OUR CURRENT UNDERSTANDING OF THE … Vol. 2, No. 4 August 2005 ABSTRACT The pathophysiology of Alzheimer’s disease (AD) is complex, involving many neurotransmitter systems and …

126 Vol. 2, No. 4 n August 2005

ABSTRACT

The pathophysiology of Alzheimer’s disease(AD) is complex, involving many neurotransmittersystems and pathophysiologic processes.Although a great deal remains to be understood,the number and diversity of aberrant mechanismsincreases the hope of finding effective anddiverse therapeutic interventions for the preven-tion and treatment of AD. The long-held hall-marks of AD—β-amyloid plaques, neurofibrillarytangles (NFTs), and neuronal cell death—arewell-known and central factors in the neurode-generative process. However, plaques and NFTsare not unique to AD. These same structuralchanges occur with normal aging and in manyother neurodegenerative disorders. In AD, theyare uniquely distributed to particular regionswithin the brain. This article reviews the patho-physiology and course of neurological damageseen during AD progression, from preclinical tosevere stages, and describes currently heldhypotheses on the causes and mechanisms lead-ing to neuronal cell death. Recent studies anddebate are expanding our understanding beyondthe definitions and details of AD pathology toexplore how the sequence and the timing ofevents within the life cycle influence the develop-ment of symptoms and illness progression. A fas-cinating (and somewhat alarming) finding is thatthe initial changes in neuronal function and neu-rotransmitter availability may actually beginyears, even decades, before there is any evi-

dence of clinical symptoms. Grasping the details ofAD pathophysiology is fundamental to understand-ing illness progression and assessing the potentialmerits and side effects of therapeutic interventions.(Adv Stud Pharm. 2005;2(4):126-135)

Alzheimer’s disease (AD) has 3 consistentneuropathologic hallmarks: plaques of β-amyloid protein (amyloid plaques), neu-rofibrillary tangles (NFTs), and neuronal

degeneration. Plaques and NFTs were first discoveredby Alois Alzheimer in a 1906 autopsy of a dementedpatient. Figures 1 and 2 show examples of plaques andNFTs after histologic staining. However, plaques andNFTs are not unique to AD, as these same structuralchanges occur with normal aging and in many otherneurodegenerative disorders. A distinguishing featureof AD is that the plaques and NFTs are localized toareas in the brain corresponding to the clinical symp-toms. Although the development of plaques and NFTseventually leads to a noticeable clinical condition, theprocess is thought to start years before the initial onsetof symptoms (preclinical AD). To understand theserelationships, it is useful to have a general knowledgeof the brain landscape.

STRUCTURAL CHANGES IN THE

ALZHEIMER’S DISEASE BRAIN

The brain areas involved in memory include thecortex, entorhinal cortex, and hippocampus (Figure3).1 A defining symptom of AD is memory loss, par-ticularly in short-term recall. Preclinical AD begins inthe entorhinal cortex, which connects the hippocam-pus, the structure responsible for memory formation(short- and long-term memory; Figure 4), to the cere-bral cortex.1 Magnetic resonance imaging studies indi-

REVIEW

OUR CURRENT UNDERSTANDING OF THE PATHOPHYSIOLOGYOF ALZHEIMER’S DISEASE

—

Jeanne Jackson-Siegal, MD*

*Assistant Clinical Professor of Psychiatry, Yale UniversitySchool of Medicine, Medical Director, Alzheimer’s ResourceCenter, New Haven, Connecticut.

Address correspondence to: Jeanne Jackson-Siegal, MD,Assistant Clinical Professor of Psychiatry, Yale UniversitySchool of Medicine, Medical Director, Alzheimer’s ResourceCenter, 1261 S Main Street, Southington, CT 06479. E-mail: [email protected].

Advanced Studies in Pharmacy n 127

REVIEW

cate that neuronal loss (measured by atrophy in theseregions) may start years before signs of memory lossemerge.2,3 As the brain atrophies, cerebrospinal fluid fillsin the space previously occupied by brain tissue.

Mild to moderate AD is characterized by promi-nent memory loss and a decline in the ability toprocess complex thoughts. In the brain, atrophy con-tinues within the cerebral cortex (Figure 5).1 Thesechanges result in the common symptoms of short-term memory loss, difficulty with balancing the check-book or other complex activities, difficulty recallingwell-known names, confusion about familiar places,and even mood and personality changes.

By the severe stage of AD, cortical atrophy hasbecome prominent in the areas that control speech,reasoning, sensory processing, and conscious thought.As larger areas of the cortex and hippocampus atrophy,the size of the lateral and third ventricles increases(Figure 6).1,4 As expected with this degree of brain atro-phy, the symptoms of severe AD increase in severity(ie, impaired long-term memory, seizures, inconti-nence, weight loss, no recognition of loved ones,unable to sit up, and groaning/moaning/grunting).Interestingly, the cortical motor strip and brain stemare left relatively intact. The progression of atrophy isshown in Figure 7.1

LEARNING

The human brain is estimated to have 100 billionneurons, which form 100 trillion synapses.1

Connections between neurons are established duringbrain development, primarily through the formationof synapses between the presynaptic neurons and post-synaptic neurons. Experience and learning strengthenneuronal connections through repeated excitatory andinhibitory inputs to postsynaptic neurons. In earlyAD, the strengthening of new inputs is diminishedbecause of the depletion of specific neurotransmitters(acetylcholine, serotonin, and norepinephrine; Figure8) and neuronal cell death. As AD progresses, theseestablished neuronal connections become increasinglydisrupted throughout much of the cerebral cortex,slowly destroying a lifetime of experiences while newlearning becomes more difficult, then impossible.

DEGENERATIVE PROCESSES IN ALZHEIMER’S DISEASE

Recent studies and debate are expanding ourunderstanding beyond the definitions and details ofAD pathology to explore how the sequence and timing

Figure 1. β-Amyloid Plaques

Image courtesy of Albert Enz, PhD, Novartis Pharmaceuticals Corporation.

Figure 2. Neurofibrillary Tangles

Image courtesy of Albert Enz, PhD, Novartis Pharmaceuticals Corporation.

Figure 3. Major Areas of the Brain

Copyright ©2002, Christy Krames.

HippocampusBrain Stem

Cerebellum

Thalamus

Cerebral CortexCorpus Callosum

Side view of the brain

Hypothalamus

128 Vol. 2, No. 4 n August 2005

REVIEW

of events within the life cycle influence the develop-ment of symptoms and illness progression. Central tothe degenerative process is the development ofplaques and tangles, although many other abnormal-ities are also occurring during the progression of AD.First, we discuss information known about the hall-mark pathologic features (plaques and tangles) andthen review other processes thought to be involvedduring the course of AD.

AMYLOID HYPOTHESIS AND

THE DEVELOPMENT OF PLAQUES

The central role of β-amyloid plaques amongst themany diverse pathophysiologic processes of AD neu-ronal degeneration is now well established. Plaques areseen in all persons as they age, but in AD, the densityis higher and the plaque distribution correlates withthe areas of neuronal degeneration and clinical symp-toms. β-amyloid plaques are clumps of insoluble pep-tides that result from the aberrant cleavage of amyloidprecursor protein (APP), a transmembrane protein. Theexact function of APP is uncertain, although severalroles have been suggested, including that of an adhesionprotein for cell-to-cell contact and structure, a factorinvolved in promoting neurite growth and perhaps dif-ferentiation, and as a modulator of gene transcriptionalactivity. APP is cleaved by 3 enzymes—β-secretase, γ-secretase, and α-secretase. Normally, cleavage by β-sec-retase, followed by γ-secretase, yields a soluble 40 aminoacid peptide (Figure 9). In AD, a variant form of the γ-secretase cleaves APP at an incorrect place, creating a 42amino acid peptide called Aβ42 or Aβ, which is not sol-uble and aggregates into identifiable clumps termed β-amyloid plaques (Figure 10).

α-secretase serves a protective function as it cleavesAPP at a site that prevents Aβ formation. Potentially,α-secretase agonists and γ-secretase inhibitors mayprovide alternate mechanisms useful in developingnew therapeutic interventions for the prevention andtreatment of AD.

There are 3 genes identified in familial AD (APP,PS1 [presenilin 1], and PS2 [presenilin 2]), and all areknown to be involved with the formation of Aβ. APP(identified as a gene rather than the protein by the useof italics) is the gene that codes for APP (the protein).It is located on chromosome 21, which may explainwhy people with Down’s syndrome (trisomy 21—3copies of chromosome 21) ultimately develop AD anddo so at a younger age than the general population.5

Figure 4. Brain Changes in Preclinical Alzheimer’sDisease

Reprinted with permission from Alzheimer’s Disease Education and ReferralCenter Web site. Available at: http://www.alzheimers.org/unraveling/04.htm.Accessed June 17, 2005.1

Figure 5. Brain Changes in Mild to ModerateAlzheimer’s Disease


Figure 6. Brain Changes in Severe Alzheimer’sDisease


CerebralCortex

EntorhinalCortex

Hippocampus

ModeratelyEnlargedVentricles

Shrinkage ofHippocampus

CorticalShrinkage

Extreme Shrinkage ofCerebral Cortex Severely

EnlargedVentricles

ExtremeShrinkage ofHippocampus


There are at least 20 known mutations of APP, morethan 140 known mutations in PS1, and approximate-ly 10 in PS2. PS1 and PS2 code for presenilin, the cat-alytic subunit of γ-secretase.5,6

In mice bred specifically for mutant APP, increasesin Aβ correlate with neuropathologic and behavioralchanges similar to AD.7 When these findings are con-sidered together, the amyloid hypothesis is very strong-ly supported, especially in familial AD. Immunizationagainst Aβ is under investigation in animals andhumans.7,8 Although promising results have beenachieved, there have been problems with encephalitis.Achieving better tolerability is the focus of intenseresearch at this time.

Recent studies also show that Aβ can aggregate inseveral ways—as oligomers (eg, aggregates of 2 or 3peptides), polymers (larger aggregates or plaques), andfibrils. Oligomers are soluble, yet appear to inducedamage to the neuron and synapse as their presence inthe synapse strongly correlates with the severity of cog-nitive impairment. Therefore, it appears that neuro-toxicity from the Aβ peptide likely occurs throughseveral mechanisms.7,9

TANGLES

Neurofibrillary tangles are seen as dying or dead neu-rons when viewed after histologic staining (Figure 2).NFTs result from the destruction of neuronal micro-tubules caused by the modification of their supportingprotein, tau (Figure 11).1 Microtubules are essentialcomponents of neuronal cell structure; they delivernutrients and assist in synaptic transmission along thelength of the neuronal axon. During AD pathogenesis,tau proteins become hyperphosphorylated, disruptingtheir bonds to microtubules, thus collapsing micro-tubule structure and destroying the neuron’s transportand communication system. Neuronal cell death ensues.As with β-amyloid plaques, hyperphosphorylation of tauwith resulting development of NFTs occurs in severalother neurodegenerative disorders, in addition to normalaging.10 It is the distribution within the brain and theirco-occurrence with plaques that make NFTs a distinctiveAD hallmark. Although the causal relationship isunclear, hyperphosphorylation of tau is thought to occurafter, thus secondary to, plaque formation.9,11

BEYOND PLAQUES AND TANGLES

In addition to plaques and NFTs, several otherpathophysiologic processes are associated with AD

REVIEW

Figure 7. Brain Damage and Atrophy During Alzheimer’sDisease Progression

A, preclinical Alzheimer’s disease; B, mild to moderate Alzheimer’s disease; and C, severe Alzheimer’s disease.Reprinted with permission from Alzheimer’s Disease Education and Referral CenterWeb site. Available at: http://www.alzheimers.org/unraveling/04.htm. Accessed June17, 2005.1

Figure 8. Neurotransmitter Deficiencies in Alzheimer’sDisease

A B C

Figure 9. The Structure of Amyloid PrecursorProtein Relative to the Cell Membrane


Acetylcholine Norepinephrine Serotonin

β-amyloid

Enzymes

130 Vol. 2, No. 4 n August 2005

progression. However, it is unclear if they occur con-currently or sequentially with regard to plaques andNFT formation. If they occur sequentially, are theseother processes the cause or result of plaques andNFTs? Although the AD pathophysiologic cascade iscomplicated, each discovery of a neurotransmitter orpathologic process leads to the potential discovery ofnew therapeutic targets through research and drugdevelopment.

CHOLINERGIC HYPOTHESIS

Acetylcholine is an important neurotransmitter inbrain regions involving memory. As expected, loss ofcholinergic activity correlates with cognitive impairment(Figure 8). In AD, cholinergic abnormalities are themost prominent of neurotransmitter changes, primarilybecause of the reduced activity of choline acetyltrans-ferase (an enzyme involved in acetylcholine synthesis).By late-stage AD, the number of cholinergic neurons ismarkedly reduced, particularly in the basal forebrain (ie,more than 75% loss of cholinergic neurons).12

Acetylcholine binds to 2 postsynaptic receptortypes: muscarinic and nicotinic. Presynaptic nicotinicreceptors influence the release of neurotransmittersimportant for memory and mood (ie, acetylcholine,glutamate, serotonin, and norepinephrine). It isknown that blocking nicotinic receptors impairs cog-nition (seen in humans and animals), whereas nico-tinic agonists may improve memory (based on studiesin rodents and nonhuman primates).13,14 Loss of nico-tinic receptor subtypes in the hippocampus and cortexhas been observed in AD.15 Muscarinic receptors arenot involved in the development of dementia, butblocking these receptors (as seen with anticholinergicagents) can cause confusion. Four of 5 approved med-ications for AD increase acetylcholine levels in thesynapse. This medication class, the cholinesteraseinhibitors, acts by blocking the enzyme acetyl-cholinesterase, which is responsible for acetylcholinedegradation. It is thought that by maintaining orincreasing acetylcholine levels in the synapse, memoryloss and cognitive function through cholinergic neu-rons could be restored or maintained, even duringneuronal degeneration.

GLUTAMATERGIC AND EXCITOTOXIC HYPOTHESIS

Glutamate, the primary excitatory neurotransmit-ter, is virtually ubiquitous in the central nervous sys-tem (CNS) and involved in essentially all CNS

functions. It is estimated to be involved in roughly66% of all brain synapses.16,17 Glutamatergic neuro-transmission is involved in learning, memory, and theshaping of neuronal architecture (plasticity).Importantly, most glutamatergic neurons are projec-tion neurons that provide information from one brainarea to another. As projection neurons, they influencecognition through connections to cholinergic neuronsin the basal forebrain and cerebral cortex. A significantfinding is that AD brains have fewer NMDA receptors(1 of 3 types of glutamate receptors, including AMPA

REVIEW

Figure 10. Insoluble Aggregates of Aβ42


Figure 11. The Role of Tau Protein in NeuronalCell Death


β-amyloidPlaque

StabilizingTau Molecules

Healthy Neuron

Microtubules

Diseased Neuron

DisintegratingMicrotubule

DisintegratingMicrotubules

Microtubule SubunitsFall Apart

Tangled Clumpsof Tau Proteins


REVIEW

and kainate) than normal.18 There also appears to beexcessive or unregulated glutamate signaling, whicheventually leads to neurotoxicity. This is not caused byexcess glutamate production or release, but it is causedby postsynaptic receptor defects that result in sus-tained low-level activation. The role of NMDA recep-tors in learning and the effect of sustained low-levelactivation are shown in Figure 12.19 Based on animalstudies, dysregulation at the glutamate NMDA recep-tor is thought to perpetuate a vicious cycle of neuronaldamage. Continuous activation of the glutamateNMDA receptor leads to chronic calcium influx thatinterferes with normal signal transduction and, overtime, increases production of APP. Increases in APPare associated with higher rates of plaque developmentand, as noted earlier in this article, hyperphosphoryla-tion of tau protein (thus NFT formation) followed byneuronal toxicity.20-23 In Part 2 of this series, we willdiscuss the role of memantine as a treatment for thecalcium dysregulation that occurs at the NMDAreceptor because of excessive glutamate signaling.

The second type of receptor defect occurs duringglutamate reuptake. During normal transmission, glu-tamate is cleared from the synapse by reuptake into thenerve terminal and surrounding glial cells. Studies inbrain tissue of patients with dementia are now show-ing reduced numbers of glutamate reuptake sites, andan association between aberrant (increased) expressionof one type of glutamate transporter and NFTs. Thisparticular transporter is normally expressed in glia, butin patients with AD, it is found in high numbers inneurons. Although a correlation between increasedexpression of a glutamate transporter and the develop-ment of AD pathology is counterintuitive (ie,increased transporter expression would be expected tocause an increase in glutamate reuptake), the investi-gators suggest that the aberrant expression may lead toNFTs, may be caused by NFTs, or may be a mecha-nism to control or prevent excitotoxic damage.24,25

OXIDATIVE STRESS

Oxidation can include the combination of a sub-stance with oxygen or free radical damage. The oxida-tive process of adding an oxygen molecule to a proteincan be a normal part of cellular function or may beaberrant, with a resulting change in the protein’s formand ability to function properly. Free radical damageoccurs when an oxygen or nitrogen molecule contain-ing an unpaired extra electron (termed species) reacts

with other molecules to achieve a stable configuration.During this process a high-energy electron is thrownoff (termed a free radical) that can cause cellular andmolecular damage. Either of these oxidative processescan cause oxidative stress with resulting cellular dam-age. The brain is especially vulnerable to damage fromoxidative stress because of its high oxygen consump-tion rate, abundant lipid content, and relative paucityof antioxidant enzymes compared to other organs. In

Figure 12. NMDA Receptor Transmission inHealthy and Degenerating Neurons

SignalDetected

Cognitive Activity

Ca2+

Rest

Magnesium

Glutamate

Ca2+

NoiseSignal

Noise

At rest, the normal NMDA receptor is blocked by magnesium. When alarge release of glutamate (such as with an action potential) occurs, gluta-mate binds. Magnesium can then leave the receptor, allowing calcium influx,causing depolarization (the cell becomes less negatively charged), and signalpropagation ensues. Learning requires a large change in voltage at the post-synaptic receptor. When the receptor is at rest, the level of noise is keptlow by the magnesium block. When the receptor is activated, the voltagechange of the membrane from the large calcium influx allows a signal to bedetected. A large absolute difference between signal and noise is requiredfor learning to occur. When the NMDA receptor is undergoing sustainedactivation, the signal-to-noise ratio is virtually one, thus learning is prevent-ed. The sustained release of calcium also causes neuronal damage.Reprinted with permission from Danysz et al. Neurotox Res. 2000;2:85-97.19

Signal Not Detected

Impairment ofNeuron Function

Neurodegeneration

Rest No Learning

Magnesium

Glutamate

Pathological Activationof NMDA Receptors

NoiseSignal Noise

DamagedNeurons

Chronic Excitation

Ca2+ Ca2+ Ca2+

A

B

132 Vol. 2, No. 4 n August 2005

REVIEW

fact, Aβ induces lipid peroxidation and generates reac-tive oxygen and nitrogen species. Oxidative damagehas been demonstrated in virtually all types of neu-ronal macromolecules (eg, lipids, carbohydrates, pro-teins, and nucleic acids). Oxidation can impairneuronal function. Oxidized lipids change the struc-ture and function of synaptic membranes resulting inneuronal death or synapse loss. Modifications to sug-ars lead to advanced glycation end-products (theattachment of sugars to proteins, lipids, or nucleicacids, usually a sign of disease or age), which causeirreversible protein cross-linking and the developmentof protein aggregates. For example, oxidation is knownto damage the principle glutamate transporter.Advanced glycation end-products themselves can cre-ate free radicals and reactive species, in addition to up-regulating proinflammatory cytokines (discussed laterin this article).26 In addition, because neurons are post-mitotic cells (ie, they have exited the cell cycle and arefully differentiated), oxidative damage is cumulativeand usually not repaired. In AD, neuronal DNA andRNA also become oxidized. DNA breaks, nicking, andfragmentation are seen, which suggest deficient DNArepair mechanisms.27 Oxidative stress is thought to beimportant early in the progression of AD and is tem-porally linked to the development of plaques andNFTs.28 Antioxidants, including vitamin E, have beeninvestigated as a possible treatment for AD. However,no satisfactory interventions have been developed yetin this category. This may be an example where timingis a critical aspect of a drug’s efficacy. For example, per-haps antioxidants may be more efficacious very earlyin the AD process, but later, after the oxidative dam-age is done, are unable to impact the pathologic courseor clinical presentation.

CHRONIC INFLAMMATION

Neurons are not the only brain cells affected in AD.Microglia, 1 of 3 glial cell types (along with astrocytesand oligodendrocytes) in the CNS, are involved inimmune and inflammatory responses to injury or infec-tion within the brain. During the AD process, microgliaare activated, releasing potentially cytotoxic molecules,such as proinflammatory cytokines, reactive oxygenspecies, proteinases, and complement proteins.29 Withinthe CNS, cytokines stimulate inflammatory processesthat may promote apoptosis (programmed cell death) ofneurons and oligodendrocytes and induce myelin dam-age. Another measure of inflammation is the finding of

increased prostaglandins, which are produced by thecyclo-oxygenases COX-1 and COX-2, in the AD brain.30

β-amyloid deposits, NFTs, and damaged neurons maystimulate inflammation as a natural response to celldamage, as occurs outside the CNS, but inflammatoryprocesses can backfire, causing more damage than pro-tection. As noted by Akiyama et al, because these stimuliare discrete, microlocalized, and present from early pre-clinical stages of AD, inflammation in AD is alsomicrolocalized, discrete, and chronic.31 In Part 2 of thisseries, we will review the use of anti-inflammatory drugsunder investigation for AD prevention and treatment.

OTHER NEUROTRANSMITTER DEFICIENCIES

During the AD process, brain regions where acetyl-choline, serotonin, and norepinephrine are prominentneurotransmitters become altered (Figure 8). Theseneuronal pathways originate in the basal forebrain andbrain stem and innervate widespread areas of the cerebralcortex. Research is also focusing on neurotransmittersthat emerge from the midbrain, notably norepinephrineand serotonin (5-hydroxytryptamine). Many studieshave underscored the importance of serotonin in affec-tive illness, and depression is a common comorbiditywith AD. It has been shown that the number of specif-ic serotonin receptors (5-HT2A and possibly the sero-tonin transporter, which is responsible for serotoninreuptake) are altered in AD brains.32 In the temporalcortex of patients with AD, the number of 5-HT2A

receptors is reduced and the extent of the reduction cor-relates with the decline in Mini-Mental StateExamination score, but it is independent of the presenceof behavioral symptoms.33 Patients with late-onset ADwho are homozygous for selected polymorphisms of the5-HT2A and 5-HT2C receptor are 5 times more likely tohave major depressive illness than heterozygotes.34 Thenumber of serotonin transporters in the temporal cortexof patients with AD with significant anxiety beforedeath appears to be preserved, but it is reduced innonanxious patients with AD. Patients with both copiesof a “high-activity allele” of the serotonin transporterhave higher rates of anxiety before death.35 Currently,dual inhibitors of serotonin transporter and acetyl-cholinesterase are under investigation in animal studiesand appear to affect levels of these neurotransmitters.36,37

Norepinephrine levels are reduced and norepi-nephrine neurons are lost in AD. Evidence suggeststhat norepinephrine may play an important role insome of the behavioral and psychological symptoms of


REVIEW

dementia (aggression, agitation, and psychosis), inaddition to memory loss.38 Animal data also suggestthat loss of norepinephrine neurons in the midbrainand norepinephrine depletion potentiate the inflam-matory responses to Aβ, suggesting that the loss ofthese neurons exacerbates neuronal cell death andinflammation in AD.39

CHOLESTEROL

The brain contains the highest amount of choles-terol of any human organ.40 Although the CNSaccounts for only 2% of body mass, it contains almost25% of the body’s unesterified cholesterol.40

Cholesterol is now also implicated in AD pathogene-sis. In vivo and in vitro data suggest that elevated cho-lesterol levels increase Aβ production, whereas reducedcholesterol synthesis (eg, as a result of statin drugs)reduces Aβ levels.41 Cholesterol reduction may alsoreduce the risk or severity of dementia through pro-tection from vascular risk factors, which often coexistin patients with AD. The APOE (apolipoprotein E)-«4allele is associated with higher cholesterol levels, is adose-dependent risk factor for AD, and correlates withan increase in Aβ pathology.42,43 In a study of 218Japanese-American men, as part of the Honolulu-AsiaAging Study, cholesterol levels were measured at mid-life and late-life and compared to the number ofplaques and NFTs. On autopsy, the results showed astrong linear association between mid-life and late-lifehigh-density lipoprotein cholesterol and the numberof these AD hallmarks in the cortex.44 In Part 2 of thisseries, we will look at the role of statins for the pre-vention and treatment of AD.

TIMING AND SEQUENCE OF NEURONAL DAMAGE

As we develop a clearer picture of the brain cell com-ponents damaged during AD, it is clear that the patho-physiologic processes are complex. An importantremaining issue is to determine whether there is a singleinitiating event that prompts a cascade of events or ifmultiple events must be present simultaneously.Traditionally, Aβ generation has been considered to bethe first step of an “amyloid cascade.” This process, alongwith oxidation, excitotoxicity, inflammation, and thehyperphosphorylation of tau, contributes to the neuro-transmitter abnormalities and neuronal cell death, whichare then manifested as cognitive and behavioral abnor-malities (Figure 13).45 Other hypotheses have also beenproposed, as described later in this article.

TWO-HIT HYPOTHESIS

Zhu et al have recently proposed a “2-hit” hypothe-sis for AD pathology—neurons may encounter aberrantmitotic signaling, causing them to re-enter the cell cycle,or be subjected to oxidative stress.46 It is unclear whyneurons would encounter aberrant mitotic signaling,causing them to re-enter the cell cycle, but some com-ponents of the cell-cycle mechanism are activated in vul-nerable neurons in AD.27,47 Re-entering the cell cycle mayultimately cause death because the mitotic structuresneeded for cell division are not observed in neurons. Thecell would then undergo a “mitotic catastrophe,” con-tributing to its eventual death.46 Either of these stressesalone could be survived through compensatory mecha-nisms. However, if both insults occur (oxidative stressand aberrant mitotic signaling), a threshold is reachedfrom which the neuron cannot recover and the AD phe-notype of neuronal loss and clinical symptoms ensue.46

Figure 13. Diagram of the Amyloid Cascade

APP = amyloid precursor protein.Reprinted with permission from Cummings. N Engl J Med. 2004;351:56-67.45

Cytoplasm

Cellmembrane

Extracellularspace

p3APP

α-APPβ-APP

Aββ-secretase

γ-secretaseγ-secretase

α-secretase

Aβ generation

Oxidation Excitotoxicity Aβ aggregation Inflammation Tau hyperphosphorylation

Cognitive and behavioralabnormalities

Neuro-transmitter

deficit

Celldeath

Neuron

Senile plaque with microglial activation

Neurofibrillarytangles

134 Vol. 2, No. 4 n August 2005

REVIEW

SYNAPTIC FAILURE

Selkoe describes the clinical presentation of ADsymptoms as a synaptic failure, rather than a disorderof neurodegeneration.11 Synapses appear to be the ini-tial target in AD pathogenesis, with subtle reductionsin synaptic efficacy before neuronal cell death.“Synapse loss matters; loss of whole neurons comeslater and matters less.” He points out that the earliestsymptom of AD (memory loss) is “remarkably pure”beginning with the loss of short-term memory (usual-ly of trivial details) and progressing onto longer-termmemory, with more serious ramifications. Thesechanges occur while motor and sensory functionsremain intact.11 Histologic analysis reveals substantial-ly decreased (approximately 30%) total numbers ofsynapses and synapses per cortical neuron in ADbrains 2 to 4 years after clinical onset.48 There may alsobe a closer correlation between synapse loss and cogni-tive deficits than with the number of plaques or tan-gles, although studies measuring Aβ peptide levels andcognitive decline also showed a strong correlation.49,50

CONCLUSIONS

As we learn more of the details involved in thepathophysiology of AD, we gain insight into potentialtherapeutic targets. The long-held hallmarks of AD—β-amyloid plaques and NFTs—are certainly major fac-tors amongst the myriad neurodegenerative processes.Perhaps more interesting is that these changes in neu-ronal function and neurotransmitter availability startyears before clinical symptoms, and they appear to cor-relate, to greater or lesser extents, to the hallmarksymptoms of AD. Optimal treatment approaches willneed to focus on the pathophysiologic changes in ADand address multiple mechanisms because of the com-plex nature of this devastating disease.

REFERENCES

1. Alzheimer’s Disease Education and Referral Center Website. Alzheimer’s disease-unraveling the mystery. Availableat: http://www.alzheimers.org/unraveling/01.htm.Accessed June 17, 2005.

2. Fox NC, Crum WR, Scahill RI, et al. Imaging of onset andprogression of Alzheimer’s disease with voxel-compressionmapping of serial magnetic resonance images. Lancet.2001;358:201-205.

3. Kaye JA, Swihart T, Howieson D, et al. Volume loss of thehippocampus and temporal lobe in healthy elderly personsdestined to develop dementia. Neurology. 1997;48:1297-1304.

4. Forstl H, Zerfass R, Geiger-Kabisch C, et al. Brain atrophyin normal aging and Alzheimer’s disease. Volumetric dis-crimination and clinical correlations. Br J Psychiatry.1995;167:739-746.

5. Selkoe DJ. Defining molecular targets to prevent Alzheimerdisease. Arch Neurol. 2005;62:192-195.

6. Bertram L, Tanzi RE. Alzheimer’s disease: one disorder, toomany genes? Hum Mol Genet. 2004;13:R135-R141.

7. Walsh DM, Selkoe DJ. Deciphering the molecular basis ofmemory failure in Alzheimer’s disease. Neuron.2004;44:181-193.

8. Walker LC, Ibegbu CC, Todd CW, et al. Emergingprospects for the disease-modifying treatment of Alzheimer’sdisease. Biochem Pharmacol. 2005;69:1001-1008.

9. Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer’sdisease: progress and problems on the road to therapeu-tics. Science. 2002;297:353-356.

10. Guillozet AL, Weintraub S, Mash DC, Mesulam MM.Neurofibrillary tangles, amyloid, and memory in aging andmild cognitive impairment. Arch Neurol. 2003;60:729-736.

11. Selkoe DJ. Alzheimer’s disease is a synaptic failure.Science. 2002;298:789-791.

12. Perry EK, Tomlinson BE, Blessed G, et al. Correlation of cholin-ergic abnormalities with senile plaques and mental test scoresin senile dementia. Br Med J. 1978;2:1457-1459.

13. Levin ED, Simon BB. Nicotinic acetylcholine involvement incognitive function in animals. Psychopharmacology.1998;138:217-230.

14. Levin ED, Rezvani AH. Development of nicotinic drug thera-py for cognitive disorders. Eur J Pharmacol. 2000;393:141-146.

15. Guan ZZ, Zhang X, Ravid R, Nordberg A. Decreased pro-tein levels of nicotinic receptor subunits in the hippocampusand temporal cortex of patients with Alzheimer’s disease. J Neurochem. 2000;74:237-243.

16. Francis PT. Glutamatergic systems in Alzheimer’s disease. IntJ Geriatr Psychiatry. 2003;18(suppl 1):S15-S21.

17. Rogawski MA, Wenk GL. The neuropharmacological basisfor the use of memantine in the treatment of Alzheimer’s dis-ease. CNS Drug Rev. 2003;9:275-308.

18. Sze C, Bi H, Kleinschmidt-DeMasters BK, et al. N-methyl-D-aspartate receptor subunit proteins and their phosphoryla-tion status are altered selectively in Alzheimer’s disease. J Neurol Sci. 2001;182:151-159.

19. Danysz W, Parsons CG, Möbius HJ, et al. Neuroprotectiveand symptomological action of memantine relevant forAlzheimer’s disease—a unified glutamatergic hypothesis onthe mechanism of action. Neurotox Res. 2000;2:85-97.

20. Parsons CG, Danysz W, Quack G. Glutamate in CNS dis-orders as a target for drug development: an update. DrugNews Perspect. 1998;11:523-569.

21. Abramov AY, Canevari L, Duchen MR. Calcium signalsinduced by amyloid beta peptide and their consequencesin neurons and astrocytes in culture. Biochim Biophys Acta.2004;1742:81-87.

22. Gordon-Krajcer W, Gajkowska B. Excitotoxicity-inducedexpression of amyloid precursor protein (beta-APP) in thehippocampus and cortex of rat brain: an electron-microscopy and biochemical study. Folia Neuropathol.2001;39:163-173.

23. Harkany T, Abraham I, Timmerman W, et al. β-amyloidneurotoxicity is mediated by a glutamate-triggered excitotox-ic cascade in rat nucleus basalis. Eur J Neurosci.2000;12:2735-2745.

24. Scott HL, Tannenberg AG, Dodd PR. Variant forms of neu-ronal glutamate transporter sites in Alzheimer disease cere-


REVIEW

bral cortex. J Neurochem. 1995;64:2193-2202.25. Scott HL, Pow DV, Tannenberg AE, Dodd PR. Aberrant

expression of the glutamate transporter excitatory aminoacid transporter 1 (EAAT1) in Alzheimer’s disease. J Neurosci. 2002;22:RC206.

26. Butterfield DA, Griffin S, Munch G, Pasinetti GM. Amyloidbeta-peptide and amyloid pathology are central to theoxidative stress and inflammatory cascades under whichAlzheimer’s disease brain exists. J Alzheimers Dis.2002;4:193-201.

27. Zhu X, Raina AK, Smith MA. Cell cycle events in neurons.Proliferation or death? Am J Pathol. 1999;155:327-329.

28. Zhu X, Lee HG, Casadesus G, et al. Oxidative imbalance inAlzheimer’s disease. Mol Neurobiol. 2005;31:205-218.

29. Kim SU, de Vellis J. Microglia in health and disease. J Neurosci Res. In press.

30. McGeer P, McGeer EG. Inflammation, cytotoxicity, andAlzheimer disease. Neurobiol Aging. 2001;22:799-809.

31. Akiyama H, Barger S, Barnum S, et al. Inflammation andAlzheimer’s disease. Neurobiol Aging. 2000;21:383-421.

32. Mossner R, Schmitt A, Syagailo Y, et al. The serotonin trans-porter in Alzheimer’s and Parkinson’s disease. J NeuralTransm Suppl. 2000;60:345-350.

33. Lai MK, Tsang SW, Alder JT, et al. Loss of serotonin 5-HT2A receptors in the postmortem temporal cortex corre-lates with rate of cognitive decline in Alzheimer’s disease.Psychopharmacology. 2005;179:673-677.

34. Holmes C, Arranz M, Collier D, et al. Depression inAlzheimer’s disease: the effect of serotonin receptor genevariation. Am J Med Genet B Neuropsychiatr Genet.2003;119:40-43.

35. Tsang SW, Lai MK, Francis PT, et al. Serotonin transportersare preserved in the neocortex of anxious Alzheimer’s dis-ease patients. Neuroreport. 2003;14:1297-1300.

36. Abe Y, Aoyagi A, Hara T, et al. Pharmacological character-ization of RS-1259, an orally active dual inhibitor of acetyl-cholinesterase and serotonin transporter, in rodents: possibletreatment of Alzheimer’s disease. J Pharmacol Sci. 2003;93:95-105.

37. Toda N, Tago K, Marumoto S, et al. Design, synthesis, andstructure-activity relationships of dual inhibitors of acetyl-cholinesterase and serotonin transporter as potential agents

for Alzheimer’s disease. Bioorg Med Chem. 2003;11:1935-1955.

38. Herrmann N, Lanctot KL, Khan LR. The role of norepinephrinein the behavioral and psychological symptoms of dementia. J Neuropsychiatry Clin Neurosci. 2004;16:261-276.

39. Heneka MT, Galea E, Gavriluyk V, et al. Noradrenergicdepletion potentiates beta-amyloid-induced cortical inflam-mation: implications for Alzheimer’s disease. J Neurosci.2002;22:2434-2442.

40. Dietschy JM, Turley SD. Cholesterol metabolism in the brain.Curr Opin Lipidol. 2001;12:105-112.

41. Reiss AB. Cholesterol and apolipoprotein E in Alzheimer’s dis-ease. Am J Alzheimers Dis Other Demen. 2005;20:91-96.

42. Rebeck GW, Reiter JS, Strickland DK, Hyman BT.Apolipoprotein E in sporadic Alzheimer’s disease: allelic varia-tion and receptor interactions. Neuron. 1993;11:575-580.

43. Schmechel DE, Saunders AM, Strittmatter WJ, et al.Increased amyloid beta-peptide deposition in cerebral cor-tex as a consequence of apolipoprotein E genotype in late-onset Alzheimer disease. Proc Natl Acad Sci U S A.1993;90:9649-9653.

44. Launer LJ, White LR, Petrovitch H, et al. Cholesterol andneuropathologic markers of AD: a population-based autop-sy study. Neurology. 2001;57:1447-1452.

45. Cummings JL. Alzheimer’s disease. N Engl J Med.2004;351:56-67.

46. Zhu X, Raina AK, Perry G, Smith MA. Alzheimer’s disease:the two-hit hypothesis. Lancet Neurol. 2004;3:219-226.

47. Raina AK, Zhu X, Rottkamp CA, et al. Cyclin’ toward demen-tia: cell-cycle abnormalities and abortive oncogenesis inAlzheimer disease. J Neurosci Res. 2000;61:128-133.

48. Davies CA, Mann DM, Sumpter PQ, Yates PO. A quantita-tive morphometric analysis of the neuronal and synapticcontent of the frontal and temporal cortex in patients withAlzheimer’s disease. J Neurol Sci. 1987;78:151-164.

49. Naslund J, Haroutunian V, Mohs R, et al. Correlation betweenelevated levels of amyloid beta-peptide in the brain and cogni-tive decline. JAMA. 2000;283:1571-1577.

50. Terry RD, Masliah E, Salmon DP, et al. Physical basis ofcognitive alterations in Alzheimer’s disease: synapse loss isthe major correlate of cognitive impairment. Ann Neurol.1991;30:572-580.

Documents

OUR CURRENT UNDERSTANDING OF THE … Vol. 2, No. 4 August 2005 ABSTRACT The pathophysiology of Alzheimer’s disease (AD) is complex, involving many neurotransmitter systems and …