er’s disease according to clinical criteria (Sulkava et al. 1985, Rocca et al. 1986,Breteler et al. 1992). Both the prevalence and the incidence of AD double approxi-mately every 5 years after the age of 60 (Jorm et al. 1987, Cummings et al. 1998).The prevalence of probable Alzheimer’s disease in East Boston was 3.0% in the age-group of those aged 65 to 74 years, 18.7% of those aged 75 to 84, and 47.2% ofthose over 85 (Evans et al. 1989). Annual incidence rates of 2.4 cases per 100,000population between ages 40 and 60, and 127 cases per 100,000 population after age60 have been reported (Rocca et al. 1986).

AD has an insidious onset and manifests with progressive impairment of memoryand of other higher cortical functions. However, the diagnosis of definite AD is neu-ropathological. Classical neuropathological hallmarks of AD include neuriticplaques, intraneuronal neurofibrillary tangles, amyloid angiopathy, and neuronal loss(Mirra et al. 1991, Braak et Braak 1991, Lippa et al. 2000b).

Like many other neurodegenerative diseases, AD occurs in both familial and spo-radic forms. Epidemiological studies indicate that approximately 30% of AD pa-tients have at least one affected first-degree relative (Rosenberg 2000).

The majority of AD cases occur at a late age (>60 years), and show no Mendelian in-heritance. However, familial aggregation of Alzheimer’s disease has been recognizedsince the beginning of the 20th century. A subset of these kindreds have been iden-tified with early-onset AD transmitted as an autosomal dominant disease with age-dependent penetrance. It has been estimated that autosomal dominant AD compris-es about 5% of all AD cases (Lendon et al. 1997, Cruts et al. 1998). The allelic vari-

14

ant ε4 of the apolipoprotein E gene on chromosome 19 is a well-documented riskfactor for familial and sporadic late-onset AD (Saunders et al. 1993, Kuusisto et al.1994, Polvikoski et al. 1995), whereas α-2 macroglobulin (Blacker et al. 1998, Myl-lykangas et al. 1999) on chromosome 12 is a likely risk factor in sporadic late-onsetAD.

Autosomal dominant AD manifests before or during the fifth and sixth decades.It is genetically heterogeneous. At present, three genes have been shown to be in-volved in the etiology, the amyloid precursor protein (APP) gene on chromosome 21(Goate et al. 1991), the presenilin 1 (PS-1) gene on chromosome 14 (Schellenberg etal. 1992, Sherrigton et al. 1995, Clark et al. 1995), and the presenilin 2 (PS-2) geneon chromosome 1 (Levy-Lahad et al. 1995, Rogaev et al. 1995). Close to one hun-dred mutations responsible for autosomal dominant AD have been identified in thesethree genes, and most of these are missense mutations causing single amino acid sub-stitutions. However, defects in exons of these three genes are not responsible for ADin all early-onset families (Cruts and Van Broeckhoven 1998). It is possible that in-tronic mutations of the APP and the presenilin genes or other thus-far unidentifiedgenes exist that lead to autosomal dominant AD (Lippa et al. 2000b). The clinicaland neuropathological features of autosomal dominant AD are usually indistinguish-able from sporadic AD (Campion et al. 1995a, Mann et al. 1992, Lippa et al. 1996).Although autosomal dominant AD is rare, it has proven to be of crucial scientificvalue in elucidating the molecular pathogenesis of AD.

The subject of this thesis is to describe a novel inherited variant of Alzheimer’sdisease identified in a Finnish family. This new variant Alzheimer’s disease (varAD)is associated with a distinctive clinical and neuropathological phenotype caused byan unusual genomic deletion mutation encompassing exon 9 of the presenilin 1 (PS-1) gene and parts of the flanking introns. The pathogenesis of varAD will providenew insights of wider interest in regards to AD in general.

15

REVIEW OF THE LITERATURE

1. SPORADIC ALZHEIMER’S DISEASE

1.1. Clinical features of sporadic AD

1.1.1. Symptoms and signsThe illness remains subclinical for years, and the onset of AD is typically insidious,making it difficult to specify the exact time of onset. Furthermore, gradual progres-sion of symptoms is typical. At the late stage, patients become mute, uncompre-hending, and bedridden (Morris 1996). Duration from onset of the disease to deathis on average from 8 to 10 years (Walsh et al. 1990). According to the report of aWork Group on the Diagnosis of Alzheimer’s Disease established by the National In-stitute of Neurological and Communicative Disorders and Stroke (NINCDS) and theAlzheimer’s Disease and Related Disorders Association (ADRDA), the diagnosis ofpossible and probable AD is clinical (McKhann et al. 1984). Previous exclusion ofother diseases causing cognitive impairment such as delirium, depression, or hy-pothyreosis is needed before establishment of the diagnosis of AD.

The evaluation of a patient with cognitive decline begins by patient and inform-ant history. The salient cognitive feature of AD is memory impairment and mostcommonly it is also the presenting symptom (Morris 1996). Typical symptoms in-clude difficulty recalling names of individuals, telephone numbers, and details ofevents of the day. In addition, repetitive comments are typical. After the history, isa cognitive evaluation often by the Mini-mental state (MMS) test (Folstein et al.1975). At the early stage, the patient has problems in short-term memory, in verbaland nonverbal recognition memory, and a deficit in acquisition of novel information,whereas remote memory is relatively spared. In addition, in the initial mild stage ofAD manifests characteristically in difficulty recalling details of recent events or con-versations, minor temporal and geographic disorientation, difficulties with calcula-tion, and reduced insight and initiative (Morris 1996). Apraxia or incapacity to per-form motor tasks unexplained by a primary motor disturbance is present in up toone-third of patients of the early stage of AD and in an increasing proportion in laterstages (Della Sala et al. 1987, Cummings and Benson 1992).

However, social skills, basic self-care abilities, emotional life and responses aregenerally well-preserved at the early stage of the disease. Clinical examination iscompleted by medical and neurologic examination. Besides memory and visuospatial

16

disturbances, extrapyramidal signs such as rigidity, bradykinesia, shuffling gait, andpostural change are relatively common in AD but, in contrast, primary motor func-tion and cerebellar function are spared (Cummings et al. 1985, Funkenstein et al.1993).

At the moderate stage of AD, memory of recent events is lost and only highlylearned material is recalled. Furthermore, language dysfunction is common, and of-ten both language comprehension and output are compromised (Morris 1996). Thelanguage disorder resembles but is not identical to transcortical sensory aphasia, andas the disease progresses, the language disorder resembles Wernicke’s aphasia (Cum-mings et al. 1985). Behavioral disturbances such as aggression and delusions oftencomplicate the moderate stage of AD, and supervision for daily self-care functions isneeded.

At the late stage of AD, self-care becomes impossible as patients lose their inde-pendence in daily activities, and most patients are institutionalized.

1.1.2. NeuropsychologyIn the clinical and preclinical diagnosis of AD, neuropsychological assessment is animportant tool (Petersen et al. 1994, Linn et al. 1995, Jacobs et al. 1995, Almkvistet al. 1996a, Almkvist et al. 1996b, Diaz-Olavarrieta et al. 1997). In addition, itcan be used in follow-up studies of disease progression and clinical trials. However,the most significant aim of neuropsychological evaluation is to gain facts with re-spect to the presence, magnitude, and nature of cognitive compromise. Memory im-pairment is the first sign of incipient disease and the most prominent, central, andconsistently observed deficit in all types of AD. Importantly, a preclinical stage ofdetectable cognitive deficits can precede the clinical diagnosis of probable AD bymany years (Linn et al. 1995). On the other hand, in different stages of AD, neu-ropsychological findings are variable.

Measures of verbal learning and memory and of immediate auditory attentionspan are among the most sensitive instruments to detect early changes in cognitivefunctioning (Linn et al. 1995). Already at a very early stage of AD, acquisition anduse of semantic cues are disturbed (Petersen et al. 1994). Other neuropsychologicalsymptoms such as impairments in language, praxis, calculation, and behavioral Aβ-normalities show more diversity.

1.1.3. NeurophysiologyThe importance of electroencephalography (EEG) as a routine method of diagnosingAD has diminished, while neuroimaging has become more accurate. At the earlystage of AD, EEG may remain normal. Furthermore, during a one-year follow-up,50% of AD patients have shown normal EEG recordings with no deterioration(Soininen et al. 1989). At more advanced stages of AD most patients have abnormalEEG. These EEG changes include diffuse slowing of the alpha rhythm and appear-

17

ance of theta and delta activity (Penttilä et al. 1985, Helkala et al. 1991).

1.1.4. NeuroimagingBrain imaging by computed tomography (CT) or magnetic resonance (MRI) is gener-ally used in the clinical diagnosis of dementia for two reasons: First, to exclude otherillnesses that are potentially amenable to treatment, and second, to seek evidence asto an imaging procedure that may offer some specific additional diagnostic informa-tion such as medial temporal- or focal atrophy and vascular changes (Scheltens 2001).Although generalized brain atrophy can be detected by CT, in AD, it is of very limit-ed diagnostic value.

MRI has higher sensitivity and can be focused on those areas that are affected ear-liest. Atrophy of the medial temporal lobes, especially atrophy of the entorhinal cor-tex and of the hippocampi and perihippocampal areas are characteristic in AD (Moriet al. 1997, Laakso et al. 1995, Juottonen et al. 1999).

In AD patients, PET has demonstrated bilateral reduction in regional cerebralblood flow (rCBF) and diminished oxygen utilization in temporoparietal and medialtemporal cerebral cortices (Frackowiak et al. 1981, Kumar et al. 1991, Ishii et al.1996, Ishii et al. 1997). The decreasing regional cerebral metabolic rate for glucose(rCMRglc) involves the parietal and temporal lobes that is already detectable in thevery early stage of AD (Julin et al. 1998). Although it is not a typical PET feature,some AD patients show metabolic asymmetry (Friedland et al. 1985). Interestingly,the site of this asymmetry often correlates with clinical manifestation. Patients withgreater left hemisphere hypometabolism tend to demonstrate language disorders,whereas patients with greater right hemisphere hypometabolism show more promi-nent visuospatial disturbances (Haxby et al. 1988, Koss et al. 1985).

1.2. Main neuropathological featuresThe diagnosis of definite AD requires histopathologic confirmation. Of 106 autop-sied patients with a clinical diagnosis of probable AD, 87% had AD according to theCERAD (Consortium to Establish a Registry for Alzheimer’s Disease) neuropatho-logical criteria (Gearing et al. 1995).

End-stages of AD are easily recognized at neuropathological examination. Incontrast, evaluation of cases with mild to moderate affection is more difficult (Braaket Braak 1991). AD is characterized by reduction in brain volume and weight andventricular enlargement (Esiri et al. 1997). Histological hallmarks of AD brains areβ-amyloid plaques, intraneuronal neurofibrillary tangles, and neuronal degenerationand loss, preceded by loss of synaptic density (Lippa et al. 2000b).

1.2.1. PlaquesSenile plaques are lesions composed of diverse amyloid peptides and associated mole-cules, degenerating neuronal processes, and reactive glia. In general, an essential

18

component of senile plaques in AD is the β amyloid peptide (Aβ), a 4 kDa proteinderived from the amyloid precursor protein (APP). Patients with AD usually showseveral types of plaques; the term “senile plaque” is used to cover a number of theirdifferent subtypes.

Senile plaques are stained by different methods. In general, they are difficult tosee in sections stained with hematoxylin or cresyl violet but are easily demonstratedthe silver impregnation method described by Bielschowsky or the methenamine sil-ver technique. Congo red PAS, or thioflavin S, in particular, are the best techniquesfor demonstrating the amyloid component.

Several classifications of plaques have been published. However, a recent author-itative article lists three main types of plaques (Dickson 1997): Plaques with exten-sive or poorly circumscribed Aβ deposits are categorized as “diffuse plaques”. Theseare composed of amorphous aggregates of Aβ protein and contain at most, only a fewamyloid fibrils. They lack dense amyloid cores and are devoid of pathological neur-ites. Diffuse plaques in the cerebellar molecular layer and striatum may be associatedwith ubiquitin-immunoreactive dystrophic neurites. However, diffuse plaques arenot identifiable in sections stained by hematoxylin-eosin but can be visualized bymethenamin silver staining and are immunoreactive for Aβ peptide.

Plaques with prominent neuritic elements are referred to as “neuritic plaques”.The typical neuritic plaque is a spherical structure 50 to 200 µm in diameter, con-sisting of a central immunoreactive amyloid core surrounded by dystrophic neurites(Cummings et al. 1998). These neurites are heterogeneous. Some neurites are ofdystrophic type and contain degenerating synaptic elements with accumulation ofmembranous material and lysosomal dense bodies, while others are PHF-type neur-ites containing paired helical filaments (Dickson 1997). PHF-type neuritic degener-ation correlates well with cognitive dysfunction (Dickson 1997). Dystrophic neur-ites are the major type seen in aged human brains that can be detected by differentstaining methods. Methods exits to detect lysosomal enzyme activity (acid phos-phatase) or molecules (cathepsin) that readily stain dystrophic neurites, while othermarkers used in detecting dystrophic neurites recognize specific glycosyl groups(concanavalin A) or components of the degenerating synaptic terminals (antibodies tochromogranin A). In addition, antibodies to ubiquitin can detect dystrophic neur-ites (Dickson 1997).

Plaques composed exclusively of amyloid are classified as “amyloid plaques”(Dickson 1997). At an early stage of AD, plaques are found in isolated areas of theassociation cortex of the frontal and anterior temporal lobes and in the amygdala, theentorhinal cortex, and the hippocampus. Neuritic plaques tend to occur in deeperlayers of the cortex, whereas diffuse plaques being predominant in the more superfi-cial layers (Esiri et al. 1997).

In addition to the three main types of plaques, a fourth type has been describedin the hippocampal region of AD patients and in individuals of advanced age. This

19

type has been called “plaque A” (Probst et al. 1991). This non-congophilic (pre-amyloid) plaque A is spherical in shape and 100-200 µm in diameter, with sharplyoutlined borders. Its main feature is a single centrally located cell, which stains withRCA-1 lectin and can therefore be considered a microglial cell (Probst et al. 1991).

1.2.2. Neurofibrillary changesAD is associated with severe cytoskeletal alterations of various kinds of lesions,namely neuritic plaques and neurofibrillary tangles, and neurophil threads (Braakand Braak 1991). Neuritic plaques are marked by a dense feltwork of argyrophilicnerve cell processes, neurofibrillary tangles are composed of bundles of paired helicalfilaments containing the microtubule-associated protein tau. These neurofibrillarytangles are present intraneuronally in AD and are composed predominantly of hyper-phosphorylated tau. They develop within the nerve cell soma from whence they mayextend into the dendrites, while the proximal axon remains free of any (Braak andBraak 1991). Neurofibrillary tangles also occur in non-AD diseases and therefore areless specific to AD than neuritic plaques (Cummings et al. 1998). Neurofibrillarytangles consist of argyrophilic processes of nerve cells loosely scattered throughoutthe neurophil, and in the isocortex they are localized in dendrites of tangle-bearingpyramidal cells (Braak and Braak 1991).

1.2.3. Amyloid angiopathyIn AD, Aβ deposits not only in the cerebral parenchyma of the amygdala, hippocam-pus, and neocortex, but also in the walls of leptomeningeal and intracortical bloodvessels, resulting in amyloid angiopathy. Vascular amyloid stains with Congo red.The amyloid in cerebral blood vessels most commonly consists of Aβ1-40 (Prelli etal. 1988) and is usually deposited in the walls of small arteries and arterioles. Amy-loid angiopathy is associated with increased risk for intracerebral hemorrhage (Gil-bert et al. 1983).

1.3. Genetic risk factors for Alzheimer’s disease

1.3.1. Apolipoprotein EGenetic linkage studies in pedigrees with predominantly late-onset familially aggre-gated AD have provided evidence for AD susceptibility locus which maps to chro-mosome 19q12-q13 (Pericak-Vance et al. 1991).

Two years later it was shown that the ApoE ε4-allele is a significant genetic riskfactor not only in familial (Strittmatter et al. 1993) but also in sporadic late-onsetAD (Saunders et al. 1993). Apolipoprotein E (ApoE) is a glycoprotein of approxi-mately 34 kDa which plays a critical role in lipid transport, homeostasis, and catabo-lism (Mahley 1988, Davignon et al. 1988). At the single ApoE locus there are threealleles: ε2, ε3, and ε4, corresponding to six phenotypes (Rall et al. 1982, Mahley

20

1988). Elevated risk for late-onset AD among individuals with the ε4-allele hasbeen confirmed in many studies (Poirier et al. 1993, Saunders et al. 1993, Kuusistoet al. 1994, Farrer et al. 1997). Meta-analysis of 40 studies representing nearly 30000 ApoE alleles demonstrated that the ε4-allele is a major risk factor for sporadicAD in Caucasians, African-Americans, Hispanics, and Japanese, and in both men andwomen (Farrer et al. 1997). Caucasian individuals with the ε2/ε4, ε3/ε4, and ε4/ε4genotypes have an increased risk for late-onset AD (Farrer et al. 1997, Evans et al.1997). In addition, a dose-dependent relationship exists between the number of cop-ies of ε4, and the age of onset of AD so that subjects with ε4/ε4 genotype have anearlier onset than do heterozygous ε4 subjects and have a higher plaque burden thando subjects with no ε4 alleles (Corder et al. 1993, Strittmatter et al. 1993, Polviko-ski et al. 1995). It has been suggested that the effect of the ApoE ε4 allele on therisk of AD is age-dependent, and is no longer present in very old age (Blacker et al.1997, Juva et al. 2000), but recent studies show that the ApoE ε4 effect is evident atall ages between 40 and 90 years (Farrer et al. 1997, Polvikoski et al. 2001).

The ApoE ε2 allele appears to be protective, both by lowering risk for AD andincreasing the age of onset (Saunders 2000). Consequently, the risk decreases forpeople with genotypes ε2/ε2 and ε2/ε3, and the ε2/ε3 genotype appears equally pro-tective across different ethnic groups (Farrer et al. 1997).

The mechanisms of ApoE in relation to AD are insufficiently understood. ApoEplays a significant role in the peripheral and central nervous system in regenerationafter injury as well as in normal brain lipid metabolism (Strittmatter et al. 1993).On the other hand, ε4 shows increased affinity for Aβ and is found in senile plaquesin AD, thus acting as a pathological chaperone (Strittmatter et al. 1993). Althoughinheritance of the ApoE ε4 allele confers increased risk for AD, it is neither sufficientnor necessary to cause the disease.

1.3.2. OthersLate-onset AD is the most common form of the disease, but 30 to 40% of patientswith late-onset sporadic AD do not carry the ApoE ε4 allele (Roses 1995). For thesereasons, the search for new genetic risk factors in different ethnic groups has beenvigorous. Statistical evidence has been presented for a susceptibility locus on chro-mosome 12 in a complete genome screen in familial AD (FAD) (Pericak-Vance et al.1997). In addition, other evidence is suggestive of linkage to an overlapping regionon chromosome 12 (Rogaeva et al. 1998). Alpha-2-macroglobulin (A2M), encodedby the A2M gene on chromosome 12, has been reported to bind tightly to the β-amyloid protein which is the major component of amyloid plaques (Du et al. 1997).Furthermore, A2M is a serum pan-protease inhibitor implicated in AD, based on itsability to mediate the clearance and degradation of Aβ (Blacker et al. 1998). Previ-ous studies have suggested an association between the alpha-2 macroglobulin andlate onset-Alzheimer’s disease (Blacker et al. 1998, Myllykangas et al. 1999). How-ever, several other studies show controversial results (Wu et al. 1998, Gibson et al.

21

2000, Sodeyama et al. 2000, ). In addition, linkage studies have suggested that ge-netic risk factors for late-onset AD may be located on chromosomes 3, 5, and partic-ularly on chromosome 10 (Tanzi 1996, Kehoe et al. 1999, Bertram et al. 2000, Mye-rs et al. 2000, Ertekin-Taner et al. 2000). Several other risk genes for AD have beensuggested but not constantly confirmed.

2. AUTOSOMAL DOMINANT ALZHEIMER’S DISEASE

2.1. GeneticsThe majority of AD cases occur at a late age and do not show Mendelian inheritance.However, epidemiological studies indicate that approximately 25 to 40% of AD pa-tients have at least one affected first-degree relative (Van Broeckhoven et al. 1995,Rosenberg 2000). A subset of these kindreds have been identified with early-onsetAD transmitted as an autosomal dominant trait with age-dependent penetrance.

Autosomal dominant AD manifests before or during the fifth and sixth decadeand comprises about 5 to 10% of all AD cases (Lendon et al. 1997, Cruts et al.1998). It is genetically heterogeneous.

At present, three genes have been shown to be involved in AD etiology: theamyloid precursor protein (APP) gene on chromosome 21 (Goate et al. 1991), thepresenilin 1 (PS-1) gene on chromosome 14 (Schellenberg et al. 1992, Sherrigton etal. 1995, Clark et al. 1995) and the presenilin 2 (PS-2) gene on chromosome 1 (Levy-Lahad et al. 1995, Rogaev et al. 1995). In these three genes have been identifiedabout 100 mutations responsible for autosomal dominant AD. However, defects inthese three genes are not responsible for AD in all early-onset families (Cruts andVan Broeckhoven 1998). Autosomal dominant AD is rare but has proven to be ofimportant scientific value in elucidating the molecular pathogenesis of AD.

2.1.1. Amyloid precursor protein (APP) gene mutationsAPP gene was the first gene to be identified in association with inherited suscepti-bility to AD, and the first point mutation of the APP gene was found at codon 717(Goate et al. 1991). This discovery was preceded by the observation that Down syn-drome (trisomy 21) is associated with augmented deposition of Aβ, and virtually allindividuals with Down syndrome who live to the age of 40 develop the histopatho-logical manifestation of AD (Mann and Esiri 1989). The early neuropathologicalchanges are apparently due to increased gene dosage and overproduction of β-amy-loid, which as in other APP mutations is a likely underlying mechanism of the ADhistopathology in Down syndrome (Lannfelt et al. 1994).

APP is a type I, integral membrane glycoprotein encoded by a gene on chromo-some 21q21.2. (Yu et al. 1992) The APP protein undergoes a series of endoproteo-

22

lytic cleavages and is metabolized via one of at least three pathways. One of these oc-curs in part in the endosomal-lysosomal compartment and involves the putative β-and δ-secretases which give rise to a series of peptides containing the 40-42 aminoacid Aβ peptide. Furthermore, Aβ42 and Aβ43 are thought to be more fibrillogenicand more neurotoxic than the Aβ40, that is the predominant isoform produced dur-ing the normal metabolism of APP (Yankner et al. 1990).

Although the normal function of APP is unknown, mutations have been shownto alter endoproteolysis of APP such that more of the Aβ42 form is produced (Hardy1997, Citron et al. 1992). In addition, FAD mutations in APP may also alter the in-tracellular trafficking and overall maturation or processing of APP (Tanzi et al.1996).

In general, mutations of APP gene are rare. Molecular genetic studies haveshown that APP mutations are responsible for only approximately 5% to 7% ofearly-onset FAD (Van Broeckhoven 1995, Tanzi et al. 1996). At present, at leastseven mutations at eight codons of the APP gene have been identified in over 20families (Goate et al. 1991, Hendricks et al. 1992, Mullan et al. 1992, Tanzi et al.1996, Cruts and Van Broeckhoven 1998).

These mutations have been shown to be nearly 100% penetrant (Tanzi et al.1996). The mutations are clustered within and around the amyloidogenic region(Tanzi et al. 1996, Cruts and Van Broeckhoven 1998). Mutations at codon 717 arepresent in several families from different ethnic groups, but the mutations at codon692 and 670/671 are rare (Goate et al. 1991, Karlinsky et al. 1992, Mullan et al.1992, Hendricks et al. 1992). A double mutation in APP at codons 670/671 hasbeen shown to cosegregate with the disease in two large Swedish pedigrees (Mullanet al. 1992, Lannfelt et al. 1994, Almkvist et al. 1995). Mutations in codons 692and 693 of the APP gene lead to presenile dementia and cerebral hemorrhages (Levyet al. 1990, Van Broeckhoven et al. 1990, Hendricks et al. 1992). Patients with he-reditary cerebral hemorrhage with amyloidosis of the Dutch type (HCHWA-D) havea mutation in codon 693 in exon 16 of the APP gene resulting in a glutamate-to-glutamine substitution (Levy et al. 1990). HCHWA-D is associated with a deposi-tion of Aβ in cerebral blood vessels causing recurrent cerebral hemorrhages. A muta-tion in which alanine is substituted for the glycine of the neighboring codon 692causes presenile dementia and cerebral hemorrhage (Hendriks et al. 1992).

2.1.2. Presenilin 1 (PS-1) gene mutationsAfter the discovery that APP mutations were a rare cause of early onset FAD, molec-ular genetic linkage studies indicated a locus on chromosome 14 (Schellenberg et al.1992). Soon it was shown that the presenilin-1 (PS-1) gene located on chromosome14q24.3 is the major gene responsible for early-onset autosomal dominant forms ofFAD (Sherrington et al. 1995, Clark et al. 1995, Campion et al. 1995a, Poorkaj et al.1998).

23

The PS-1 gene encodes a 467 amino acid membrane serpentine protein witheight transmembrane domains and two hydrophilic acidic domains, one at the N-ter-minus (Hutton et al. 1997, De Strooper et al. 1998, Haass et De Strooper 1999).Presenilins are expressed in neurons throughout the central nervous system and alsoin other organs and are localized in intracellular membranes and mainly in endoplas-mic reticulum (De Strooper et al. 1998). Immunoblotting studies suggest that thePS-1 protein is about 50 kDa in size (De Strooper et al. 1997). At genomic level thisgene comprises 12 exons spanning approximately 75 kb, with the open readingframe limited to exons 3 to12 and spanning approximately 24 kb (Clark et al. 1995,Tanzi et al. 1996). Approximately 75% of known PS-1 mutations occur in exons 5,7, and 8 (Smith et al. 1999). Most PS-1 gene mutations are missense mutationscaused by a single amino-acid substitution. In 1995 a different kind of mutationwas reported (Perez-Tur et al. 1995). This mutation in a British kindred involves apoint mutation in the splice acceptor site at the 5’ end of exon 10 (Perez-Tur et al.1995). Two years later, deletion of exon 9 of the PS-1 gene was identified in an Aus-tralian family with dementia and spastic paraparesis (Kwok et al. 1997). The thirdsplicing defect mutation was reported in a Japanese kindred (Sato et al. 1998).

Estimates of the proportion of early-onset FAD due to mutations in the PS-1gene vary from 18% to 70% (Campion et al. 1995b, Hutton et al. 1997, Cruts et al.1998). To date, over 60 mutations of PS-1 have been discovered, and have beenidentified all over the world in more than 80 families of different ethnic origins(Tanzi et al. 1996, Cruts et al. 1998, Rosenberg 2000 ).

Mutations in the PS-1 gene result in the most severe form of the disease with au-tosomal dominant inheritance, complete penetrance, and onset occurring as early as30 years of age (Poorkaj et al. 1998). The range of onset for PS-1 gene mutations isbetween 30 to 65 years (Campion et al. 1996, Lendon et al. 1997, Kwok et al. 1997,Poorkaj et al.1998, Rosenberg 2000). The onset is earlier and the disease durationshorter than in sporadic patients or patients with the APP gene mutation (V717I)(Lippa et al. 1996, Russo et al. 2000).

2.1.3. Presenilin 2 (PS-2) gene mutationsSoon after the discovery of the PS-1 gene underlying AD, a homologous gene calledthe PS-2 gene was identified on chromosome 1q31-q42 (Levy-Lahad et al. 1995, Ro-gaev et al. 1995). The amino acid sequence identity between PS-1 and PS-2 proteinsis approximately 67% (Tanzi et al. 1996). However, the pattern of expression of thePS-2 gene is somewhat different from that of PS-1. PS-2 is expressed maximally incardiac muscle, skeletal muscle, and the pancreas but less homogeneously in thebrain and in the peripheral tissues (Rogaev et al. 1995). In comparing the frequencyof PS-1 gene mutations with PS-2 gene mutations in a large set of patients, it wasobvious that mutations of the PS-2 gene are rare (Sherrington et al. 1996). Atpresent, only three missense mutations have been identified (Levy-Lahad et al. 1995,

24

Rogaev et al. 1995, Cruts et al. 1998). The first mutation (N141I) was detected inpatients of families of Volga German ancestry (Levy-Lahad et al. 1995) and the sec-ond mutation (M239V) in an Italian family (Rogaev et al. 1995), while the third(R62H) was observed in a sporadic Dutch AD case with an onset of 62 years (Crutset al. 1998).

It is however, possible that PS-2 families have been overlooked because of thewide variation in age of onset and because of incomplete penetrance of the PS-2 genemutations (Cruts et al. 1998). In general, the age of onset of AD in individuals withPS-2 gene mutations is higher than in individuals with PS-1 mutations, the age-range being between 40 and 88 years (Lendon et al. 1997, Sherrington et al. 1996).The underlying mechanism for this variation is unsolved. In addition, death occurslater in the PS-2 AD group than in the PS-1 or APP AD groups (Lippa et al. 2000b).In contrast to PS-1 and APP gene mutations, there is at least one instance of non-penetrance in an asymptomatic individual over 80 years whose offspring was affected(Bird et el. 1989).

2.1.4. Autosomal dominant Alzheimer’s disease and ApoEPrevious studies suggest that the age of onset of FAD is modulated by the ApoE gen-otype (Blacker et al. 1996). Consequently, carriers of the APP (V717I) mutationhaving one or more ε4 alleles at ApoE have an earlier onset of the disease than do car-riers who have no ε4 alleles of ApoE (St George-Hyslop et al. 1994).

However, the relationship between presenilin mutations and apolipoprotein Ehas remained to some extent controversial. Most studies show that ApoE does notmodulate the age of onset in families with a mutation in the PS-1 gene (Van Broeck-hoven et al. 1994, Kwok et al. 1997). Furthermore, evidence exists that the ApoEgenotype does not influence neuropathologic features in patients with FAD withlinkage to chromosome 14 or in patients with FAD with linkage to chromosome 21(Lippa et al. 1996). These observations suggest the existence of other genetic and en-vironmental factors modifying the AD phenotype in PS-1 families.

2.2. Clinical featuresIn contrast to the marked genetic heterogeneity, distinctive clinical features are rare-ly reported in FAD. Before molecular genetic studies were available, the main fea-tures of FAD were suggested to be earlier age of onset with more rapid course, great-er prevalence of language disturbances, seizures, myoclonic jerks, and shorter survivalthan in sporadic AD (Bird et al. 1989, Kennedy et al. 1995, Swearer et al. 1996).Nowadays, specific mutations can be related to characteristic clinical features.

Most reported patients with PS-1 mutations have onset of symptoms before age50 years (Van Broeckhoven 1995), while the age-range for PS-2 gene mutations isbetween 40 and 88 (Lendon et al. 1997). Early progressive aphasia, myoclonus, gen-eralized seizures, and paratonia are apparently more common in PS-1-encoded FAD

25

than in patients with APP (codon 717) mutations (Lampe et al. 1994).Those affected persons with the PS-2 gene mutation in the Volga German

kindreds are reported to have generalized seizures, rigidity, tremor, myoclonus, andlanguage deficits (Bird et al. 1989).

In addition, a few reports describe phenotypic heterogeneity between differentmutations of the same gene. Mutations of the APP gene that lie on either end of theβ-peptide (Aβ) sequence cause a typical presenile FAD. However, a mutation of co-don 693 of the APP gene within the Aβ sequence causes hereditary cerebral hemor-rhage of the Dutch type (Levy et al. 1990), whereas a mutation of codon 692 leads topresenile dementia and cerebral hemorrhage (Hendriks et al. 1992).

Mutations of the PS-1 are the most common of autosomal dominant FAD, andthe majority of over 60 mutations are associated with typical clinical features of AD.However, there are a few exceptions: headache is a distinctive symptom in FAD, dueto the E280A mutation of the PS-1 gene (Lopera et al. 1997); early and prominentmyoclonus mimicking Creutzfeldt-Jakob disease is associated with the M146V mu-tation of the same gene (Haltia et al. 1994, Clark et al. 1995).

2.3. Neuropsychological featuresKindreds with specific mutations provide an opportunity to study neuropsychologi-cal features related to certain mutations. However, surprisingly few studies with alimited number of patients provide neuropsychological data on FAD in associationwith specific mutations.

Patients with a missense substitution of isoleucine for valine at codon 717 of theAPP gene show early neuropsychological manifestations such as deficits in memory,in cognitive processing speed, and in attention to complex cognitive sets (Karlinskyet al. 1992). Substitution of phenylalanine for valine at codon 717 of the APP geneis associated with a neuropsychological profile in which recent memory, information-processing speed, sequential tracking, and conceptual reasoning are the earliest cog-nitive functions affected. However, at least early in the course of the disease, lan-guage and visuoperceptual skills are largely spared (Farlow et al. 1994).

In general, APP mutations are associated with disturbance in attention and withpsychiatric symptoms (Matsumura et al. 1996, Almkvist et al. 1995), while languagefunctions are usually preserved.

PS-1 mutations are more often associated with language impairment (Campionet al. 1995a, Kennedy et al. 1995, Fox et al. 1997, Roselli et al. 2000). Chromo-some 14- linked FAD was initially associated with a neuropsychological profile char-acterized by early memory deficit with early dyscalculia and an impairment in speechproduction. Later this mutation was established as a point mutation M139V in thepresenilin 1 gene (Kennedy et al. 1995, Fox et al. 1997). Further neuropsychologicalevaluation disclosed a sequence of deficits with early memory loss initially selectivefor verbal memory in some individuals, followed by loss of arithmetic skills, whereasnaming and object-perception skills were relatively preserved. A speech productiondeficit characterized by stammering was seen in all surviving affected members of a

26

British family (Fox et al. 1997).

2.4. Neuropathological featuresApart from the more intense pathology, more numerous lesions, and a faster diseaseprocess, neuropathological findings in FAD overlap with and are similar to those ofsporadic AD (Lippa et al. 1996, Lippa et al. 2000a).

Actually, it is surprising that mutations of different genes of chromosomes 1,14, and 21 produce exactly the same neuropathological phenotype. It is possiblethat slight pathological differences may not have been recognized in earlier studies.However, careful study of neuropathological features of FAD sometimes shows subtledifferences in features among AD groups.

Mutations at codon 717 of the APP gene result in classic but severe neuropatho-logical features typical of AD without unusual characteristics (Mann et al. 1992,Karlinsky et al. 1992). However, two mutations of the APP gene cause more exten-sive angiopathy than that seen in sporadic AD. A mutation at codon 693 of the APPgene, causing E693Q substitution, leads to hereditary cerebral hemorrhage withamyloidosis of the Dutch type (HCHWA-D) without mature neuritic plaques (Levyet al. 1990). A mutation at the neighboring codon 692 causing A692G substitutionleads to HCHWA of the Flemish type (Hendriks et al. 1992). Patients withHCHWA of the Flemish type also develop dementia, and their neuropathologyshows senile plaques with large Aβ cores surrounded by a meshwork of dystrophicneurites (Cras et al. 1998).

FAD associated with the PS-1 gene mutations is, in general, pathologically indis-tinguishable from sporadic AD, apart from exceptionally florid amyloid and neurofi-brillary pathology (Lippa et al. 1996, Campion et al. 1995a, Fox et al. 1997, Smithet al. 1999). However, careful analysis reveals that there is scarce pathological diver-sity caused by PS-1 gene mutations. Neuropathological analysis of patients with theM146V mutation of the PS-1 gene showed mild vacuolar changes resembling spong-iform changes seen in Creutzfeldt-Jakob disease (CJD). In contrast to CJD, thesevacuolar changes were restricted to the medial temporal cortex and amygdala (Haltiaet al. 1994). Furthermore, in a kindred with the E280A mutation, the entire cere-bral cortex showed prominent vacuolation of neurophils in superficial layers, accom-panied by the severe gliosis often seen in CJD (Lopera et al. 1997). Mutation A260Vwas associated with discrete silver- and ubiquitin-positive inclusions resembling Pickbodies in the neurons of the dentate gyrus (Ikeda et al. 1996). Histopathology ofL250S FAD showed a large number of senile and diffuse plaques in the cerebellum.This mutation was associated with large numbers of small, diffuse “cloud-like”plaques in the region of the nucleus basalis of Meynert. There was a selective distri-bution of these plaques and neurofibrillary tangles in the deep gray nuclei. Theputamen contained many senile plaques which contrasted with a paucity of senileplaques in the globus pallidus. Neurofibrillary tangles were plentiful in the midline

27

nuclei of the thalamus, but were scarce in the dorsomedial nucleus (Harvey etal.1998).

Very few reports concern neuropathology associated with PS-2 mutations, proba-bly because of the rarity of mutations in this gene. Neuropathological analysis ofpatients with a PS-2 gene mutation in the Volga German kindred showed typicalAD changes with amyloid angiopathy, but, surprisingly, there were neuritic plaquesin the cerebellum (Bird et al. 1989).

3. HYPOTHESES FOR THE PATHOGENESIS OF ALZHEIMER’S

DISEASE

3.1. Amyloid cascade hypothesis

The amyloid cascade hypothesis of AD is based on the idea that the primary cause ofthe neurodegenerative process in AD is abnormal generation or deposition of Aβ, ei-ther by direct neurotoxic effect or by the induction of associated toxic factors (Hardyet Higgins 1992). Aβ is derived from amyloid precursor protein (APP), a complextransmembrane protein with several different isoforms, which in its longest isoformis a single transmembrane-spanning polypeptide of about 700 amino acids (Hardy etDuff 1993, Tanzi et al. 1996). APP is processed by several alternative metabolicpathways. The major pathways are endosomal-lysosomal and secretory. Three en-zymes: α-, β-, and γ-secretase, cleave APP into Aβ fragments of different sizes. Thesecretory pathway (α-route) for APP begins with cleavage by an α-secretase whichcuts within the Aβ domain between residues 16 and 17. This cleavage produces acarboxy terminal fragment (C-100) which is considered to be irrelevant to amyloido-sis because it lacks the first 16 amino acid residues of Aβ. However, Aβ in AD is de-rived from APP metabolism via the endosomal-lysosomal pathway. This pathway in-volves the putative β- and γ-secretases which give rise to a series of peptides contain-ing the 40-42 amino acid Aβ peptide. AD brains have demonstrated the major Aβpeptides to be Aβ 1-40 and Aβ 1-42, Aβ42 and Aβ43 are considered particularly fi-brillogenic and neurotoxic (Yankner et al. 1990). This endosomal-lysosomal pathway(β-route) results in the cleavage event of the liberation of Aβ. In the first phase ofthe β-route, β-secretase cleaves the APP molecule at the amino terminus of Aβ.Then a γ-secretase cleaves the molecule at the carboxy terminus, releasing the Aβfragment, of which approximately one-fifth is in 42-amino-acid form (Cummings etal. 1998, Hardy et Higgins 1992). However, it is unclear whether Aβ has directneurotoxity or if neurotoxity is due to induction of a secondary mechanism such asinflammatory changes (Tanzi et al. 1996, Selkoe 1997). The gradual cerebral build-

28

up of Aβ first in soluble and then in particulate form, especially in plaques, appearsto result in inflammatory changes. This in turn results in local microglial and astro-cytic activation, with concomitant release of cytokines and acute-phase proteins (Mc-Geer et al. 1994, Selkoe 1997). Secondly, it has been suggested that filamentous Aβalters the glia and neurons by causing changes in calcium homeostasis as well ascausing oxidative injury by free radical formation (Selkoe 1997, Behl et al. 1994,Cummings 1998).

3.2. Other hypothesesNeurofibrillary tangles, which contain paired helical filaments (PHFs) and straightfilaments, are characteristic of AD pathology. PHFs are formed primarily from ab-normal aggregation of tau proteins. The second most important theory of AD em-phasizes the view that abnormal phosphorylation of tau proteins play an importantrole in the pathophysiology of AD (Grundke-Iqbal et al. 1986) possibly because ofimpaired intracellular transport. However, it has been shown that specific mutationof the tau gene on chromosome 17q21-22 is causal for frontotemporal dementia andnot for AD (Hutton et al. 1998, Baker et al. 2000). In addition, other hypothesis ofAD pathophysiology include theories of an autoimmune etiology, inflammatoryprocesses, and oxidative stress (Cummings et al. 1998).

3.3. Function and dysfunction of the presenilinsThe actual biological roles of PS proteins or the mechanisms by which they causeFAD are not fully understood. However, mutant presenilins interfere with the nor-mal processing of APP. Mutations in the PS-1 gene, PS-2 gene, and APP genes havebeen shown to alter APP processing so that more of the peptide Aβ42 or Aβ43 isproduced (Haass et DeStrooper 1999). Long-tailed Aβ is thought to be more neuro-toxic and fibrillogenic than the Aβ ending at residue 40, which is the most commonisoform produced during normal metabolism of APP (Yankner et al. 1990).

Because presenilins are required for the γ-secretase activity, it has therefore beenproposed that presenilins either modulate the activity of γ-secretase or both β-and γ-secretase, or that PS-1 itself is γ-secretase ( Haas and De Strooper 1999, Russo et al.2000).

In addition, the presenilins appear to control the proteolysis of the membrane-associated proteins of integral membrane domains of APP, of Notch, of the mamma-lian APP homolog of amyloid precursor-like protein 1 (APLP-1) and of the tyrosinekinase receptor (TrkB) (De Strooper et al 1998, DeStrooper et al. 1999, Naruse et al.1998 ).

29

4. FAMILIAL DISEASES CAUSING DEMENTIA AND SPASTIC

PARAPARESIS

4.1. Prion diseases

4.1.1. Familial Creutzfeldt-Jakob disease (FCJD)Clinical features in Creutzfeldt-Jakob disease (CJD) are often preceded by a prodomalperiod of non-specific features such as fatigue and sleep disturbance. However, themost characteristic features include rapid mental deterioration, myoclonus, motordisturbance (extrapyramidal, cerebellar, pyramidal, and/or anterior horn cell), and anelectroencephalogram (EEG) showing periodic short-wave activity (DeArmond etPrusiner 1998). The peak age of onset is about 60 in sporadic CJD, while in FCJDthe onset is often earlier, about 46 to 55 (Brown et al. 1991).

4.1.2. Gerstmann-Sträussler-Scheinker diseaseGertmann-Sträussler-Scheinker disease (GSS) is an autosomal dominant disordercaused by certain mutations of the prion protein gene. Codon 105 mutation (P105L)is associated with a clinical picture of spastic paraparesis (Ghetti et al. 1995). Pa-tients become affected in their fourth or fifth decades. The clinical course is long(9±3 years), progressing from paraparesis to quadriparesis, loss of emotional control,and dementia. In contrast to most other prion diseases, in GSS the EEG remainsnormal, and neuropathological analysis shows no spongiform changes (Ghetti et al.1995).

4.2. Hereditary spastic paraplegiaHereditary spastic paraplegia (HSP) comprises a clinically and genetically diversegroup of disorders characterized by insidiously progressive lower extremity weaknessand spasticity. HSP is also known as familial spastic paraparesis and Strumpell-Lor-rain syndrome (MIM *18260; Fink et al. 1996). The prevalence of HSP in Europevaries in different studies from 2.0 to 9.6/100 000 (McDermott et al. 2000). Al-though there are patients with HSP in Finland, the exact prevalence has not beenstudied. Based on their clinical features, patients with HSP are divided into twogroups, depending on whether the disorder is a pure spastic paraplegia (uncomplicat-ed HSP) or a more complex syndrome with other codominating findings (complicat-ed HSP) (Harding 1981). Uncomplicated HSP is limited to spastic gait, urinarytract symptoms, lower extremity hyperreflexia, clonus, extensor plantar responses,and often mildly impaired vibration sensation in the distal lower extremities withlittle or no involvement of the upper limbs (Harding 1981, Heinzlef et al. 1998,

30

Scheltens et al. 1990, McDermott 2000, Fink 2000). The major neuropathologicalfeature in HSP is axonal degeneration that is maximal in terminal portions of thelongest descending and ascending tracts within the spinal cord (Behan and Maia1974). The most severely affected pathways are the crossed and uncrossed corticospi-nal tracts going to the lower extremities and the fasciculus gracilis fibers comingfrom the lower extremities (Harding et al. 1981, McDermott et al. 2000). These de-generating axons are the longest axons in the central nervous system (Fink et al.1996). In addition, involvement of spinocerebellar tracts is seen in about 50% ofcases (McDermott et al. 2000).

In complicated HSP, other abnormalities may include optic atrophy, retinopathy,extrapyramidal disturbance, supranuclear gaze palsy, cerebellar signs, epilepsy, ichty-osis, mental retardation, and deafness (Cross and McKusick 1967, Fink et al. 1996,McDermott et al. 2000). The psychiatric symptoms associated with complicatedHSP are mental retardation, presenile dementia, and hypomania (Skre 1974, Primoreet al. 1995). Complicated HSP with dementia is probably not as rare as earlierthought. Recent studies have suggested that this kind of HSP is linked to chromo-some 2p and is associated with a subcortical type of dementia (Webb et al. 1998,McDermott et al. 2000). In a chromosome 2p21-24-linked family the patients de-veloped late onset dementia, and autopsy confirmed specific cortical pathology, maxi-mal in the hippocampus and medial temporal cortex. There were neuronal depletionand tau-immunoreactive neurofibrillary tangles in the hippocampus and frequentballoon cells seen in the limbic system and neocortex (White et al. 2000).

HSP may be transmitted in an autosomal dominant, autosomal recessive, or X-linked recessive manner (Behan and Maia 1974, Fink 2000). At present, mutationsin four genes: L1CAM, PLP, paraplegin, and spastin have been identified to underliean HSP phenotype, while the genes at seven other loci remain unknown (McDermottet al. 2000). Inheritance of HSP is most commonly autosomal dominant, and thephenotype has been linked to loci on chromosome 2p, 14q, 15q, 8q, 10q, and 12q,but not all families map to these loci (Hazan et al. 1993, Fink et al. 1995,McDermott et al. 2000). The most common locus of autosomal dominant HSP ischromosome 2 (Webb et al. 1998, McDermott et al. 2000). Recently mutations in agene called spastin has been identified in uncomplicated HSP and localized onchromosome 2p22-p21 (Hazan et al. 1999).

Genetic loci for autosomal recessive uncomplicated HSPs have been identifiedon chromosomes 8p11-8q13 and 15q13-q15 (Fink 2000). Mutations in a genecalled paraplegin on chromosome 16q24.3 have been found in autosomal recessivecomplicated HSP (McDermott et al. 2000). X-linked HSP is rare. Families withcomplicated HSP have been linked to Xq28 and Xq22, whereas uncomplicated HSPhas been linked to Xq22 (McDermott et al. 2000).

4.3. Familial British dementia (FBD)Familial British dementia, previously known as familial cerebral amyloid angiopathy

31

of British type, is clinically characterized by its autosomal dominant mode of inherit-ance, progressive spastic tetraparesis, cerebellar ataxia, and dementia, with onset inthe sixth decade (Worster-Drought et al. 1940, Plant et al. 1990, Mead et al. 2000).The first symptom in some patients is paralysis, whereas other patients only have de-mentia (Worster-Drought et al. 1940). In addition, cerebellar signs are typical. Pa-tients have, infrequent, indistinct, and progressively slurred speech. Less constantclinical features include headaches, occasional stroke-like episodes, personalitychanges and “blackouts” (Plant et al. 1990). The mean age of onset of symptoms is47 years (range from 40 to 57) and the duration of the disease ranges from 2 to 18(Plant et al. 1990).

Extensive cerebral periventricular white matter changes visualized in MRI aretypical for FBD. White matter changes are also seen in the brainstem and cerebel-lum (Plant et al. 1990). In neuropathological examination there is severe, extensiveamyloid angiopathy with deposition of amyloid in small cerebral and spinal arteriesand arterioles. Three different types of plaques have been described. Classified aslarge, small, and perivascular (Plant et al. 1990, Revesz et al. 1999). Large plaquesappear as large globular masses, often with a central congophilic star-shaped core andperipheral radiating fine spicules or coarser processes. Small plaques resemble thecentral core of the large plaques, but are naked and show no peripheral pale zone.Perivascular plaques appear as pale, homogenous, or fibrillated structures aroundcapillaries or arterioles. They are faintly PAS-positive, but not uncommonly have astrongly congophilic core. Light and electron microscopic examination shows thathippocampal amyloid plaques are primarily of the “nonneuritic” type (Plant et al.1990, Revesz et al. 1999). These nonneuritic amyloid plaques also affect the cerebel-lum (Plant et al. 1990). In addition, the presence of numerous neurofibrillary tan-gles (NFTs) has been reported (Plant et al. 1990).

Neuropathological examination reveals myelin pallor in the pyramids of the me-dulla oblogata (Plant et al. 1990). There is mild pallor of myelin staining of thegracile and cuneate fasciculi. In addition, the corticospinal and dorsal and ventralspinocerebellar tracts are affected (Plant et al. 1990). Recently, a unique 4K proteinsubunit called ABri has been isolated from the amyloid fibrils of FBD patients (Vidalet al. 1999). ABri is a highly insoluble peptide which is a fragment of a putativetype-II single-spanning transmembrane precursor protein that is encoded by a novelgene, BRI, located on chromosome 13.

4.4. Autosomal dominant Alzheimer’s disease and spasticparaparesis

Spastic paraparesis is not a classical feature of AD. However, a few case reports existof a rare combination of dementia and spastic paraparesis. The earliest description ofthe unusual combination of dominantly inherited Alzheimer-type presenile dementiawith spastic paralysis affecting members of a family in Belgium was published in1940 (van Bogaert et al. 1940).

32

Postmortem neuropathological analysis of the brain of one patient showed nu-merous neurofibrillary tangles as well as a multitude of various types of senileplaques not only in the cerebral but also in the cerebellar cortex. In addition, “druse-nartige Entartung” of cerebral blood vessels as well as degeneration of the pyramidaltracts were observed (van Bogaert et al. 1940). Alzheimer’s disease with either para-or tetraparesis has also been reported in an Australian family (Kwok et al. 1997), inone Japanese patient (Aikawa et al. 1985), and in two Japanese sisters (Sodeyama etal. 1995 ). After the publication of our first article (IV), several other families withpresenile Alzheimer’s disease and spastic paraparesis have been described (see Discus-sion, Table II).

33

AIMS OF THE PRESENT STUDY

The purpose of this study was:

1. To define the clinical and neuropsychological features of a novel

variant of presenile familial Alzheimer’s disease called variant

Alzheimer’s disease (varAD) (I,II)

2. To describe the neuropathological features associated with variant

Alzheimer’s disease (III)

3. To establish the molecular genetic basis of variant Alzheimer’s dis-

ease in a Finnish family (IV, V)

34

PATIENTS AND METHODS

1. PATIENTS

The study family was ascertained through a family member who contacted TheFamily Federation of Finland in the 1980’s requesting genetic counseling because ofa strong family history of an early-onset dementia combined with impaired gait.The pattern of transmission was compatible with autosomal dominant inheritance offull penetrance, but the precise diagnosis was unknown. When this study began in1996, sixteen affected relatives had been identified in four generations of the family,and three of them were still living. Further pedigree information came from familymembers, medical records, and parish registers. The first affected individual hadbeen born in 1860 in southern Finland. One branch of the family was discovered in1997 when her typical cotton wool plaques were identified. This patient had notbeen in any contact with her relatives. Three new patients started to show symptomsduring this study. Finally, 23 affected individuals (16 males and 7 females) wereidentified (Figure 1). Medical records were available for 15 and six of them werepersonally examined. Additional evaluations such as functional imaging andneurophysiologic investigations were performed at the neurological departments offour university hospitals in Helsinki, Turku, Tampere, and Oulu. All PET imagingwas performed at the Turku PET Center. Molecular genetic analysis was completedfor samples sent to the Neurogenetic Laboratory, Mayo Clinic, Jacksonville, Florida.Medical and neurologic records were available from nine additional deceasedpatients. Eight other patients had had a history of a similar disorder, but medicalrecords were unavailable. These 17 patients, plus the previous 6 patients comprisethe 23. This study includes data on neuropathologic autopsies performed on fivecases of the 15 cases (Table 1).

To date this is the only family with presenile dementia and spastic paraparesis thathas been identified in Finland.

2. ETHICAL ISSUES

This study was approved by the Ethics Committees of the Department of Neurologyof the Helsinki University Central Hospital and the Family Federation of Finland.Patients or their caregivers gave written consent for participation in the study. Be-cause the blood samples were collected for research purposes family members weretold not to expect individual results. After the mutation was found they received aninformation letter recommending that they contact a genetic counseling clinic incase they would consider presymptomatic testing. In that case, a new blood samplewould be taken.

35

For reasons of confidentiality, the family tree has been modified, and no pedigreesymbols for gender have been used.

3. CLINICAL METHODS (I, II)

3.1. Neurologic examinationA total of eleven family members (six affected and five non-affected individuals) werepersonally examined. The function of their cranial nerves was investigated, and men-tal status, motor function, gait, tendon reflexes, and sensation were examined.

3.2. Neuropsychological examinationA comprehensive neuropsychological test battery was administered to four of thesesix affected. In addition, retrospective neuropsychological data were collected fromfour other cases (during the years 1982 to 1998). Patients’ ages ranged from 43 to63 years at the time of testing, and duration of illness from the first symptomsranged from 0 to 6 years. Severity of the disease ranged from the mild to the moder-ate stage. The four retrospective cases were examined once, whereas the other fourpatients took part in the follow-up of two successive examinations. The neuropsy-chological evaluation was based on tests of verbal and visual memory, abstract think-ing, visuoconstructional and spatial functions, language ability, calculation, and tasksof ideomotor apraxia. Orientation was assessed by the Orientation questions of theWechsler Memory Scale (WMS) (Wechsler 1945). Memory tests included LogicalMemory (story A) and Visual Reproduction (pictures A and C) of the WMS-Revised(Wechsler 1987). A 10-word learning test was administered to most of the patients(Christensen 1975, Erkinjuntti et al. 1986 ).

Abstract thinking was evaluated by the Similarities Test form the Wechsler AdultIntelligence Scale (WAIS of Wechsler 1955: items 1-4, 6, 8, 9, 12, 13). Visuocon-structive Abilities were measured by Block Design from the WAIS (Wechsler 1955):items 1, 2, 4-10, and copying a cube (Christensen 1975, Erkinjuntti et al. 1986).Calculation included either mental arithmetic from a dementia test (Christensen1975, Erkinjuntti et al. 1986) or basic paper and pencil arithmetical problems usedin clinical settings. Apraxia of hand movements was mainly evaluated by the recip-rocal and spatial tasks from the Dementia test (Christensen et al. 1975, Erkinjunttiet al. 1986) and by Finger Tapping. A cohort of 13 healthy individuals were selectedas case-controls on the basis of age, education, and gender. In sporadic AD, becauseage of onset is approximately 20 years later; such patients were not considered appro-priate as controls. Healthy control subjects were recruited from a study examiningsporadic dementia (Erkinjuntti et al. 1986).

36

3.3. Psychiatric evaluationPsychiatric symptoms of eight of the 23 patients were evaluated by specialists inclinical psychiatry and neurology with particular attention to emotions and to mooddisturbances such as depression, euphoria, anxiety, delusions, and hallucinations.

3.4. Cerebrospinal fluid (CSF) analysisCerebrospinal fluid was taken from nine patients by standard procedures. Proteinlevel, cell counts of white and red blood cells, and glucose of CSF were analyzed.

3.5. Neurophysiologic studiesConventional EEG recordings were taken of 9/15 affected individuals. Five EMGand nerve conduction studies were performed. Visual evoked potentials (VEP) stud-ied in one patient and somatosensory evoked potentials (SEP) were performed onthree patients.

3.6. NeuroimagingCranial CT:s were done on nine affected individuals and on three non-symptomaticrelatives at risk for the disorder. An MRI system operating at 1.5T (GE, Signahorit-zon) was used. All four individuals subjected to brain MRI were affected. In addi-tion, spinal MRI was carried out on these four plus one patient. Before MRI wasavailable, myelography had been performed on four patients with spastic paraparesis,whereas SPECT (99mTc-HM-PAO) had been performed on four patients by standardmethods.

PET in addition to MRI was performed on the four patients with a GE (Milwau-kee, WI, USA) advance scanner giving 35 transaxial planes at 4.3 intervals. 2-[18F]-fluoro-2-deoxy-D-glucose (FDG) served as a ligand. FDG was prepared as described(Hamacher et al. 1986, Solin et al. 1988); radiochemical purity exceeded 99%, andspecific radioactivity at the time of injection was about 75 Gbq/mmol. A dynamic55-minute study was performed after intravenous injection of a bolus of 3.7 MBq/kgof FDG. Arterialized blood samples were drawn from the antecubital vein to meas-ure plasma radioactivity concentration and glucose level. The four patients studiedby PET took part in a neuropsychological follow-up study.

4. NEUROPATHOLOGICAL METHODS (III)

A neuropathological autopsy on five patients was performed 5-48 hours after death(with the exception of case III:20); these were three males and two females. Age atdeath ranged from 54 to 69 years and disease duration from 5 to 12 years. The im-mediate cause of death was bronchopneumonia in four of these cases and pulmonarythromboembolism in one case. Four of these patients had suffered spastic paraparesiswith presenile dementia; one suffered presenile dementia alone without definite evi-

37

dence of paraparesis. The brains and spinal cords of all five deceased patients werefixed in 4% phosphate-buffered formaldehyde.

4.1. Histology and immunohistochemistryRepresentative tissue samples from frontal, temporal, parietal, and occipital associa-tion cortices, the precentral motor cortex (both lateral and interhemispheric aspects),hippocampus, amygdala, basal ganglia, midbrain, cerebellum, pons, medulla, andspinal cord were embedded in paraffin. The samples were stained with the hematox-ylin-eosin (HE), periodic acid-Schiff (PAS), Luxol Fast Blue-cresyl violet (LFB-CV),and modified Bielschowsky stains. Amyloid was identified after Congo red staining(Puchtler) by red-green dichroism in polarized light and by fluorescence after thiofla-vin S staining. Selected specimens were studied after appropriate pretreatments bystandard immunoperoxidase methods for the presence of APP, various APP-derivedpeptides, hyperphosphorylated (PHF) and non-phosphorylated tau protein, neurofila-ment peptides, synaptophysin, ubiquitin, apoE, amyloid P-component, protease-re-sistant prion protein, and complement components, as well as for markers of macro-and microglia. The bound primary antibodies were visualized by use of an appropri-ate peroxidase-labeled secondary antibody (Vector Laboratories, Burlingame, CA,USA) with diaminobenzidine as the chromogen and hematozylin as the counterstain.

4.2. Confocal microscopyParaffin sections of various thickness (5 to 20 µm) were doubly immunostained withpolyclonal rabbit antisera raised against Aβ1-40, Aβ1-42, or Aβ1-43, and mousemonoclonal antibody against neurofilament or to PHF-tau. FITC-conjugated goatanti-mouse and TRITC-conjugated swine anti-rabbit secondary antibodies were usedto label the bound primary antibodies. The sections were viewed under a Leica TCSSP confocal microscope with version 1.6.587 software, and the pictures were editedwith the Adobe Photoshop 5.5 program.

4.3. Electron microscopyFor electron microscopy, samples were taken from the cerebral cortex of brains rou-tinely fixed in 4% phosphate-buffered formaldehyde. After postfixation with os-mium tetroxide, the samples were dehydrated and embedded in epon. Semi-thinsections were stained with toluidine blue. Regions of interest were selected for thecutting of thin sections which were contrasted with uranyl acetate and lead citrateand examined in a JEOL JEM 1200 electron microscope.

4.4. Biochemical analysesAβ40 and Aβ42 were extracted from fresh-frozen brain tissue of two patients (III:16,III:17) autopsied 14 and 5 hours post mortem, captured using BNT77 antibody, anddetected by use of BA27 HRP (for Aβ40) and BC05HRP (for Aβ42) antibodies withperoxidase substrate/solution (Tomita et al. 1997)

38

5. MOLECULAR GENETIC METHODS (IV, V)

5.1. Linkage analysisDNA was extracted from whole blood samples of patients and healthy family mem-bers. Markers were chosen from the vicinity of three genes, the APP, the PS-1, andthe PS-2 gene. The PS-1 gene was the only gene that preliminary showed weaklypositive lod scores.

Linkage analysis was performed of markers from around the PS-1 gene. The fol-lowing markers were used: D14S277, D14S268, D14 S77, 2 cM gap, D14S71, 2 cMgap, D14S43, D14S273, D14S284, 12 cM gap and D14S256. The lod score was cal-culated with MLINK (http://linkage.rockefeller.edu/soft/list.html#m) set at a mu-tant allele frequency of 0.01, a disease frequency of 0.01, and full penetrance by age60. Genetic data using D14S77 was generated by an ALF-Pharmacia sequencer(Uppsala, Sweden) and analyzed with the Linkage baggage programs. These datawere compared with simulation data generated for the pedigree by SIMLINK.

5.2. Analysis of cDNATo analyze the PS-1 gene in more detail, RNA was isolated from immortalized lym-phoblastoid cell lines withTrizol Reagent (Gibco BRL, Pittsburgh, PA, USA) follow-ing the manufacturer’s protocol. First strand cDNA was obtained with the Super-Script preamplification system ( Gibco BRL) and a gene specific primer (5’-GT-TCGCAGTGTGCAGTGAAATCG-3’) designed toward the 3’ end of PS-1 in exon12. The target cDNA was amplified by use of 4 different primer pairs spanning var-ious exons, designed to give products of ~700bp.

Primers in exon 5 (5’-CTGAATGCTGCCATCATG-3’) and exon 10 (5’-TGCT-GGAAAGTTCCTGGAC-3’) gave alternately spliced products which were isolatedfrom an agarose gel with a QIAquick gel extraction kit (Qiagen). Purified productswere sequenced on an ALF automated sequencer (Pharmacia) using exonic fluores-cently labeled primers and Thermosequenase cycle sequencing kit (Amersham, UK).

5.3. Defining the mutationTo characterize the ∆9 deletion at genomic level, intronic primers flanking exon 9were designed to use the PAC sequence available in the Genebank sequence database(clone DJ0054D12; accession number AC006342).

Western blotting was used to analyze PS-1 ∆9 mutant protein expression in hu-man lymphoblasts. Lymphoblasts grown in RPMI 1640 medium (Gibco) supple-mented with 15% FBS and antibiotics were harvested, washed, and then lysed bybrief sonication in lysis buffer (2% SDS in Tris buffered saline PH 7.8 containing acocktail of protease inhibitors). Fifty micrograms of total lysate protein was loadedand separated on 10% tricine gels, transferred to immobilon, and probed with PS-1antibody against residues 2 to 12.

39

5.4. ApoE genotypingIn five patients, APOE alleles ε2, ε3, and ε4 were detected following standard PCRand RFLP methods. Briefly, a 231 bp fragment of the APOE gene including sites ofthe ε2 and ε4 variants was amplified in a 25 µl PCR reaction containing 1.5 mMMgCl

2, 150 µM dNTPs, 0.3 µM of each primer, 10% dimethyl sulfoxide (DMSO),

and 0.5U Taq DNA polymerase (Promega). Amplification was done by a 60-50touchdown protocol. PCR products were digested with 3U of Hha I enzyme for 4hours, and then the DNA fragments were separated on a 4.5% agarose gel afterethidium bromide staining. The APOE genotype was determined by the pattern ofDNA fragments present (Crook et Duff 1994).

6.0 Statistical analysisThe level of significance of the differences between mean ages at onset and ages atdeath of affected men and women were analyzed by the ANOVA test with the SPSS/PC (Cary, NC, USA) program. Neuropsychological test results between our patientsand control subjects were compared by the Mann-Whitney U-test. P values of lessthan 0.05 were accepted as significant.

RESULTS

1. CLINICAL FINDINGS

1.1. Pedigree analysisThe distribution of affected individuals within the pedigree was compatible with anautosomal dominant mode of inheritance with full penetrance. Among the 23 affect-ed subjects, the male/female ratio (16/7) was 2.28.

1.2. Clinical courseThe age at onset of varAD symptoms ranged from 40 to 61 years, the mean age(±SD) being 50.9±5.0 for all affected individuals. Age of onset for men was 50.7±5.4 and 51.3 ±4.9 for women. All patients had dementia usually preceded by orco-existing with spastic paraparesis. In patients (11/15) with spastic paraparesis, itpreceded dementia by approximately 5 years. Impaired gait appeared at age 45 to 60.The course of spastic paraparesis was progressive, and most patients needed somekind of walking aid. In the late stage of the disease, all patients were bedridden.They were also mute, incontinent, and unable to cooperate. Furthermore, some pa-tients developed flexion contractures, lower facial weakness, and difficulties in swal-lowing. The duration of the disease from onset of symptoms to death was 9 years on

40

average (range 5 to 12 years), with mean age at death (± SD) being 60.7 ± 5.8 yearsfor all patients, 62.3± 4.5 for men, and 58 ± 7.4 for women. The immediate causeof death in most patients was bronchopneumonia.

1.3. Neurological signs

1.3.1. DementiaDementia was the neurologic sign shared by all affected individuals. They had diffi-culties in immediate and delayed recall of recently presented material. Progressiveloss of memory was diagnosed by the age of 40 to 61. In addition, all patients hadother neuropsychological deficits such as impairment of spatial skills, apraxia, acalcu-lia, and aphasia. Both memory deficit and impairment of other cognitive functionsprogressed, causing dementia in all patients. Thus only three patients of the fifteenmaintained some insight into their cognitive decline.

1.3.2. Spastic paraparesisSpastic paraparesis was the second most common and the most characteristic clinicalfeature of variant AD. Furthermore, spastic paraparesis was often the first symptomof the disease preceding dementia. In some patients, however, spastic paraparesis ap-peared simultaneously with dementia; while dementia preceded it in one case only.Age at onset of paraparesis was 45 to 55 years, and most patients were referred to aneurologist for impaired gait. Reliable medical data were available for 15 of 23 af-fected individuals. A total of eleven patients (three personally examined) had spasticparaparesis verified by medical examination (Table 1). Six of them had subjective ex-perience of one leg’s being weaker than the other, but the neurological examinationrevealed that all patients had bilateral spastic paraparesis. Of eleven patients withspastic paraparesis, five became wheelchair-bound after 6 years on average (range 5 to8 years), and three further patients used crutches. Six patients with spastic parapare-sis suffered lower back pain about 1.5 to 5 years before their gait disorder. Two ofthese had retired because of severe chronic back pain before the manifestation of oth-er symptoms. Neurologic examination of patients with disturbed gait showed hyper-reflexia or clonus of the lower extremities. Furthermore, in most cases the Babinskiresponse was abnormal on both sides. Sensation in the trunk remained normal. Ex-cept for three patients of whom two had polyneuropathy, sensation in the extremitiesalso remained normal. Of the 15 patients, four developed no spastic paraparesis.However, three of these were personally examined, and in neurological examinationtwo of them showed unusually brisk knee and ankle jerks.

1.3.3. Clumsiness of the hands and dysarthriaTen of the 15 affected individuals showed clumsiness of the hands, six of them also

41

had paraparesis. Neurologic examination revealed hyperreflexia in ten of whom spas-ticity of the upper extremities appeared in three. In most cases, clumsiness of thehands was present bilaterally. In three of the 10, the clumsiness was so severe that itcomplicated activities of daily life. In the late stage of the disease, these three pa-tients were unable to move their left hands but were able to use their right. Howev-er, no tetraparesis in the sense of inability to move all four extremities was seen.Dysarthria was noted in six of the fifteen patients.

1.3.4. Other clinical findingsOf 14, seven patients had motor dysphasia that progressed to aphasia. Only two ofthe 15 had myoclonus. One had myoclonus in the body and the extremities sevenyears after of onset, and the other already had facial myoclonus at the onset of thedisease. Three patients had gaze palsy resembling the supranuclear pathway type.They had difficulties in following a moving object, especially upwards. Epilepticseizures of the grand mal type were diagnosed in four. However, epilepsy was not anearly sign; it was seen from five to ten years after onset of the first symptoms.

1.4. Neuropsychological findingsThe patients showed statistically a significant cognitive decline in all tests: orienta-tion, logical memory, visual reproduction, similarities, Block Design, and copying acube. All except one patient had difficulties in orientation. Furthermore, immediaterecall in Logical Memory in most of the patients was reduced to two or three items.Visual Reproduction was impaired, and the patients were able to learn only five toseven of the ten words presented in the Learning test.

All patients had difficulties in intellectual functions. However, abstract think-ing, measured by the Similarities test, showed some variability and fell within thenormal range in two patients. The most characteristic feature besides memory defi-cits was early and prominent impairment of visuoconstructive functions. Conse-quently, all patients had difficulties in the Block Design test. In addition, most pa-tients also showed impairment of spatial functions. Two of the eight patients couldcopy a cube and a cross, but had difficulties in the Block Design, while others couldnot produce copies and were almost incapable of placing the cubes in Block Design.

There was variation concerning impairment in language functions and calcula-tion. One patient developed aphasia, while others showed only mild language diffi-culties. Difficulties in calculation was not a prominent feature. However, all pa-tients exhibiting apraxia, when tested, showed ideomotor apraxia. Four patientstook part in a follow-up study including neuropsychological examination, and thefindings correlated with findings in PET (see 1.8.4.).

Patient IV:1, who had least cognitive difficulties, showed progression only inmemory, visuoconstructional functions, apraxia, and executive functions.

42

1.5. Psychiatric signsAmong the 23 affected family members, eighth patients had a combination of de-mentia and psychiatric symptoms. The psychiatric symptoms showed great variabil-ity: In general, the most typical were emotional lability and mood disturbances.Three patients suffered from mild to moderate depression.

One patient experienced depression with anxiety, while another one had depres-sion alone. The third patient experienced depression in the early stage of her disease,but as the disease progressed, she suffered from delusions and visual hallucinations.In contrast, two patients developed abnormal euphoria.

1.6. Cerebrospinal fluid (CSF) findingsCSF was available for nine patients. In all samples, the cell count and glucose levelof the CSF were normal. In contrast, the protein concentration was elevated in five,with a range of 490 to 971mg/L (normal range, 150-450mg/L). Four of these fivepatients had spastic paraparesis.

1.7. Neurophysiologic findings

1.7.1. EEGIn this study, a total of nine conventional EEG recordings of affected individualswere taken. All EEGs were abnormal, consisting of the sporadic generalized slowdelta-theta activity (3-4 Hz) that is frequently seen in the severe stage of AD.

1.7.2. EMGOf the five EMG and nerve conduction studies performed, three indicated mild distalsensorimotor polyneuropathy; one patient also had diabetes, while the etiology forpolyneuropathy in two patients remained unknown. However, in all these casespolyneuropathy was very mild and thus could not have explained the impaired gait.One examination was abnormal in an unspecified way.

1.7.3. Somatosensory evoked potentialsVisual evoked potentials (VEP) was performed on one patient, and the results werenormal; somatosensory evoked potentials (SEP) performed on three showed normalresults, with the exception of one patient with polyneuropathy.

1.8. Findings in imaging studies Most of the nine CTs of the head showed both central and cortical atrophy. In addi-tion, some incidental old infarcts without clinical relevance were detected. Corticalatrophy ranged from moderate to severe. The hippocampal region was particularlyatrophied. All three cranial CTs done on healthy family members were normal.

Myelographies had been done on four patients with spastic paraparesis. Two ofthese had lumbar spondylosis and one was operated on for a diagnosis of spinal steno-

43

sis. However, because his spastic paraparesis deteriorated also after the operation,this indicated that spinal stenosis was not the true cause of his spastic paraparesis.Myelography of two other patients with spastic paraparesis revealed no significantabnormality

Results of a SPECT scan taken of four patients in the early stage of the diseasewere normal.

PET showed temporo-parietal hypometabolism compatible with AD in all fourpatients scanned. Variable patterns of additional hypometabolism areas were relatedto individual deficits in cognitive performance. Temporo-parietal hypometabolismwas the only PET finding in one patient who was in the early stage of the disease butalready had spastic paraparesis and slight difficulties in logical Memory, Visual Re-production, and Block Design tests.

1.8.1. PET and neuropsychological follow-upThe PET scan of patient III:24 demonstrated widespread bilateral hypometabolismin the temporo-parietal areas, with slight hypometabolism noted also in the occipitalcortex; the frontal cortex and striatum were relatively preserved. This patient alreadyhad severe visuoconstructive and spatial difficulties in the first testing, and these def-icits progressed into visual agnosia in the follow-up of 5 years and 8 months. Thepatient was able to identify common objects, but not simple drawings. Althoughprogression in all cognitive domains were noticeable, the most prominent featureswere memory and visual impairments. The PET scan of patient III:25 showed hy-pometabolism in temporal and parietal areas bilaterally, but more clearly on theright. He had memory problems, and his visuconstructive abilities were poor. ThePET scan of patient IV:2 showed bilateral hypometabolism in the temporal and pari-etal cortices, but more severely on the left. He showed severe memory impairmentand progressive dysphasia, despite only mild to moderate difficulties in visuocon-structive and spatial functions. At the follow-up seven months later, his memory per-formance had slightly deteriorated, and progressive aphasia caused impairment in theSimilarities test.

2. NEUROPATHOLOGIC FINDINGS

2.1. Macroscopic observationsPost-mortem brain and spinal cord tissues from five patients were available for neu-ropathological analysis. At macroscopic neuropathological examination, the brainweights ranged from 1075 g to 1475 g, the highest weight (1470g) being due tomarked brain edema. Except for the patient with brain edema there existed moder-ate to severe generalized cerebral gyral atrophy and pronounced shrinkage of the me-dial temporal lobes.

44

In addition, there was atrophy of the hippocampi and, at most, slight atheroscle-rosis of the intracranial arteries. One patient (III:16) had suffered an old left frontalintracerebral hemorrhage which was of the lobar type.

2.2. Histologic findings

2.2.1. General descriptionHistopathologic analysis of the brains of these patients showed unexpected but char-acteristic features. The cerebral isocortex both the primary cortex and associationcortices, including the precentral motor cortex showed numerous plaques. Severaltypes of plaques occurred, but one type predominated. Because of their characteristicmorphology, the predominant ones were termed “cotton wool “ plaques (CWP). Inaddition to these CWPs, variable numbers of diffuse and cored plaques appeared inthe cerebral cortex. The number of plaque-related macro-and microglial cells wassmall. However, although neuronal loss was observed in all areas of isocortex, it wasparticularly pronounced in the hippocampus and entorhinal cortex. Amyloid angio-pathy was present in the meningeal and cortical blood vessels and it was particularlyextensive in the precentral cortex and in the cerebellum. Diffuse and nonneuriticcored amyloid plaques appeared in the cerebellum. Degeneration of the lateral corti-cospinal tracts was observed at the level of the medulla oblongata and the spinalcord.

2.2.2. PlaquesCWPs were the most prominent neuropathological change of varAD. They wereeosinophilic structures with distinct borders but devoid of a congophilic core. Fur-thermore, they were unusually large, their diameters often exceeding 100 µm. TheseCWPs were immunoreactive for Aβ but, in contrast to neuritic and amyloid plaques,they did not, after thioflavin S staining, show congophilia/red-green dichroism orbright fluorescent cores.

In silver impregnation, only minor plaque-associated neuritic pathology was evi-dent. Interestingly, CWPs did not stain for the amyloid P-component and only in-consistently and minimally for ApoE and the complement components C1q, C3d,and C9. Their immunoreactivity for synaptophysin was accentuated in comparisonwith the surrounding neurophil. The CWPs avidly bound antibodies raised againstAβ1-16, Aβ42, and Aβ43 but were weakly nor not at all visualized by antibodiesraised against Aβ40.

Individual CWPs were frequently composed of homogeneous granular materialonly, but several of them embraced macro-and microglial cells or neuronal perikarya.Macro-and microglial cells, as identified by GFAP or HLA-DP, DQ, DR immunopo-sitivity, were very rarely seen in association with CWPs when compared with otherforms of AD.

45

The density of axons within individual CWPs was usually markedly reducedwhen compared to those of the surrounding neurophils, as demonstrated by silverimpregnation, LFB-CV, and double immunostaining for Aβ peptide and neurofila-ment proteins, including analysis with the confocal microscope. Most of the axonsseemed either to wind around the CWP or degenerate on entering the CWP.

CWPs were most commonly separate from each other but sometimes coalescedand formed conglomerates. Especially large conglomerates of CWPs occurred in theinterhemispheric precentral cortex, corresponding to the motor representation of thelower extremities, whereas the lateral motor cortex showed fewer CWPs. CWPswere common in all deep layers of the cortex (III to VI). Occasional CWPs were alsodetected in the subcortical white matter and a few even deep in the white matter.They were also present in the amygdala, putamen, globus pallidus, and claustrum.

CWPs were frequent in the entorhinal cortex and hippocampus. They were par-ticularly abundant in the subiculum, and displaced neurons and even interrupted anddistorted the granular cell layer of the dentate gyrus and the pyramidal cell band ofthe hippocampus. No CWPs were present in the cerebellum.

In the cerebral cortex, in addition to the CWPs, there appeared diffuse plaquesand cored and non-cored plaques with dystrophic or PHF type neurites. However,CWPs were always the predominant type of plaques, with the relative proportions ofother types of plaques varying among different cortical regions and individual pa-tients. The number of cored plaques was highest in patient III:16, who had sufferedan intracerebral hemorrhage. However, even in this patient, CWPs were predomi-nant in the deeper cortical layers.

2.2.3. Neurofibrillary pathologyOnly a few silver- or hyperphosphorylated tau (PHF-tau)-positive dystrophic neuriteswere associated with CWPs. Even the number of neurophil threads in the isocortexwas relatively low. In contrast, neurons bearing neurofibrillary tangles (NFT) werecommon in the isocortex and comprised the majority in the entorhinal cortex andhippocampus. Apart from NFT, no tau-or ubiquitin-immunoreactive intraneuronalor glial inclusions were observed.

2.2.4. Amyloid angiopathyAmyloid angiopathy was present in the meningeal and cortical blood vessels in allautopsied patients. Extensive congophilic amyloid deposits, immunoreactive for Aβ,were seen in the walls of these vessels. After thioflavin S staining, the vascular amy-loid deposits were strongly fluorescent in ultraviolet light. Amyloid angiopathy wasparticularly prominent in the cerebellum. In addition, there was a discontinuoussubpial layer of Aβ-immunoreactivity in a band-like pattern.

46

2.2.5. Cerebellar pathologyThere were no CWPs found in the cerebellar cortex. Instead, there were scatteredamyloid plaques with compact cores in the molecular layer. These plaques werestrongly immunoreactive for both Aβ40 and Aβ42/43 and were surrounded by a haloof Aβ42/43 positivity. These cored plaques were associated with clusters of micro-glial cells. In addition, in all three layers of the cerebellar cortex were numerous ir-regular diffuse plaques, which were immunopositive for Aβ42/43 but not for Aβ40.No obvious difference was observed between the vermis and the hemispheres. PHF-tau-positive material was absent from around the cerebellar plaques.

2.2.6. Brain stem and spinal pathologyIn the medulla oblongata and spinal cord, Luxol Fast Blue (LFB) and neurofilamentstainings disclosed a marked loss of myelin and axons in the corticospinal tracts.

Immunoreactivity for Aβ was observed only as occasional small diffuse plaques inthe central gray matter of the cord, whereas the tracts in the white matter were Aβ-negative. No PHF-tau-positive perikarya or neurites were encountered in the spinalcord.

2.3. Electronmicroscopic findingsAt the ultrastructural level in electron microscopy, apart from occasional elongatedstructures amongst the granular osmiophilic material, no definite filaments to sug-gest presence of fibrillar amyloid were detected within the CWPs. In the markedlythickened blood vessel walls, abundant bundles of straight non-branching filamentswith a diameter of about 10 nm, corresponding to amyloid fibrils, were present inthe lamina elastica interna and media where smooth muscle cells were destroyed.

2.4. Biochemical findingsBiochemical analyses of the brain tissue were available for two cases (III:16, III:17).These analyses disclosed exceptionally high total Aβ concentrations. Furthermore,in the frontal lobes the concentrations of Aβ40 were 3.88 and 22.53µg/g, and thoseof Aβ42 were 9.55 and 14.90 µg/g. These values are markedly higher than the re-ported concentrations of Aβ40 (1.66 µg/g) and Aβ42 (3.14µg/g) in sporadic AD(Houlden et al. 2000).

3. MOLECULAR GENETIC FINDINGS

3.1. Linkage analysisLinkage analysis of genetic markers from the vicinity of the APP and PS-2 genesshowed multiple examples of non co-inheritance with the disease, ruling out a muta-

47

tion at either of these two loci. Linkage analysis of markers from around the PS-1gene were inconclusive because all patients and some unaffected individuals carriedthe same haplotype.

However, modest positive two-point lod scores were obtained with several mark-ers around PS-1. Based on these lod scores, SIMLINK analysis suggested that thefamily had prior odds of having a lesion on chromosome 14q24.3 of more than 85%.

3.2. Analysis of cDNA

RNA was prepared from isolated lymphoblasts. The structure of the cDNA of thePS-1 gene was examined through RT-PCR, showing a deletion of exon 9 in the PS-1gene. Furthermore, resequencing of the intron-exon boundaries of exon 9 confirmedthat there was no splice site mutation.

3.3. Defining the mutation

Intronic primers flanking exon 9 were used. In this final analysis, an approximately6.5 kb region encompassing exon 9 showed the presence of an aberrant product thatwas present in all patients’ samples but not in healthy family members’ samples andnot in more than 150 alleles of a Finnish control population. This band was con-firmed by sequencing to be the result of a 4,555 bp deletion between exons 8 and 10and encompassing exon 9.

In Western blotting, the ∆9-PS-1 protein was not proteolytically cleaved, andlysates from lymphoblasts expressing this mutation resulted in a truncated 39-kbprotein and a reduced amount of the endogenous 28-kd N-terminal fragment.

3.4. ApoE genotypesThree of five patients whose apoE genotype was analyzed had the most common gen-otype ε3/ε3, and two patients had the genotype ε2/ε3 (Table 1). These apoE geno-types did not explain the variation in age at onset.

DISCUSSION

1. CLINICAL PHENOTYPE (I)

The earliest description of the combination of dominantly inherited Alzheimer-typepresenile dementia and spastic paraparesis was published in 1940 (van Bogaert et al.1940). Offspring of this Belgian family have not been found, and therefore molecular

48

genetic verification is lacking, but in a recent study cortical cotton wool plaques in asample from this family were documented (Houlden et al. 2000). More than fiftyyears later, a similar syndrome, characterized by the unusual combination of clinicalfeatures of presenile dementia, cerebellar symptoms, and spastic paraparesis appearedin Finnish patients (I, IV) Neither spastic paraparesis nor cerebellar symptoms hadup until then been considered typical of AD. Furthermore, although the neu-ropathological lesions indicate that the disease in our patients falls within the spec-trum of Alzheimer syndrome they do not unequivocally fulfill the established neu-ropathological criteria for AD. Consequently, we designated this disease “variantAlzheimer’s disease with spastic paraparesis”. Our southern Finnish pedigree includ-ed 23 affected individuals of both sexes in four successive generations, which is com-patible with autosomal dominant inheritance with full penetrance. The oldest fami-ly member known to be affected was born in 1860, but not enough data were availa-ble on his parents or siblings to make predictions on their affectedness status. It ispossible that some of his sibling may have passed the disease to their offspring, andnew patients from this family will turn up. Furthermore, many individuals in thefourth generation had not yet reached the age of typical onset. There was an excessof male patients (M/F ratio 16/7=2.28), and this skewed ratio may be due to the factthat, by chance, there were more men than women both in the second (8/3) and third(14/9) generations. Mean age (±SD) at onset was 50.9±5.0 years, ranging from 40 to61. This is not early when compared to ages at onset in other PS-1 gene mutations(from 29 to 64) (Haltia et al. 1994, Campion et al. 1996, Lendon et al. 1997, Crutsand Van Broeckhoven 1998, Poorkaj et al. 1998). In contrast to APP gene muta-tions (St George-Hyslop et al. 1994) and similarly to other PS-1 gene mutations(Van Broeckhoven et al. 1994, Kwok et al. 1997), the apoE genotype did not modu-late age at onset in this family. The patients died 5 to 12 years after onset, the maincause of death being bronchopneumonia. Affected men reached a slightly higher age(62.3± 4.5 years) than affected women (58 ± 7.4 years). At present, all living pa-tients are receiving cholinesterase inhibitors, and postmenopausal estrogen therapyhas been recommended for women in this family. The effect of these medications onduration of the disease remains to be seen.

The disease is clinically characterized by a slowly progressive cognitive decline fi-nally leading to dementia in all patients. In the early stage of varAD, neuropsycho-logical evaluation is of diagnostic value. In addition to disturbed immediate and de-layed recall, neuropsychological tests showed a distinctive profile. Strong impair-ment of visuoconstructive and spatial skills were early and prominent features. HSPand varAD may have several clinical features in common, but in HSP dementia is arare finding. In varAD the neuropsychological profile is different from that of sub-cortical dementia, which is characteristic of those cases of HSP with dementia (Webbet al. 1998, McDermott et al. 2000). It is of further interest that in three familieswith a mutation on chromosome 2p, HSP has been associated with cognitive impair-ment and dementia (Webb et al. 1998, Heinzlef et al. 1998, White et al. 2000). In

49

questionable cases, neuropsychological examination may be of differential diagnosticvalue.

In most cases, dementia was preceded by or appeared simultaneously with spasticparaparesis, which was the most characteristic additional clinical feature of this vari-ant AD, manifesting at 45 to 55 years of age. At the early stage, patients experi-enced walking difficulties and were frequently falling down. Consequently, in mostpatients, the initial referral to a neurologist had been because of impaired gait. Be-fore paraparesis, some but not all patients suffered from lower back pain, and two pa-tients had retired because of severe chronic back pain before the manifestation of oth-er symptoms.

From the diagnostic point of view, the clinical picture is crucial because routinelaboratory tests, neurophysiologic studies, or neuroimaging alone will not aid in thediagnosis of varAD, although they are needed to exclude other neurological diseases.In a number of disorders in the literature, dementia and spastic paraparesis are themain clinical findings.

Tropical spastic paraparesis is associated with HTLV-1 (type-C lymphotropic ret-rovirus) infection and can be excluded by clinical features alone because painful par-aesthesia in the legs and acute vasculitis do not exist in varAD. Vitamin B

12 and vi-

tamin E deficiencies can be excluded by simple laboratory tests. Amyotrophic lateralsclerosis (ALS) when involving mainly the lower extremities is often diagnosed byone or repeated EMG. Normal-pressure hydrocephalus, central nervous system infec-tions, multiple sclerosis, brain tumors, and spinal stenosis or tumors are verified byneuroimaging (CT, MRI) and by cerebrospinal fluid (CSF) analysis.

FBD in its classic form is characterized by spastic tetra- and not paraparesis.Furthermore some patients have sustained stroke-like episodes, and less constantclinical features include headaches and “blackouts”(Plant et al. 1990) that are notseen in varAD. Other clinical findings in FBD are so similar to those in varAD thatmolecular genetic analysis may be needed to confirm the clinical diagnosis.

However, neuroimaging may be useful because FBD is associated with extensiveperiventricular white matter changes also visualized in MRI of the head (Plant et al.1990). It is important to diagnose varAD because the patients can benefit from earlydiagnosis in many ways, not only from early medical treatment. Although the dis-ease cannot yet be cured, more powerful treatments may be available in the future.In addition, some family members at risk may consider presymptomatic genetic test-ing in order to make reproductive and other personal choices. This, however, re-quires careful genetic counseling.

2. NEUROPATHOLOGICAL PHENOTYPE (III)

Neuropathological investigations showed numerous, unusually distinct, large, round,eosinophilic plaques in the cerebral cortex. These “cotton wool” plaques (CWPs)

50

were distinct from all types of plaques previously described as characteristic of AD(Dickson 1997, Probst et al. 1991). In contrast to diffuse plaques, CWPs were easilydetectable even in hematoxylin-eosin stained sections. CWPs lacked a dense con-gophilic core but were immunoreactive to Aβ.

They showed strong immunoreactivity for Aβ42/43, but only weak or no immu-noreactivity for Aβ40. In addition, biochemical analysis of cortical brain tissue dem-onstrated markedly higher concentrations of Aβ42 and Aβ40 than in sporadic AD.Despite the high levels of fibrillogenic Aβ42, no definite amyloid fibrils could befound within CWPs by electron microscopy. This phenomenon may be associatedwith the paucity of macro-and microglial cells observed and the lack of complementactivation and amyloid P-component in CWPs. CWPs have been identified in mostbut not all varAD families (Table 2). Even though CWPs are not included in theneuropathological criteria of AD and are not typical of sporadic AD, in one studyscattered CWPs were reported to occur in occasional nonfamilial late-onset AD cases(Dickson 2000).

Plaque morphology is important in the distinction between varAD, FBD, andGSS. CWPs do not occur in other diseases causing dementia and spastic paraparesis.In FBD three types of plaques have been described (Plant et al. 1990, Revez et al.1999), but none of these resembles the CWPs typical of varAD. In addition to dif-ferent morphology, plaques in FBD are immunoreactive to ABri but not to beta-amyloid peptide (Plant et al. 1990, Revez et al. 1999) and white matter changes arenot seen in varAD, whereas amyloid plaques in GSS are multi-lobulated and showimmunoreactivity to PrP antibody (Masters et al. 2001).

Spastic paraparesis is a characteristic feature present in most varAD patients inthe present Finnish family. Neuropathological data provide clues as to the etiologyof spastic paraparesis. Patients show degeneration of the corticospinal tracts at thelevel of the medulla oblongata and spinal cord. The earliest description of this kindof degeneration in Alzheimer’s disease with spastic paraparesis is from the Belgianfamily (van Bogaert et al. 1940), and later this finding was observed in a Japanesepatient (Aikawa et al. 1985). In the present family, plaques were virtually absentfrom the brain stem and spinal cord, and no significant amyloid angiopathy waspresent. This finding indicates that the primary lesions may well occur at cerebrallevel. Remarkably large conglomerates of CWPs were present in the interhemispher-ic precentral cortex (paracentral lobule), corresponding to the motor representation ofthe lower extremities, while the more lateral motor area was less affected. A recentstudy showed that extensive Aβ deposition occurred even in the primary motor cor-tex of AD patients and was proposed to cause motor dysfunction in late stage of AD(Suva et al. 1999). Similarly, in GSS with spastic paraparesis the neuropathologicalanalysis does not show Aβ, but instead prion protein deposits in the cerebral cortex,especially in the motor cortex (Ghetti et al. 1995).

In varAD, in addition to abundant CWPs in the interhemispheric motor cortex,numerous NFTs and pronounced amyloid angiopathy were present in this area and

51

may have further aggravated motor dysfunction. However, amyloid angiopathy invarAD involving the spinal cord vessels is not as extensive as in FBD where amyloid,followed by ishemic damage to the cord parenchyma, is suggested to be responsiblefor the spasticity, at least in part (Plant et al. 1990). In HSP, axonal degeneration oc-curs in the tracts with the longest axons. A “dying-back” mechanism has been sug-gested beginning at the distal axon terminals and proceeding towards the cell body,which eventually disappears (Behan and Maia 1974, McDermott et al. 2000). In ourpatients, axonal injury may be a factor contributing to spastic paraparesis because thedensity of axons was reduced within the CWPs, and CWPs were particularly abun-dant in the motor cortex.

Changes in the spinal cord are present in other familial diseases associated withspastic paraparesis; however, Aβ plaques do not exist. The major neuropathologicalfeature in FSP is axonal degeneration that is maximal in terminal portions of thelongest descending and ascending tracts within the spinal cord (Behan and Maia1974), and in GSS there is degeneration with vacuolation and loss of myelin in thepyramidal tracts in the brain stem and spinal cord (Ghetti et al. 1995). There is de-generation of corticospinal tracts in the spinal cord also in FBD (Plant et al. 1990).In both varAD and FBD are plaques affecting the cerebellum and possibly causingclinical symptoms. However, the plaques affecting the cerebellum in FBD are mostfrequently of the perivascular type (Plant et al. 1990).

3. GENOTYPE PHENOTYPE CORRELATIONS (I, IV, V)

Unlike earlier examples of the ∆9 variant (Perez-Tur et al. 1995, Kwok et al. 1997,Sato et al. 1998) this deletion was not caused by a splice acceptor site mutation. Thedisease of our patients was caused by a novel genomic deletion encompassing exon 9of the PS-1 gene.

Previously, the combination of presenile autosomal dominant AD and spasticparaparesis has been described in two more Australian families (Kwok et al. 2000).One of them (AusAD1) is a 5.9kb genomic deletion in the PS-1 gene which removesa portion of intron 8, the entire exon 9, and a portion of intron 9, while the other oneis a splice acceptor mutation (Kwok et al. 2000).

After the description of the ∆9 mutation in our varAD family, varAD caused byPS-1 mutations has been reported in another British family (Houlden et al. 2000), ina third Australian family (Taddei et al. 1998, Houlden et al. 2000), in a Scottishfamily (Steiner et al. 2001), and in two families in two different states of the USA,(Moretti et al. 2000, Farlow et al. 2000) (Table 2). In varAD the presentation ofspastic paraparesis is variable, even within different members of the same family, asin this Finnish family and in an Australian one (Smith et al. 2001). In addition,whole families have been reported with ∆9 splice mutations of PS-1 gene withoutspastic paraparesis (Perez-Tur et al. 1995, Sato et al. 1998, Kwok et al. 2000). An-

52

other Finnish family with a genomic 4.6-kb deletion of exon 9 of the PS-1 gene wasrecently described (Hiltunen et al. 2000). In the pedigree there had been an affectedmother who had four children, three of whom became affected. These patients didnot have spastic paraparesis, and CWPs were not found in the one available brain bi-opsy from this family (Hiltunen et al. 2000).

It is noteworthy that in Alzheimer’s disease spastic paraparesis has been reportedonly in association with PS-1 gene mutations (Table 2) but not in APP or PS-2 genemutations. It is becoming increasingly important to identify distinctive clinical fea-tures of subtypes of AD so that molecular genetic investigations can focus on theright gene. The combination of spastic paraparesis or brisk tendon reflexes of thelower extremities, presenile dementia, and cerebellar signs should alert the clinicianto consider varAD and focus molecular genetic studies on the PS-1 gene.

4. PATHOGENETIC IMPLICATIONS

Several neuropathological findings associated with varAD may be of significance forthe understanding of the pathogenesis of AD in general.

Severe cytoarchitectural changes are associated with varAD. CWPs were fre-quent in the entorhinal cortex and hippocampus and especially abundant in thesubiculum. The expansive CWPs displaced neurons and even interrupted and dis-torted the granular cell layer of the dentate gyrus as well as the pyramidal cell bandof the hippocampus. It has been suspected that lesions of the entorhinal cortex andother limbic structures, potentially important in the genesis of memory impair-ments, may serve to disconnect the hippocampus and the neocortex (Hyman et al.1984). The functional role of the distorted cytoarchitecture in the entorhinal cortexand hippocampus in varAD and their possible contribution to dementia need furtherinvestigation.

In confocal microscopy the axons seemed either to wind around the CWP or todegenerate on entering into the CWP. It is likely that these findings are indicatorsof axonal damage and may be associated with functional disturbances. A previousstudy has indicated that in sporadic AD neurites are disrupted within a subset of Aβdeposits (Knowles et al. 1998). Neurites that pass through Aβ deposits lose theirnormal straight configuration, and this change in geometry has marked consequencesin terms of the signal transduction properties of dendrites (Knowles et al. 1999).

According to the amyloid cascade hypothesis of AD, the β-amyloid peptide istoxic to nerve cells and thereby is the main cause of the neurodegeneration in AD(Hardy et Higgins 1992). Our observations in varAD constitute a strong argumentagainst this classic amyloid cascade hypothesis in its simple reading. The presence ofdominant CWPs without amyloid cores rather than neuritic plaques in our dementedpatients supports the view that the extracellular, poorly soluble, congophilic amyloidplaques are not the only initiators of neurodegeneration in AD. However, the excep-

53

tionally high production of Aβ42/43 in varAD is likely to be of pathogenetic signifi-cance, and therefore we and others (Yankner 1998, Steiner et al. 2001) suggest thatAβ is neurotoxic at a stage prior to amyloid fibril formation.

CONCLUSIONS

On the basis of the present study, the following conclusions can be drawn:

1. Twenty-three affected individuals with presenile dementia were identified in foursuccessive generations of a Finnish family. The distribution of affected individualswithin the pedigree was compatible with an autosomal dominant mode of inherit-ance with full penetrance. The mean age (±SD) at onset was 50.9±5.0, ranging from40 to 61 years. Memory impairment was the neurological sign shared by all pa-tients. Spastic paraparesis was the second most common clinical feature of this vari-ant AD. It was often the first symptom of the disease preceding dementia. In somepatients, however, spastic paraparesis appeared simultaneously with dementia where-as it was preceded by dementia in one case only. It is noteworthy that spastic para-paresis was present in most but not all patients. Besides memory deficits, early andprominent impairment of visuoconstructive functions was characteristic. Most af-fected individuals showed clumsiness of the hands and dysarthria suggesting cerebel-lar involvement. The combination of spastic paraparesis and dementia constitutes adistinct syndrome with the frame of Alzheimer’s disease designated variant AD withspastic paraparesis. Any patient with spastic paraparesis should be tested for cogni-tive function.

2. The primary and association cortices and hippocampus of the affected individualsshowed a profusion of eosinophilic roundish structures with distinct borders termed“cotton wool” plaques (CWPs). These unusual plaques have not been described earli-er in AD or in any other neurodegenerative diseases. CWPs were immunoreactivefor Aβ42/43 but weakly or not at all for Aβ40 isoforms of the amyloid β peptide.However, the CWPs were devoid of a congophilic core, and fibrillar amyloid couldnot be seen within them by electron microscopy. In addition to the CWPs, degener-ation of the lateral corticospinal tracts was a characteristic feature. Diffuse and non-neuritic cored amyloid plaques but no CWPs occurred in the cerebellum. Our obser-vations in varAD constitute a strong argument against the classic amyloid cascadehypothesis in its simple reading. The predominance of CWPs without amyloid coresand definite amyloid fibril formation in the cerebral cortex of demented patients sug-gests that the abnormal processing of APP resulting in the accumulation of Aβ isdeleterious at a stage prior to amyloid fibril formation.

54

3. Our molecular genetic analysis began by sequencing of the exons and the imme-diate flanking intronic regions of the PS-1 gene from genomic DNA amplified byPCR. This approach, however, did not disclose the mutation, which was first identi-fied as a deletion of exon 9 through sequencing of PCR-amplified cDNA obtainedfrom lymphoblastoid mRNA. By use of intronic primers flanking exon 9, the muta-tion was eventually defined as a 4,555kb genomic deletion between exons 8 and 10of the PS-1 gene. This is the first genomic deletion mutation associated with AD.

ACKNOWLEDGMENTS

I sincerely thank the patients and their relatives of the Finn II family for their en-thusiastic cooperation. I think of them with love and respect and I am looking for-ward to more effective treatments and a better future for them.

I am grateful that I had the opportunity to work with great scientists who arenot only great as researchers but also as persons.

I wish to express my heartfelt gratitude to Matti Haltia, who gave me the op-portunity to work on this study. He introduced me to the interesting world of neu-roscience and educated me. I am very honored that I have had the chance to workwith him.

I admire Mirja Somer as a scientist, as a wise and warm person, and as an excel-lent clinical doctor. Mirja is the “mother” of this study, because she was the firstdoctor who became interested in these patients. I learnt a great deal of genetics andethics concerning genetic study from her and from Minna Pöyhönen.

I wish to express my deep appreciation to John Hardy for the fine opportunity towork in his molecular genetic laboratory at the USF in Tampa, Florida, USA. Aftermy visit the Neurogenetics Laboratory was moved to the Mayo Clinic Jacksonville,in Jacksonville, Florida. I send thanks to Richard Crook, Guy Prihar, Jordi Perez-Tur, Sarah Lincoln, Matt Baker, and Mike Hutton. I thank all the other excellentresearchers in John Hardy’s laboratory who have been working on this study andTakeshi Iwatsubo from the Graduate School of Pharmaceutical Sciences, Universityof Tokyo.

I owe my deepest gratitude also to Hannu Kalimo for his great expertise. Hewas always helpful and gave me advice on how to improve the manuscripts. I thankhim and Matti Viitanen for very valuable advice concerning this study, their encour-agement, and scientific education. I am deeply grateful to my co-author Juha Rinnefor his valuable comments and for the PET studies. I thank Timo Kurki for theMRIs and neuroradiologic consultation.

I am deeply grateful also to Raija Ylikoski for her excellent neuropsychologicalskills and her wonderful cooperation. It has been a pleasure to work with her.

55

I wish to express my sincere gratitude to the head of the Department ofPathology, Eero Saksela for providing me with exellent working facilities.

I am very grateful to my co-authors and colleagues Liisa Myllykangas, Aki Hie-taharju, Anders Peatau and Timo Erkinjuntti for their advice and scientific discus-sions.

I am deeply grateful as well to my excellent reviewers Hilkka Soininen andAnna-Elina Lehesjoki for their valuable comments and supportive criticism.

I have learnt my clinical skills by working for over ten years at the HelsinkiUniversity Central Hospital, Neurological Department. I owe my deepest gratitudeto the head of the Department of Neurology, Markku Kaste, for my clinicaleducation and for support towards this work. I thank all my wonderful friends andcolleagues at HUCH. I thank Leena Hänninen, Anne Saari, and Kaija Lindberg fortheir kind help.

I owe my warmest gratitude to Carol Norris for author-editing the language ofthis thesis. I owe my deep appreciation to Tuija Järvinen for her valuable laboratorywork.

I thank the great researchers Raimo Sulkava, Tuomo Polvikoski and Pentti Tie-nari with whom I have worked on the Vantaan Vanhimmat project

I send my warmest thanks to Raija Ahlfors for her excellent secretarial work, lab-oratory work and for her taking care of Camilla. Raija is a very special friend to meand my daughter.

I warmly thank my brother Juha Jokinen, for fixing my computer several timesand helping me with soft ware problems.

I have been blessed with extremely good friends. This thesis would not havebeen finished without the most wonderful support I got from my closest friends whohave helped me and Camilla in many ways. I am very grateful to Sari Rastas and hermother Sinikka Rastas, Johanna Lahdenperä, Pia Nordström, Minna Pöyhönen, KatiJuva, Kastehelmi Vihko, and Minna Kostiainen.

Camilla, my precious, lovable daughter, is the sunshine and joy of my life.Jyrki Uurasmaa is the man of my life whom I love strongly. Jyrki and Camilla

give me an enormous amount of love and happiness, part of which makes work hap-pen.

This work has received financial support from the Päivikki and Sakari SohlbergFoundation, the Uulo Arhio Foundation, the Neurological Foundation, the Academyof Finland (project 48173), and the Helsinki and Turku University Central Hospitals(EVO-funding).

56

Tabl

e 1

Cli

nica

l fea

ture

s an

d ap

olip

opro

tein

E g

enot

ype

in 1

5 ca

ses

of v

aria

nt A

lzhe

imer

’ s di

seas

e w

ith

spas

tic

para

pare

sis

Subj

ect

II:1

II:1

1II

I:6

III:

7II

I:10

III:

15II

I:16

*II

I:17

*II

I:20

III:

21II

I:23

III:

24*

III:

25*

IV:1

*IV

:2*

Num

ber

Age

at

onse

t48

4551

6148

5457

5455

5149

5149

5140

____

___

Age

at

deat

h/pr

esen

t#53

5658

6956

6669

6461

6254

60#

57#

53#

46#

____

___

Apo

E g

enot

ype

ND

ND

ND

ND

ND

ND

ε3/ε

3ε3

/ε3

ND

ND

ND

ε2/ε

3ε2

/ε3

ND

ε3/ε

3__

____

_

Dem

enti

a+

++

++

++

++

++

++

++

++

++

++

++

++

++

++

+15

/15

Psy

chia

tric

sym

ptom

s-

--

-+

-+

-+

++

++

++

+-

8/15

Epi

leps

y-

-+

++

--

--

+-

--

--

4/15

Dys

phas

ia-

++

-+

-+

+-

--

+-

-+

+7/

15

Gaz

e pa

resi

s-

--

+-

--

--

--

++

--

3/15

Par

apar

esis

++

++

++

++

++

++

++

++

++

--

-+

-11

/15

Hyp

erre

flex

ia/c

lonu

s of

NE

++

++

NE

++

++

-+

-+

+11

/13

knee

and

ank

le je

rks

Clu

msi

ness

of

hand

s-

+-

++

+-

++

-+

-+

++

++

+10

/15

Inte

ntio

n tr

emor

-+

NE

++

-+

+-

NE

-+

+-

-7/

13

Abn

orm

al D

DK

-+

NE

++

-+

NE

-+

-+

++

+9/

13

Dys

arth

ria

++

++

-+

--

--

-+

--

-6/

15

1 s

ubje

ct=

ped

igre

e nu

mbe

r, *=

per

sona

lly

exam

ined

, #=

pati

ent

aliv

e, N

D=

not

done

, NE

=no

t ex

amin

ed, +

=si

gn p

rese

nt, +

+=

prom

inen

t si

gn,

- =

sign

not

pre

sent

, DD

K=

diad

ocho

kine

sis

57

Table 2 Families with varAD causing spastic paraparesis and presenile dementia, or having exon 9deletion

Author and year of spastic Nationality/code for Type of PS-1 mutation Neuropathologypublication paraparesis the family

van Bogaert 1940 Belgian/CO Not known Cotton woolHoulden et al. 2000 yes plaques

Perez-Tur et al. 1995 no British/F74 ∆9 (splice site) Nearly typicalMann et al. 1996

Kwok et al. 2000 yes Australian/EOAD1/ ∆9 (genomic deletion) Cotton woolSmith et al. 2001 AusAD-1 plaques

Aus-1 Corticospinaltractdegeneration

Kwok et al. 1997 no Australian/EOAD2/ ∆9 (splice site; G to A) TypicalKwok et al. 2000 AusAD-2

Kwok et al. 1997 yes Australian/EOAD3/ ∆9 (splice site; G to T) Not reportedKwok et al. 2000 AusAD-3

Sato et al. 1998 no Japanese ∆9 (splice site) Typicalbut severe

I 2000 yes Finnish/Finn2∆9 ∆9 (genomic deletion) Cotton woolIII 2001 plaquesIV 1998V 1999

Taddei et al. 1998 yes Australian/D P436Q Cotton woolHoulden et al. 2000 plaques

Hiltunen et al. 2000 no Finnish ∆9 (genomic deletion) Cotton woolplaques notpresent in theone corticalbiopsy sampletaken

Moretti et al. 2000 yes American (Michigan) insertional mutation in Not reportedputative intracellularloop betweentransmebrane domains2 and 3

Farlow et al. 2000 yes American (Indiana) aminoacid change at Typical, butcodon 261 (valine to severe cerebralphenylalanine) amyloid

angiopathy anddegeneration oflateralpyramidal tracts

Houlden et al. 2000 yes British/EB DelIM (from exon 4) Cotton woolplaques

Steiner et al. 2001 yes Scottish ∆I83/∆M84 Cotton wool(within TM1) plaques

58

59

REFERENCES

Almkvist O. Neuropsychological features of early Alzheimer’s disease: preclinical and clinicalstages. Acta Neurol Scand 1996a: Suppl. 165:63-71.

Almkvist O, Bråne G, Johanson A. Neuropsychological assessment of dementia: state of art.Acta Neurol Scand 1996b: Suppl. 168:45-49.

Almkvist O, Basun H, Wahlund L-O, Lannfelt L. Cognitive functions of family memberswith and without the APP

670/671 mutation. In: Iqbal K, Mortimer JA, Winblad B, Wis-

niewski HM, eds. Research Advances in Alzheimer’s Disease and Related Disorders.New York: John Wiley, 1995, pp. 251-259.

Alzheimer A. Uber eine eigenartige erkrankung der hirnrinde, Allg Z Psych 1907;64:146-148.

Aikawa H, Suzuki K, Iwasaki Y, Iizuka R. Atypical Alzheimer’s disease with spastic parapare-sis and ataxia. Ann Neurol 1985;17:297-300.

Baker M, Graff-Radford D, Wavrant DeVrieze F, Graff-Radford N, Petersen RC, Kokmen E,Boeve B, Myllykangas L, Polvikoski T, Sulkava R, Verkkoniemi A, Tienari P, Haltia M,Hardy J, Hutton M, Perez-Tur J. No association between TAU haplotype and Alzheim-er’s disease in population or clinic based series or in familial disease. Neurosci Lett2000:285:147-149.

Behan WMH, Maia M. Strumpell’s familial spastic paraplegia: Genetics and neuropathology.J Neurol Neurosurg Psychiatry 1974;37:8-20.

Behl C, Davis JB, Lesley R, Schubert D. Hydrogen peroxide mediates amyloid beta proteintoxicity. Cell 1994;77:817-827.

Bertram L, Blacker D, Mullin K, Keeney D, Jones J, Basu S, Yhu S, McInnis MG, Go RCP,Vekrellis K, Selkoe DJ, Saunders AJ, Tanzi RE. Evidence for genetic linkage of Alzheim-er’s disease to chromosome 10q. Science 2000;290:2302-2303.

Bird TD, Sumi SM, Nemens EJ, Nochlin D, Schellenberg G, Lampe TH, Sadovnick A, ChuiH, Miner GW, Tiklenberg J. Phenotypic heterogeneity in familial Alzheimer’s disease:A study of 24 kindreds. Ann Neurol 1989;25:12-25.

Blacker D, Haines JL, Rodes L, Terwedow H, Go RCP, Harrell LE, Perry RT, Bassett SS,Chase G, Meyers D, Albert MS, Tanzi R. APOE4 and age at onset of Alzheimer’s disease:the NIMH genetic initiative. Neurology 1997;48:139-147.

Blacker D, Wilcox MA, Laird NM, Rodes L, Horvath SM, Go RCP, Perry R, Watson B Jr.,Bassett SS, McInnis MG, Albert MS, Hyman BT, Tanzi RE. Alpha-2 macroglobulin isgenetically associated with Alzheimer disease. Nat Genet 1998;19:357-360.

60

Braak H, Braak E. Neuropathological staging of Alzheimer-related changes. Acta Neu-ropathol 1991;82:239-259.

Breteler MMB, Claus JJ, van Duijn CM, Launer LJ, Hofman A. Epidemiology of Alzheimer’sdisease. Epidemiol Rev 1992;14:59-82.

Brown P, Goldfarb LG, Gibbs CJ, Gajdusek DC Jr. The phenotypic expression for differentmutations in transmissible familial Creutzfeldt-Jakob disease. Eur J Epidemiol1991;7:469-476.

Campion D, Brice A, Dumanchin C, Puel M, Baulac M, De La Sayette V, Hannequin D, Duy-ckaerts C, Michon A, Martin C, Moreau V, Penet C, Martinez M, Clerget-Darpoux F,Agid Y, Frebourg T. A novel presenilin 1 mutation resulting in familial Alzheimer’s dis-ease with an onset age of 29 years. Neuroreport 1996;7:1582-1584.

Campion D, Brice A, Hannequin D, Tardieu S, Dubois B, Calenda A, Brun E, Penet C, TayotJ, Martinez M, Bellis M, Mallet J, Agid Y, Clerget-Darpoux F. A large pedigree with ear-ly-onset Alzheimer’s disease. Clinical, neuropathologic, and genetic characterization.Neurology 1995a;45:80-85.

Campion D, Flaman J-M, Brice A, Hannequin D, Dubois B, Martin C, Moreau V, Charbonni-er F, Didierjean O, Tardieu S, Penet C, Puel M, Pasquier F, Le Doze F, Bellis G, CalendaA, Heiling R, Martinez M, Mallet J, Bellis M, Clerget-Darpoux F, Agid Y, Frebourg T.Mutations of the presenilin 1 gene in families with early-onset Alzheimer’s disease. HumMol Genet 1995b;4 (12):2373-2377.

Christensen A-L. Luria’s Neuropsychological Investigation. Copenhagen: Munksgaard, 1975.

Citron M, Oltersdorf T, Haass C, McConlogue L, Hung AY, Seubert P, Vigo-Pelfrey C, Lieber-burg I, Selkoe DJ. Mutation of theb-amyloid precursor protein in familial Alzheimer’sdisease increasesb-protein production. Nature 1992;360:672-674.

Corder EH, Saunders AM, Strittmatter WJ, Schmechel DE, Gaskell PC, Small GW, RosesAD, Haines JL, Pericak-Vance MA. Gene dose of apolipoprotein E type 4 allele and therisk of Alzheimer’s disease in late onset families. Science 1993;261:921-923.

Cras P, van Harskamp F, Hendriks L, Ceuterick C, van Duijn CM, Stefanko SZ, Hofman A,Kros JM, Van Broeckhoven C, Martin JJ. Presenile Alzheimer dementia characterized byamyloid angiopathy and large amyloid core type senile plaques in the APP 692Ala—->Gly mutation. Acta Neuropathol (Berl) 1998;96:253-260.

Cross HE, McKusick VA. The Mast Syndrome. Arch Neurol 1967;16:1-13.

Crook R, Hardy J, Duff K. Single-day Apolipoprotein E genotyping. J Neurosci Methods1994;53:125-127.

Cruts M, Van Broeckhoven C. Presenilin mutations in Alzheimer’s disease. Hum Mutat1998;11:183-190.

61

Cruts M, van Duijn CM, Backhovens H, Van den Broeck M, Wehnert A, Serneels S, Sher-rington R, Hutton M, Hardy J, St George-Hyslop PH, Hofman A, Van Broeckhoven C.Estimation of the genetic contribution of presenilin-1 and -2 mutations in a population-based study of presenile Alzheimer disease. Human Mol Genet 1998;7 (1): 43-51.

Cummings JL, Benson DF. Dementia: a clinical approach. 2nd ed. Boston: Butterworth-Heinemann, 1992.

Cummings JL, Benson F, Hill MA, Read S. Aphasia in dementia of the Alzheimer type.Neurology 1985;35:394-397.

Cummings JL, Vinters HV, Cole GM, Khachaturian ZS. Alzheimer’s disease. Etiologies,pathophysiology, cognitive reserve, and treatment opportunities. Neurology 1998;51:S2-S17.

Davignon J, Gregg RE, Sing CF. Apolipoprotein E polymorphism and atherosclerosis. Arte-rioscler 1988;8:1-21.

DeArmond SJ, Prusiner SB. Prion diseases. In Markesbery WR, eds. Neuropathology of de-menting disorders. (First Edition) London: Edward Arnold Publishers Ltd. 1998 pp.356-357.

Della Sala S, Lucchelli F, Spinnler H. Ideomotor apraxia in patients with dementia of Alzhe-imer type. J Neurol 1987;234:91-93.

De Strooper B, Annaert W, Cupers P, Saftig P, Craessaerts K, Mumm JS, Schroeter EH, Schri-jvers V, Wolfe MS, Ray WJ, Goate A, Kopan R. A presenilin-1-dependent γ-secretase-like protease mediates release of Notch intracellular domain. Nature 1999;398:518-522.

De Strooper B, Beullens M, Contreras B, Levesque L, Craessaerts K, Cordell B, Moechars D,Bollen M, Fraser P, St. George-Hyslop, Van Leuven F. Phosphorylation, subcellular local-ization and membrane orientation of the Alzheimer’s disease associated presenilins. J BiolChem 1997;272:3590-3598.

De Strooper B, Saftig P, Craessaerts K, Vanderstichele H, Guhde G, Annaert W, Von FiguraK, Van Leuven F. Deficiency of presenilin-1 inhibits the normal cleavage of amyloid pre-cursor protein. Nature 1998;391:387-390.

Diaz-Olavarrieta C, Ostrosky-Solis F, de la Cadena CG, Rodriguez Y, Alonso E. Neuropsycho-logical changes in subjects at risk of inheriting Alzheimer’s disease. Neuroreport1997;8:2449-2453.

Dickson DW. The pathogenesis of senile plaques. J Neuropathol Exp Neurol 1997;56:321-339.

Dickson DW. Cotton wool plaques in nonfamilial late-onset Alzheimer’s disease. 76th Annu-al Meeting of the American Association of Neuropathologists, Inc., held in Atlanta, Geor-gia. June 8-11, 2000. (Abstract) J Neuropathol Exp Neurol 2000;49:125-129.

62

Du Y, Ni B, Glinn M, Dodel RC, Bales KR, Zhang Z, Hyslop PA, Paul SM. α2-Macroglobu-

lin as a amyloid peptide-binding plasma protein. J Neurochem 1997;69:299-305.

Erkinjuntti T, Laaksonen R, Sulkva R, Syrjäläinen R, Palo J. Neuropsychological differentia-tion between normal aging. Alzheimer’s disease and vascular dementia. Acta NeurolScand 1986;74:393-403.

Ertekin-Taner Nilufer, Graff-Radford N, Younkin LH, Eckman C, Baker M, Adamson J, Ron-ald J, Blangero J, Hutton M, Younkin SG. Linkage of plasma Aβ42 to a quantitative lo-cus on chromosome 10 in late-onset Alzheimer’s disease pedigrees. Science2000;290:2303-2304.

Esiri MM, Hyman BT, Beyreuther K, Master CL. Ageing and dementia. In: Graham DI &Lantos PL, eds. Greenfield’s Neuropathology. (Sixth Edition) London: Edward ArnoldPublishers Ltd. 1997, pp.153-188.

Evans DA, Beckett LA, Field TS, Feng L, Albert MS. Apolipoprotein E ε4 and incidence ofAlzheimer disease in a community population of older persons. JAMA 1997;277:822-824.

Evans DA, Funkenstein HH, Albert MS, Scherr PA, Cook NR, Chown MJ, Hebert LE, Hen-nekens CH, Taylor JO. Prevalence of Alzheimer’s disease in a community population ofolder persons. JAMA 1989;262:2551-2556.

Farlow MR, Murell JR, Hulette CM , Ghetti B. Hereditary lateral sclerosis and Alzheimerdisease associated with mutation at codon 261 of the presenilin 1 (PS1) gene. WorldAlzheimer Congress 2000, held in Washington, DC. July 9-13, 2000. (Abstract) Neuro-biol Aging 2000;21S62.

Farlow M, Murrell J, Ghetti B, Unverzagt P, Zeldenrust S, Benson M. Clinical characteristicsin a kindred with early-onset Alzheimer’s disease and their linkage to a G—T change atposition 2149 of the amyloid precursor protein gene. Neurology 1994;44:105-111.

Farrer LA, Cupples LA, Haines JL, Hyman B, Kukull WA, Mayeux R, Myers RH, Pericak-Vance MA, Risch N, van Duijn CM. Effects of age, sex, and ethnicity on the associationbetween apolipoprotein E genotype and Alzheimer disease. A meta-analysis. JAMA1997;278:1349-1356.

Fink JK. Hereditary spastic paraplegia overview. GeneClinics 11 August 2000. Funded byNIH. Developed at the University of Washington, Seattle. wysiwyg://21/http://www.geneclinics.org/profiles/hsp/details.html?

Fink JK, Brocade Wu C-t, Jones SM, Sharp GB, Lange BM, Lesicki A, Reinglass T, Varvil T,Otterud B, Leppert M. Autosomal dominant familial spastic paraplegia: tight linkage tochromosome 15q. Am J Hum Genet 1995;56:188-192.

Fink JK, Heiman-Patterson T. Hereditary spastic paraplegia: Advances in genetic research.Neurology 1996;46:1507-1514.

63

Folstein SE, Folstein MF, McHugh PR. “Mini-mental state”: a practical method for gradingthe cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-198.

Fox NC, Kennedy AM, Harvey RJ, Lantos PL, Roques PK, Collinge J, Hardy J, Hutton M,Stevens JM, Warrington EK, Rossor MN. Clinicopathological features of familial Alzhe-imer’s disease associated with the M139V mutation in the presenilin 1 gene. Pedigreebut not mutation specific age at onset provides evidence for a further genetic factor.Brain 1997;120:491-501.

Frackowiak RSJ, Pozzilli C, Legg NJ, Du Boulay GH, Marshall J, Lenzi GL, Jones T. Region-al cerebral oxygen supply and utilization in dementia. A clinical and physiological studywith oxygen-15 and positron tomography. Brain 1981;104:753-778.

Friedland RP, Brun A, Budinger TF. Pathological and positron emission tomographic correla-tion in Alzheimer’s disease. Lancet 1985;1:228.

Funkenstein HH, Albert MS, Cook NR, West CG, Sherr PA, Chown MJ, Pilgrim D, EvansDA. Extrapyramidal signs and other neurological findings in clinically diagnosed Alzhe-imer’s disease: A community-based study. Arch Neurol 1993;50:51-56.

Gearing M, Mirra SS, Hedreen JC, Sumi SM, Hansen LA, Heyman A. The consortium to es-tablish a registry for Alzheimer’s disease (CERAD). Part X. Neuropathology confirma-tion of the clinical diagnosis of Alzheimer’s disease. Neurology 1995;45:461-466.

Ghetti B, Dlouhy SR, Giaccone G, Bugiani O, Frangione B, Farlow MR, Tagliavini F. Gerst-mann-Sträussler-Scheinker disease and the Indiana kindred. Brain Pathol 1995;5:61-75.

Gibson AM, Singleton AB, Smith G, Woodward R, McKeith IG, Perry RH, Ince PG, BallardCG, Edwardson JA, Morris CM. Lack of association of the α2-macroglobulin locus onchromosome 12 in AD. Neurology 2000;54:433-438.

Gilbert JJ, Vinters HV. Cerebral amyloid angiopathy: incidence and complications in the ag-ing brain I. cerebral hemorrhage. Stroke 1983;14:915-923.

Goate A, Chartier-Harlin M-C, Mullan M, Brown J, Crawford F, Fidani L, Giuffra L, HaynesA, Irving N, James L, Mant R, Newton P, Rooke K, Roques P, Talbot C, Pericak-VanceM, Roses A, Williamson R, Rossor M, Owen M, Hardy J. Segregation of a missense mu-tation in the amyloid precursor protein gene with familial Alzheimer’s disease. Nature1991;349:704-706.

Grundke-Iqbal I, Iqbal K, Quinlan M, Tung Y-C, Zaidi MS, Wisniewski HM. Microtubule-associated protein Tau. A component of Alzheimer paired helical filaments. J Biol Chem1986;261:6084-6089.

Haass C, De Strooper B. The presenilins in Alzheimer’s disease-Proteolysis holds the key. Sci-ence 1999;286:916-919.

Haltia M, Viitanen M, Sulkava R, Ala-Hurula V, Pöyhönen M , Goldfarb L, Brown P, Levy E,Houlden H, Crook R, Goate A, Clark R, Korenblat K, Pandit S, Keller HD, Lilius L, Liu

64

L, Axelman K, Forsell L, Winblad B, Lannfelt L, Hardy J. Chromosome 14-encodedAlzheimer’s disease: Genetic and clinicopathological description. Ann Neurol1994;36:362-367.

Hamacher K, Coenen HH, Stocklin G. Efficient stereospecific synthesis of no-carrier-added 2[18F]-fluoro-2-deoxy-D-glucose using amionopolyether supported nucleophilic substitu-tion. J Nucl Med 1986;27:235-238.

Harding AE. Hereditary “pure” spastic paraplegia: a clinical and genetic study of 22 fami-lies. J Neurol Neurosurg Psychiatry 1981;44:871-883.

Hardy J. Amyloid, the presenilins and Alzheimer’s disease. Trends Neurosci 1997;20:154-159.

Hardy JA, Higgins GA. Alzheimer’s disease: the amyloid cascade hypothesis. Science1992;256:184-185.

Hardy J, Duff K. Heterogeneity in Alzheimer’s disease. Ann Med 1993;25:437-440.

Harvey RJ, Ellison D, Hardy J, Hutton M, Roques PK, Collinge J, Fox NC, Rossor MN.Chromosome 14 familial Alzheimer’s disease: the clinical and neuropathological charac-teristics of a family with a leucine >serine (L250S) substitution at codon 250 of thepresenilin 1 gene. J Neurol Neurosurg Psychiatry 1998;64:44-49.

Hazan J, Lamy C, Melki J, Munnich A, de Recondo J, Weissenbach J. Autosomal dominantfamilial spastic paraplegia is genetically heterogeneous and one locus maps to chromo-some 14q. Nat Genet 1993;5:163-167.

Hazan J, Fonknechten N, Mavel D, Paternotte C, Samson D, Artiguenave F, Davoine CS, Cru-and C, Durr A, Wincker P, Brottier P, Cattolico L, Barbe V, Burgunder JM, Prud’hommeJF, Brice A, Fontaine B, Heilig B, Weissenbach J. Spastin, a new AAA protein, is alteredin the most frequent form of autosomal dominant spastic paraplegia. Nat Genet1999;23:296-303.

Haxby JV, Grady CL, Koss E, Horwitz B, Schapiro M, Friedland RP, Rapoport SI. Heteroge-neous anterior-posterior metabolic patterns in dementia of the Alzheimer type. Neurolo-gy 1988;38:1853-1863.

Heinzlef O, Paternotte C, Mahieux F, Prud’homme J-F, Dien J, Madigand M, Pouget J,Weissenbach J, Roullet E, Hazan J. Mapping of a complicated familial spastic paraplegiato locus SPG4 on chromosome 2p. J Med Genet 1998;35:89-93.

Helkala E-L, Laulumaa V, Soininen H, Partanen J, Riekkinen PJ. Different patterns of cogni-tive decline related to normal or deteriorating EEG in a 3-year follow-up study of patientswith Alzheimer’s disease. Neurology 1991;41:528-532.

Hendriks L, van Duijn CM, Cras P, Cruts M, Van Hul W, van Harskamp F, Warren A, McInn-is MG, Antonarakis SE, Martin J-J, Hofman A, Van Broeckhoven C. Presenile dementia

65

and cerebral haemorrhage linked to a mutation at codon 692 of the β-amyloid precursorprotein gene. Nat Genet 1992;1:218-221.

Hiltunen M, Helisalmi S, Mannermaa A, Alafuzoff I, Koivisto AM, Lehtovirta M, PirskanenM, Sulkava R, Verkkoniemi A, Soininen H. Identification of a novel 4.6-kb genomic de-letion in preselini-1 gene which results in exclusion of exon 9 in a Finnish early onsetAlzheimer’s disease family: an Alu core sequence-stimulated recombination? Eur J HumGenet 2000;8:259-266.

Houlden H, Baker M, McGowan E, Lewis P, Hutton M, Crook R, Wood NW, Kumar-SinghS, Geddes J, Swash M, Scaravilli F, Holton JL, Lashley T, Tomita T, Hashimoto T, Verk-koniemi A, Kalimo H, Somer M, Paetau A, Martin J-J, Van Broeckhoven C, Golde T,Hardy J, Haltia M, Revesz T. Variant Alzheimer’s disease with spastic paraparesis and“cotton wool” plaques is caused by PS-1 mutations that lead to exceptionally high Aβconcentrations. Ann Neurol 2000;48:806-808.

Hutton M, Hardy J. The presenilins and Alzheimer’s disease. Hum Mol Genet 1997;6:1639-1646.

Hutton M, Lendon CL, Rizzu P, Baker M, Froelich S, Houlden H, Pickering-Brown S,Chakraverty S, Isaacs A, Grover A, Hackett J, Adamson J, Lincoln S, Dickson D, DaviesP, Petersen RC, Stevens M, de Graaff E, Wauters E, van Baren J, Hillebrand M, Joosse M,Kwon JM, Nowotny P, Che LK, Norton J, Morris JC, Reed LA, Trojanowski J, Basun H,Lannfelt L, Neystat M, Fahn S, Dark F, Tannenberg A, Dodd PR, Hayward N, Kwok JBJ,Schofield PR, Andreadis A, Nowden J, Crauford D, Neary D, Owen F, Ooostra B, HardyJ, Goate A, van Swieten J, Mann D and Heutink P. Association of missense and 5´-splice-site mutations in tau with the inherited dementia FTDP-17. Nature 1998;393:702-705.

Hyman BT, van Hoesen GW, Damasio AR, Barnes CL. Alzheimer’s disease: Cell-specific pa-thology isolates the hippocampal formation. Science 1984;225:1168-1170.

Ikeda M, Sharma V, Sumi M, Rogaeva EA, Poorkaj P,Sherrington R, Nee L, Tsuda T, Oda N,Watanabe M, Aoki M, Shoji M, Abe K, Itoyama Y, Hirai S, Schellenber GD, Bird TD, StGeorge-Hyslop PH. The clinical phenotype of two missense mutations in the presenilin Igene in Japanese patients. Ann Neurol 1996;40:912-917.

Ishii K, Kitagaki H, Kono M, Mori E. Decreased medial temporal oxygen metabolism inAlzheimer’s disease shown by PET. J Nucl Med 1996;37:1159-1165.

Ishii K, Sasaki M, Kitagaki H, Yamaji S, Sakamaoto S, Matsuda K, Mori E. Reduction of cer-ebellar glucose metabolism in advanced Alzheimer’s disease J Nucl Med 1997;38:925-928.

Jacobs DM, Sano M, Dooneief G, Marder K, Bell KL, Stern Y. Neuropsychological detectionand characterization of preclinical Alzheimer’s disease. Neurology 1995;45:957-962.

Jorm AF, Korten AE, Henderson AS. The prevalence of dementia: A quantitative integrationof the literature. Acta Psychiatr Scand 1987;76:465-479.

66

Julin P, Almkvist O, Basun H, Lannfelt L, Svensson L, Winblad B, Wahlund L. Brain vol-umes and regional cerebral blood flow in carriers of the Swedish Alzheimer amyloid pro-tein mutation. Alzheimer Dis Assoc Disord 1998;12:49-53.

Juottonen K, Laakso MP, Partanen K, Soininen H. Comparative MR analysis of the entorhi-nal cortex and hippocampus in diagnosing Alzheimer disease. Am J Neuroradiol1999;20:139-144.

Juva K, Verkkoniemi A, Viramo P, Polvikoski T, Kainulainen K, Kontula K, Sulkava R.APOE ε4 does not predict mortality, cognitive decline, or dementia in the oldest old.Neurology 2000;54:412-415.

Karlinsky H, Vaula G, Haines JL, Ridgley J, Bergeron C, Mortilla M, Tupler RG, Percy ME,Robitaille Y, Noldy NE, Yip TCK, Tanzi RE, Gusella JF, Becker R, Berg JM, CrapperMclachlan DR, St. George-Hyslop PH. Molecular and prospective phenotypic characteri-zation of a pedigree with familial Alzheimer’s disease and a missense mutation in codon717 of theb-amyloid precursor protein gene. Neurology 1992;42:1445-1453.

Kehoe P, Wavrant-De Vrieze F, Crook R, Wu WS, Holmans P, Fenton I, Spurlock G, NortonM, Williams H, Williams N, Lovestone S, Perez-Tur J, Hutton M, Chartier-Harlin MC,Shears S, Roehl K, Booth J, Van Voorst W, Ramic D, Williams J, Goate A, Hardy J andOwen MJ. A full genome scan for late onset Alzheimer disease. Hum Mol Genet.1999;8:237-245.

Kennedy AM, Newman SK, Frackowiak RSJ, Cunningham VJ, Roques P, Stevens J, Neary D,Bruton CJ, Warrington EK, Rossor MN. Chromosome 14 linked familial Alzheimer’sdisease. A clinico-pathological study of a single pedigree. Brain 1995;118:185-205.

Knowles RB, Gomez-Isla T, Hyman BT. Aβ associated neuropil changes: Correlation withneuronal loss and dementia. J Neuropathol Exp Neurol 1998;57:1122-1130.

Knowles RB, Wyart C, Buldyrew SV, Cruz L, Urbanc B, Hasselmo ME, Stanley HE, HymanBT. Plaque-induced neurite abnormalities: Implications for disruption of neural net-works in Alzheimer’s disease. Proc Natl Acad Sci USA 1999;96:5274-5279.

Koss E, Friedland RP, Ober BA, Jagust WJ. Differences in lateral hemispheric asymmetriesof glucose utilization between early-and late-onset Alzheimer-type dementia. Am J Psy-chiatry 1985;142:638-640.

Kumar A, Schapiro MB, Grady C, Haxby JV, Wagner E, Salerno JA, Friedland RP, RapoportSI. High-resolution PET studies in Alzheimer’s disease. Neuropsychopharmacology1991;4:35-46.

Kwok JBJ, Smith MJ, Brooks WS, Mclean C, Hallupp M, Schofield PR. Variable presenta-tion of Alzheimer’s disease and/or spastic paraparesis phenotypes in pedigrees with a novelPS-1 exon 9 gene deletion or exon 9 splice acceptor mutations. World Alzheimer Con-gress 2000, held in Washington, DC. July 9-13, 2000 (Abstract) Neurobiol Aging2000;21:S25.

67

Kwok JBJ, Taddei K, Hallupp M, Fisher C, Brooks WS, Broe GA, Hardy J, Fulham MJ, Ni-cholson GA, Stell R, St. George Hyslop PH; Fraser PE, Kakulas B, Clarnett R, Relkin N,Gandy SE, Schofield PR, Martins RN. Two novel (M233T and R278T) presenilin-1 mu-tations in early-onset Alzheimer’s disease pedigrees and preliminary evidence for associa-tion of presenilin-1 mutations with a novel phenotype. Neuroreport 1997;8:1537-1542.

Laakso MP, Soininen H, Partanen K, Helkala EL, Hartikainen P, Vainio P, Hallikainen M,Hanninen T, Riekkinen PJ Sr. Volume loss of the hippocampus, amygdala and frontallobes in the MRI-based diagnosis of early Alzheimer’s disease: correlation with memoryfunctions. J Neural Transm 1995;9:73-86.

Lampe TH, Bird TD, Nochlin D, Nemens E, Risse SC, Sumi SM, Koerker R, Leaird B, WierM, Raskind MA. Phenotype of chromosome 14-linked familial Alzheimer’s disease in alarge kindred. Ann Neurol 1994;36:368-378.

Lannfelt L, Bogdanovic N, Appelgren H, Axelman K, Lilius L, Hansson G, Schenk D, HardyJ, Winblad B. Amyloid precursor protein mutation causes Alzheimer’s disease in a Swed-ish family. Neurosci Lett 1994;168:254-256.

Lendon CL, Ashall F, Goate AM. Exploring the etiology of Alzheimer disease using moleculargenetics. JAMA 1997;277:825-831.

Levy E, Carman MD, Fernandez-Madrid IJ, Power MD, Lieberburg I, van Duinen SG, BotsGTAM, Luyendijk W, Frangione B. Mutation of the Alzheimer’s disease amyloid gene inhereditary cerebral hemorrhage, Dutch type. Science 1990;248:1124-1126.

Levy-Lahad E, Wasco W, Poorkaj P, Romano DM, Oshima J, Pettingell WH, Yu C, JondroPD, Schmidt SD, Wang K, Crowley AC, Fu Y-H, Guenette SY, Galas D, Nemens E,Wijsman EM, Bird TD, Schellenberg GD, Tanzi RE. Candidate gene for the chromosome1 familial Alzheimer’s disease locus. Science 1995;269:973-977.

Linn RT, Wolf PA, Bachman DL, Knoefel JE, Cobb JL, Belanger AJ, Kaplan EF, D’AgostinoRB. The ‘preclinical phase’ of probable Alzheimer’s disease. A 13-year prospective studyof the Framingham Cohort. Arch Neurol 1995;52:485-490.

Lippa CF, Saunders AM, Smith TW, Swearer JM, Drachman DA, Ghetti B, Nee L, Pulaski-Salo D, Dickson D, Robitaille Y, Bergeron C, Crain B, Benson MD, Farlow M, HymanBT, St. George-Hyslop P, Roses AD, Pollen DA. Familial and sporadic Alzheimer’s dis-ease: Neuropathology cannot exclude a final common pathway. Neurology 1996;46:406-412.

Lippa CF, Schmidt ML, Nee LE, Bird T, Nochlin D, Hulette C, Mori H, Lee VM-Y, Tro-janowsky JQ. AMY plaques in familial AD: Comparison with sporadic Alzheimer’s dis-ease. Neurology 2000a;54:100-104.

Lippa CF, Swearer JM, Kane KJ, Nochlin D, Bird TD, Ghetti B, Nee LE, St. George-HyslopP, Pollen DA, Drachman DA. Familial Alzheimer’s disease: Site of mutation influencesclinical phenotype. Ann Neurol 2000b;48:376-379.

68

Lopera F, Ardilla A, Martinez A, Madrigal L, Arango-Viana JC, Lemere CA, Arango-LasprillaJC, Hincapie L, Arcos-Burgos M, Ossa JE, Behrens IM, Norton J, Lendon C, Goate AM,Ruiz-Linares A, Roselli M, Kosik KS. Clinical features of early-onset Alzheimer diseasein a large kindred with an E280A presenilin-1 mutation. JAMA 1997;277:793-799.

Mahley RW. Apolipoprotein E: cholesterol transport protein with expanding role in cell biol-ogy. Science 1988;240:622-630.

Mann DMA, Esiri MM. The pattern of acquisition of plaques and tangles in the brains of pa-tients under 50 years of age with Down’s syndrome. J Neurol Sci 1989;89:169-179.

Mann DMA, Jones D, Snowden JS, Neary D, Hardy J. Pathological changes in the brain of apatient with familial Alzheimer’s disease having a missense mutation at codon 717 in theamyloid precursor protein gene. Neurosci Lett 1992;137:225-228.

Masters CL, Beyreuther K. The Worster-Drought syndrome and other syndromes of dementiawith spastic paraparesis: The paradox of molecular pathology. J Neuropathol Exp Neurol2001;60:317-319.

Matsumura Y, Kitamura E, Miyoshi K, Yamamoto Y, Furuyama J, Sugihara T. Japanese sib-lings with missense mutation (717Val—Ile) in amyloid precursor protein of early-onsetAlzheimer’s disease. Neurology 1996;46:1721-1723.

McDermott CJ, White K, Bushby K, Shaw PJ. Hereditary spastic paraparesis: a review ofnew developments. J Neurol Neurosurg Psychiatry 2000;69:150-160.

McGeer PL, Rogers J, McGeer EG. Neuroimmune mechanisms in Alzheimer disease patients.Alzheimer Dis Assoc Disord 1994;8:149-158.

McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosisof Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group* under the auspic-es of Department of Health and Human Services Task Force on Alzheimer’s Disease.Neurology 1984;34:939-944.

Mead Simon, James-Galton Merle, Revesz Tamas, Doshi R. Bala, Harwood George, Pan Lee E,Ghiso Jorge, Frangione Blas, Plant Gordon . Familial British dementia with amyloid an-giopathy. Early clinical, neuropsychological and imaging findings. Brain 2000;123:975-991.

Mirra SS, Heyman A, McKeel D, et al. The consortium to establish a registry for Alzheimer’sdisease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzhe-imer’s disease. Neurology 1991;41:479-486.

Moretti PM, Giordani B, Kluin KJ, Minoshima S, Kuhl DE, Seltzer WK, Foster NL. Early-onset familial Alzheimer’s disease and spastic paraparesis associated with a novel inser-tional mutation of the presenilin 1 gene. World Alzheimer Congress 2000, held inWashington, DC. July 9-13, 2000. (Abstract) Neurobiol Aging 2000;21:S175.

69

Mori E, Yoneda Y, Yamashita H, Hirono N, Ikeda M, Yamadori A. Medial temporal struc-tures relate to memory impairment in Alzheimer’s disease: an MRI volumetric study. JNeurol Neurosurg Psychiatry 1997;63:214-221.

Morris JC. Classification of dementia and Alzheimer’s disease. Acta Neurol Scand1996;165:41-50.

Mullan M, Crawford F, Azelman K, Houlden H, Lilius L, Winblad B, Lannfelt Lars. A patho-genic mutation for probable Alzheimer’s disease in the APP gene at the N-terminus of β-amyloid. Nat Genet 1992;1:345-347.

Myers A, Holmans P, Marshall H, Kwon J, Meyer D, Ramic D, Shears S, Booth J, WavrantDeVrieze F, Crook R, Hamshere M, Abraham R, Tunstall N, Rice F, Carty S, Lillystone S,Kehoe P, Rudrasigham V, Jones L, Lovestone S, Perez-Tur J, Williams J, Owen MJ, Har-dy J, Goate AM. Susceptibility locus for Alzheimer’s disease on chromosome 10. Science2000;290:2304-2305.

Myllykangas Liisa, Polvikoski Tuomo, Sulkava Raimo, Verkkoniemi Auli, Crook Richard,Tienari Pentti J., Pusa Anna-Kaisa, Niinistö Leena, O’Brien Peter, Kontula Kimmo, Har-dy John, Haltia Matti. Genetic association of α

2-macroglobulin with Alzheimer’s disease

in a Finnish elderly population. Ann Neurol 1999;46:382-390.

Naruse S, Thinakaran G, Luo J-J, Kusiak JW, Tomita T, Iwatsubo T, Qian X, Ginty DD,Price DL, Borchelt DR, Wong PC, Sisodia SS. Effect of PS1 deficiency on membraneprotein trafficking in neurons. Neuron 1998;21:1213-1221,

Penttilä M, Partanen JV, Soininen H, Riekkinen PJ. Quantitative analysis of occipital EEG indifferent stages of Alzheimer’s disease. Electroencephalogr Clin Neurophysiol.1985;60:1-6.

Perez-Tur J, Froelich S, Prihar G, Crook R, Baker M, Duff K, Wragg M, Busfiled F, LendonC, Clark RF, Roques P, Fulder RA, Johnston J, Cowburn R, Forsell C, Axelman K, LiliusL, Houlden H, Karran E, Roberts GW, Rossor M, Adams MD, Hardy J, Goate A, Lan-nfelt L, Hutton M. A mutation in Alzheimer’s disease destroying a splice acceptor site inthe presenilin -1 gene. Neuroreport 1995;7:297-301.

Pericak-Vance MA, Bass MP, Yamaoka LH, Gaskell PC, Scott WK, Terwedow HA, MenoldMM, Connealy PM, Small GW, Vance JM, Saunders AM, Roses AD, Haines JL. Com-plete genomic screen in late-onset familial Alzheimer disease: evidence for a new locus onchromosome 12. JAMA 1997;278:1237-1241.

Pericak-Vance MA, Bebout JL, Gaskell PC, Yamaoka LH, Hung W.-Y, Alberts MJ, WalkerAP, Bartlett RJ, Haynes CA, Welsh KA, Earl NL, Heyman A, Clark CM, Roses AD.Linkage studies in familial Alzheimer disease: Evidence for chromosome 19 linkage. AmJ Hum Genet 1991;48:1034-1050.

Petersen RC, Smith GE, Ivnik RJ, Kokmen E, Tangalos EG. Memory function in very earlyAlzheimer’s disease. Neurology 1994;44:867-872.

70

Pridmore S, Rao G, Abusah P. Hereditary spastic paraplegia with dementia. Aust N Z J Psy-chiatry 1995;29:678-682.

Plant GT, Revesz T, Barnard RO, Harding AE, Gautier-Smith PC. Familial cerebral amyloidangiopathy with nonneuritic amyloid plaque formation. Brain 1990;113:721-747.

Poirier J, Davignon J, Bouthillier D, Kogan S, Bertrand P, Gauthier S. Apolipoprotein Epolymorphism and Alzheimer’s disease. Lancet 1993;342:697-699.

Polvikoski T, Sulkava R, Haltia M, Kainulainen K, Vuorio A, Verkkoniemi A, Niinistö L,Halonen P, Kontula K. Apolipoprotein E, dementia, and cortical deposition of β-amyloidprotein. N Engl J Med 1995;333:1242-1247.

Polvikoski T, Sulkava R, Myllykangas L, Notkola I-L, Niinistö L, Verkkoniemi A, Kainulai-nen K, Kontula K, Perez-Tur J, Hardy J, Haltia M. Prevalence of Alzheimer’s disease invery elderly people. A prospective neuropathological study. Neurology 2001;56:1690-1696.

Poorkaj P, Sharma V, Anderson L, Nemens E, Alonso ME, Orr H, White J, Heston L, BirdTD, Schellenberg GD. Missense mutations in the chromosome 14 familial Alzheimer’sdisease presenilin 1 gene. Hum Mutat 1998;11:216-221.

Prelli F, Castano E, Glenner GG, Frangione B. Differences between vascular and plaque coreamyloid in Alzheimer’s disease. J Neurochem 1988;51:648-651.

Probst A, Langui D, Ulrich J. Alzheimer’s disease: A description of the structural lesions.Brain Pathol 1991;1:229-239.

Rall SC, Weisgraber KH, Mahley RW. Human apolipoprotein E. J Biol Chem1982;257:4171-4178.

Revesz T, Holton JL, Doshi B, Anderton BH, Scaravilli F, Plant GT. Cytoskeletal pathologyin familial cerebral amyloid angiopathy (British type) with non-neuritic amyloid plaqueformation. Acta Neuropathol (Berl) 1999;97:170-176.

Rocca WA, Amaducci LA, Schoenberg BS. Epidemiology of clinically diagnosed Alzheimer’sdisease. Ann Neurol 1986;19:415-424.

Rogaev EI, Sherrington R, Rogaeva EA, Levesque G, Ikeda M, Liang Y, Chi H, Lin C, Hol-man K, Tsuda T, Mar L, Sorbi S, Macmias B, Piancentini S, Amaducci L, Chumakow I,Cohen D, Lannfelt L, Fraser PE, Rommens JM, St George-Hyslop PH. Familial Alzheim-er’s disease in kindreds with missense mutations in a gene on chromosome 1 related to theAlzheimer’s disease type 3 gene. Nature 1995;376:775-778.

Rogaeva E, Premkumar S, Song Y, Sorbi S, Brindle N, Paterson A, Duara R, Levesque G, YuG, Nishimura M, Ikeda M, O’Toole C, Kawarai T, Jorge R, Vilarino D, Bruni AC, FarrerLA, St. George-Hyslop PH. Evidence for an Alzheimer disease susceptibility locus onchromosome 12 and for further locus heterogeneity. JAMA 1998;280:614-618.

71

Roselli M, Ardila A, Moreno S, Standish V, Arango-Lasprilla JC, Tirado V, Ossa J, Goate AM,Kosic KS, Lopera F. Cognitive decline in patients with familial Alzheimer’s disease asso-ciated with E280a prsenilin-1 mutation: A longitudinal study. J Clin Exp Neuropsychol2000;22:483-495.

Rosenberg RN. The molecular and genetic basis of AD: The end of the beginning. The 2000Wartenberg lecture. Neurology 2000;54:2045-2054.

Roses AD. Apolipoprotein E genotyping in the differential diagnosis, not prediction, ofAlzheimer’s disease. Ann Neurol 1995;38:6-14.

Russo C, Schettini G, Saido TC, Hulette C, Lippa C, Lannfelt L, Ghetti B, Gambetti P, Taba-ton M, Teller JK. Presenilin-1 mutations in Alzheimer’s disease. Nature 2000;405:531-532.

Sato S, Kamino K, Miki T, Doi A, Ii K, St. George-Hyslop PH, Ogihara T, Sakaki Y. Splic-ing mutation of presenilin 1 gene for early onset, familial Alzheimer’s disease. Hum Mu-tat 1998;1:S91-S94.

Saunders AM. Apolipoprotein E and Alzheimer Disease: An update on genetic and functionalanalyses. J Neuropathol Exp Neurol 2000;59: 751-758.

Saunders AM, Strittmatter WJ, Schmechel D, St. George-Hyslop PH, Pericak-Vance MA, JooSH, Rosi BL, Gusella JF, Crapper-MacLachlan DR, Alberts MS, Hulette C, Crain B,Goldgaber D, Roses AD. Association of apolipoprotein E allele ε4 with late-onset famil-ial and sporadic Alzheimer’s disease. Neurology 1993;43:1467-1472.

Schellenberg GD, Bird TD, Wijsman EM, Orr HT, Anderson L, Nemens E, White JA, Bon-nycastle L, Weber JL, Alonso ME, Potter H, Heston LL, Martin GM. Genetic linkage ev-idence for a familial Alzheimer’s disease locus on chromosome 14. Science 1992;258:668-671.

Scheltens P, Bruyn RPM, Hazenberg GJ. A Dutch family with autosomal dominant purespastic paraparesis (Strumpell’s disease). Acta Neurol Scand 1990;82:169-173.

Scheltens P. Structural neuroimaging of Alzheimer’s disease and other dementias. Aging ClinExp Res 2001:203-209.

Selkoe DJ. Alzheimer’s disease: Genotypes, phenotype, and treatments. Science1997;275:630-631.

Sherrington R, Rogaev EI, Liang Y, Rogaeva EA, Levesque G, Ikeda M, Chi H, Lin C, Li G,Holman K, Tsuda T, Mar L, Foncin J-F, Bruni AD, Montesi MP, Sorbi S, Rainero I, Pines-si L, Nee L, Chumakow I, Pollen D, Brookes A, Sanseau P, Polinsky RJ, Wasco W, Da Sil-va HAR, Haines JL, Pericak-Vance MA, Tanzi RE, Roses AD, Fraser PE, Rommens JM,St George-Hyslop PH. Cloning a gene bearing missense mutations in early-onset familialAlzheimer’s disease. Nature 1995;375:754-760.

72

Skre H. Hereditary spastic paraplegia in Western Norway. Clin Genet 1974:6:165-183.

Sodeyama N, Shimada M, Uchihara T, Yanagisawa K, Fujigasaki H, Yamaguchi K, MatsushitaM, Yamada M. Spastic tetraplegia as an initial manifestation of familial Alzheimer’s dis-ease. J Neurol Neurosurg Psychiatry 1995;59:395-399.

Smith MJ, Gardner RJM, Knight MA, Forrest SM, Beyreuther K, Storey E, McLean CA, Cot-ton RGH, Cappal R, Masters CL. Early-onset Alzheimer’s disease caused by novel muta-tion at codon 219 of the presenilin-1 gene. Neuroreport 1999;10:503-507.

Smith MJ, Kwok JBJ, McLean CA, Kril JJ, Broc GA, Nicholson GA, Cappai R, Hallupp M,Cotton RGH, Masters CL, Schofield PR, Brooks WS. Variable phenotype of Alzheimer’sdisease with spastic paraparesis. Ann Neurol 2001;49:125-129.

Sodeyama N, Yamada M, Itoh Y, Suematsu N, Matsushita M, Otomo E, Mizusawa H. α2-macroglobulin polymorhpism is not associated with AD or AD-type neuropathology inthe Japanese. Neurology 2000; 54:443-446.

Soininen H, Partanen PJ, Laulumaa V, Helkala E-L, Laakso M, Riekkinen PJ. LongitudinalEEG spectral analysis in early stage of Alzheimer’s disease. Electroencephalogr Clin Neu-rophysiol 1989;72:290-297.

Solin O, Bergman J, Haaparanta M, Reissell A. Production of 18F from water targets, specificradioactivity, and anionic contaminants. Int J Appl Radiat Isot 1988;39:1065-1071.

Steiner H, Revesz T, Neumann M, Romig H, Grim MG, Pesold B, Kretzchmar HA, Hardy J,Holton JL, Baumeister R, Hoilden H, Haass C. A pathogenic presenilin-1 deletion caus-es aberrant Aβ42 production in the absence of congophilic amyloid plaques. J Biol Chem2001;276:7233-7239.

Strittmatter WJ, Saunders AM, Schmechel D, Pericak-Vance M, Enghild J, Salvesen GS,Roses AD. Apolipoprotein E: high-avidity binding to β-amyloid and increased frequencyof type 4 allele in late-onset familial Alzheimer disease. Proc Natl Acad Sci1993;90:1977-1981.

Sulkava R, Wikström J, Aromaa A, Raitasalo R, Lehtinen V, Lahtela K, Palo J. Prevalence ofsevere dementia in Finland. Neurology 1985;35:1025-1029.

Sulkava R, Aromaa A, Palo J, Wikström J, Erkinjuntti T, Sneck T: Dementian esiintyvyys nytja tulevina vuosikymmeninä. Suom Lääkäril 1986;41:3157-3161.

Suva D, Fabre I, Kraftsik R, Esteban M, Lobrinus A, Miklossy J. Primary motor cortex in-volvement in Alzheimer disease. J Neuropathol Exp Neurol 1999;58:1125-1134.

St. George-Hylop P, Crapper McLachlan D, Tuda T, Karlinsky H, Lippa CF, Pollen D.Alzheimer’s disease and possible gene interaction. Science 1994;263:537.

73

Swearer JM, O’Donnell BF, Drachman DA, Woodward BM. Neuropsychological features offamilial Alzheimer’s disease. Ann Neurol 1992;32:687-694.

Tanzi RE, Kovacs DM, Kim T-W, Moir RD, Guenette SY, Wasco W. The gene defects re-sponsible for familial Alzheimer’s disease. Neurobiol Dis 1996;3:159-168.

The Alzheimer’s Disease Collaborative Group: Clark RF, Hutton M, Fulder RA, Froelich S,Karran E, Talbot C, Crook R, Lendon C, Prihar G, He C, Korenblat K, Martinez A,Wragg M, Busfield F, Behrens MI, Myers A, Norton J, Morris J, Mehta N, Pearson C,Lincoln S, Baker M, Duff K, Zehr C, Perez-Tur J, Houlden H, Ruiz A, Ossa J, Lopera F,Arcos M, Madrigal L, Collinge J, Hymphreys C, Ashworth A, Sarner S, Fox N, Harvey R,Kennedy A, Roques P, Cline RT, Philips CA, Venter JC, Forsell L, Azelman K, Lilius L,Johnston J, Cowburn R, Viitanen M, Winblad B, Kosik K, Haltia M, Pöyhönen M, Dick-son D, Mann D, Neary D, Sowden J, Lantos P, Lannfelt L, Rossor M, Rossor GW, AdamsMD, Hardy J, Goate A. The structure of the presenilin 1 (S182) gene and identificationof six novel mutations in early onset AD families. Nat Genet 1995;11:219-222.

Tomita T, Maruyama K, Saido TC, Kume H, Shinozaki K, Tokuhiro S, Capell A, Walter J,Grunberg J, Haas C, Iwatsubo T, Obata K. The presenilin 2 mutation (N141) linked tofamilial Alzheimer disease (Volga German families) increases the secretion of amyloid betaprotein ending at the 42nd (or 43rd) residue. Proc Natl Acad Sci USA 1997;94:2025-2030.

van Bogaert L, Mare M, de Smedt E. Sur les formes familiales precoces de la maladied’Alzheimer. Monatschrift fur Psychiatrie und Neurologie 1940;102:249-301.

Van Broeckhoven CL. Molecular genetics of Alzheimer disease: Identification of genes andgene mutations. Eur Neurol 1995;35:8-19.

Van Broeckhoven C, Backhovens H, Cruts M, Martin JJ, Crook R, Houlden H, Hardy J.ApoE genotype does not modulate age of onset in families with chromosome 14 encodedAlzheimer’s disease. Neurosci Lett 1994;169:179-180.

Van Broeckhoven C, Haan J, Bakker E, Hardy JA, Van Hul W, Wehnert A, Vegter-Van derVlis M, Roos RAC. Amyloid β-protein precursor gene and hereditary cerebral hemor-rhage with amyloidosis (Dutch). Science 1990;248:126-129.

Vidal R, Frangione B, Rostagno A, Mead S, Revesz T, Plant G, Ghiso J. A stop-codon muta-tion in the BRI gene associated with familial British dementia. Nature 1999;399:776-781.

Walsh JS, Welch HG, Larson EB. Survival of outpatients with Alzheimer-type dementia.Ann Intern Med 1990;113:429-434.

Webb S, Coleman D, Byrne P, Parfrey N, Burke T, Hutchinson J, Hutchinson M. Autosomaldominant hereditary spastic paraparesis with cognitive loss linked to chromosome 2p.Brain 1998;121:601-609.

74

Wechsler D. A standardized memory scale for clinical use. J Psychol 1945;19:87-95.

Wechsler D. Wechsler Adult Intelligent Scale. Manual. New York: Psychological Corpora-tion, 1955

Wechsler D. WMS-R, Wechsler Memory Scale-Revised. Manual. San Antonio: The Psycho-logical Corporation. Harcourt Brace Jovanovich, Inc., 1987.

Worster-Drought C, Greenfield JG, McMenemey WH. A form of familial presenile dementiawith spastic paralysis: (including the pathological examination of a case). Brain1940;63:237-254.

Wu WS, Holmans P, Wavrant-DeVrieze F, Shears S, Kehoe P, Crook R, Booth J, Williams N,Perez-Tur J, Roehl K, Fenton I, Chartier-Harlin M-C, Lovestone S, Williamns J, HuttonM, Hardy J, Owen MJ, Goate A. Genetic studies on chromosome 12 in late-onset Alzhe-imer disease. JAMA 1998;280:619-622.

Yankner BA. Monster plaques: What they tell us about Alzheimer’s disease. Nat Med1998;4:394-455.

Yankner BA, Duffy LK, Kirschner DA. Neurotrophic and neurotoxic effects of amyloid βprotein: Reversal by Tachykinin neuropeptides. Science 1990;250: 279-282.

Yu J, Hartz J, Xu Y, Gemmill RM, Korenberg JR, Patterson D, Kao F-T. Isolation, charac-terization, and regional mapping of microclones from a human chromosome 21 microdis-section library. Am J Hum Genet 1992;51:263-272.

VARIANT ALZHEIMER’S DISEASE WITH SPASTIC PARAPARESIS: CLINICAL ...ethesis.helsinki.fi/julkaisut/laa/kliin/vk/verkkoniemi/varianta.pdf · Department of Clinical Neurocience Helsinki

Documents