J Neuropathol Exp Neurol Copyright Ó 2014 by the American Association of Neuropathologists, Inc.

Vol. 73, No. 3 March 2014 pp. 192Y205

REVIEW ARTICLE

Alzheimer Disease and Cellular Mechanisms of Memory Storage Yuri I. Arshavsky, PhD

Abstract Most ongoing efforts to combat Alzheimer disease (AD) are focused on treating its clinical symptoms, but the neuropathologic changes underlying AD appear decades earlier and become essentially irreversible by the time the disease reaches its clinical stages. This necessitates treating AD at preclinical stages, which requires a better understanding of the primary mechanisms leading to AD pathology. Here I argue that such an understanding calls for addressing perhaps the most puzzling question in ADVwhy the underlying pathology selectively impairs neurons that are involved in memory formation and storage. Memory formation is associated with epigenetic chromatin modifications and may, therefore, be accompanied by the synthesis of proteins unique to neurons involved in memory. These proteins could be recognized by the immune system as ‘‘nonself’’ antigens. This does not happen in the healthy brain because of its isolation from the immune system by the blood-brain barrier (BBB). All risk factors for AD impair the BBB, which may allow the immune system to attack memory-involved neurons and make them vulnerable to AD-associated pathology. This hypothesis is testable and, if confirmed, could redirect therapeutic efforts toward maintaining BBB integrity people belonging to AD risk groups rather than treating them when it is too late. Key Words: Alzheimer disease, Autoimmune reaction, Blood-brain barrier, Cerebral amyloid angiopathy, Memory, Epigenetics, Risk factors.

progress. It is quite possible that one of the reasons for this failure is that all existing efforts to fight AD (i.e. antiinflammatory drugs, acetylcholine esterase inhibitors, antiamyloid immunization, and others) are focused on treating the signs and symptoms characteristic of the late preclinical or clinical stages of AD (2, 3). However, the initial neuropathologic changes of AD appear in middle ageVdecades before AD progresses into the clinical stage (Fig. 1). By the time AD reaches the clinical or even late preclinical stage, the pathologic alterations have developed to a degree such that any therapeutic intervention is unlikely to be efficient. Therapeutic treatments for patients with clinically diagnosed AD seem analogous to treating gangrenous patients with antiinflammatory drugs. Thus, alternative strategies aimed at preventing the AD clinical stage in individuals belonging to risk groups for AD are needed. This, in turn, requires a fundamental understanding of the primary mechanisms of AD pathobiology. In this review, I argue that understanding the primary mechanisms leading to AD would benefit greatly from understanding the molecular mechanisms of declarative (explicit) memory. Based on an analysis of the literature concerning both AD and cellular mechanisms of memory, I offer a hypothesis on the pathogenesis of AD and discuss possible approaches to its prevention, or at least delaying its onset (see also 4, 5).

THE MAIN MYSTERY OF AD INTRODUCTION Alzheimer disease (AD) is the most common cause of dementia in the elderly. In industrialized countries, 11% to 20% of individuals older than 70 years suffer from AD. Currently, AD affects more than 5 million individuals in the United States alone (1). Caring for AD patients costs more than $150 billion annually, which means that AD is not only an individual tragedy for patients and their families but also a serious socioeconomic problem. Alzheimer disease was described more than 100 years ago. Since that time, most areas of medicine have achieved extraordinary success; however, AD has not shared this From the BioCircuits Institute, University of California San Diego, La Jolla, California. Send correspondence and reprint requests to: Yuri I. Arshavsky, PhD, BioCircuits Institute, University of California San Diego, La Jolla, CA 92093-0402; E-mail: [email protected] The authors declare that there is no funding or conflicts of interest to report.

192

Alzheimer disease is a slowly progressing neurodegenerative disease. It starts solely as a disease of declarative (particularly episodic) memory (6Y10). Memory impairments that typically occur in the preclinical stage of AD are not accompanied by disruptions of other cognitive functions (including intelligence) or sensory and motor disturbances. Interestingly, AD is a uniquely human disease; no spontaneous AD-like pathology has been found in other species (7, 11, 12). One could suggest that AD is a ‘‘trade-off’’ for the ability of the human brain to acquire and store an enormous amount of information that is not comparable to the quantity and quality of information acquired and stored by animal brains. The destructive processes underlying AD are mainly caused by the abnormal extracellular deposition of the amyloid A (AA) peptide, leading to the formation of soluble AA oligomers and amyloid (senile) plaques, and by the hyperphosphorylation of the intraneural tau protein, leading to the formation of neurofibrillary tangles (NFTs). In agreement with clinical observations, these pathologic processes occur in J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

Alzheimer Disease and Memory

RISK FACTORS FOR AD Because of the very slow progression of AD, its symptoms are typically manifested in older individuals (Fig. 1). Thus, old age is usually regarded as the main risk factor for this disease. As emphasized by Nelson et al (23), however, AD is not ‘‘brain aging’’ but a definite form of neuropathology. Therefore, it is important to consider specific genetic and nongenetic factors leading to both early- and late-onset AD.

Genetic Risk Factors

FIGURE 1. Spatiotemporal progression of neurofibrillary pathology in Alzheimer disease. The intensity of the red color is proportional to the density of neurofibrillary tangles. The bright red color shows atrophied areas. MTL, medial temporal lobe; MC, motor cortex; SSC, somatosensory cortex; VC, visual cortex. Modified Figure 3 from Smith (14) (Reproduced with permission, Copyright 2002, the US National Academy of Sciences).

the brain structures involved in learning and memory (6Y10, 13Y16). The progression of neurofibrillary pathology begins within the structures of the medial temporal lobe, including the hippocampus and entorhinal cortex, which play a decisive role in the initial formation of memory (Fig. 1) (17, 18). The pathologic process then spreads to neocortical association areas in the prefrontal, temporal, parietal, and occipital lobes, which are involved in the final consolidation and storage of declarative memory (17, 18). Meanwhile, other neocortical areas, such as the primary sensory and motor cortices, remain mostly unaffected. Deterioration of the primary sensory and motor cortices, along with the breakdown of all other brain functions, occurs only at the terminal stages of AD. This brain destruction probably results from the development of a nonspecific inflammatory process that is initiated by AA deposition and products of neurodegeneration (19, 20). Within the cortical areas involved in memory formation and storage, only selected types of neurons are sensitive to the development of the pathologic process (6, 7, 15, 16, 21, 22). For example, in the hippocampus and entorhinal cortex, the neurofibrillary pathology mainly occurs in the pyramidal cells, but not in the granule cells. Similarly, in the association neocortical areas, the pyramidal neurons in layers II, III, and V, which form both short- and long-distance excitatory corticocortical connections, are the most vulnerable. On the other hand, local circuit interneurons are relatively spared. This raises perhaps the most puzzling questions in AD (4, 5,16): Why do AA deposition and tau hyperphosphorylation occur in brain structures related to learning and memory whereas other parts of the brain are spared? Why is it that only the neurons involved in memory formation and storage are vulnerable to AD pathology? Before discussing these questions, I will turn to an analysis of AD risk factors.

The AA peptides, consisting of 38 to 43 amino acid residues, are cleaved from the transmembrane amyloid precursor protein (APP) because of sequential APP proteolysis by A- and F-secretases; the catalytic components of the F-secretase are presenilins (8, 24). Autosomal dominant mutations in the AA region of the APP gene, as well as in the presenilin-1 and presenilin-2 genes, lead to overproduction of AA, including its insoluble AA-42 form, which plays the most significant role in forming amyloid plaques. These mutations evoke early-onset AD, that is, before the age of 65 years. Notably, even in the case of early-onset AD, pathologic changes begin in the brain at least 2 decades before the estimated onset of clinical symptoms (25, 26). Although mutations of the APP and presenilin genes account for only a small fraction (È1%) of AD cases, their discovery played a major role in the formulation of the ‘‘amyloid cascade hypothesis’’ (27). This hypothesis suggests that the abnormal deposition of AA triggers the cascade of pathologic changes observed in the brains of AD patients, including the formation of NFTs and subsequent neuronal and synaptic death. The amyloid hypothesis strongly impacted further studies of the pathologic mechanisms underlying AD (8, 24Y26). One implication of the amyloid hypothesis is the concept of anti-AA immunotherapy. This concept originated from studies of genetically modified mice expressing human mutant APP (28Y30). These mice display AD-like pathology, that is, amyloid plaque formation and memory deficiency. Immunizing these animals with synthetic AA-42 prevented or at least significantly reduced both AA deposition and memory impairment (31). These promising results initiated clinical trials of active AA immunotherapy in patients with mild to moderate AD. Unfortunately, these trials did not yield anticipated positive results (32, 33). There was no difference in memory and cognitive test performance between the immunized and placebo groups despite postmortem studies revealing a reduced quantity of amyloid plaques in the treated group. Some authors suggested that the inefficiency of AA immunotherapy was explained by its application at the very late stages of AD (26, 34). However, no attempts to use this approach at earlier disease stages were undertaken because of an unexpected side effect observed in the first trial: 18 of the 300 treated patients developed meningoencephalitis after receiving 1 to 3 injections (35). Several approaches to passive AA immunotherapy in which AD patients were treated with anti-AA antibodies have also been attempted. Whereas the clinical efficacy of passive immunotherapy was equivocal (26, 36), it was producing serious side effects of a vascular origin, such as vasogenic edema and cerebral microhemorrhages (37).

Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

193

Arshavsky

Despite the lack of major success in these early attempts, a number of new more sophisticated immunotherapy trials are currently underway (38). Despite the broad acceptance of the amyloid hypothesis, especially among investigators involved in the experimental studies of AD, a growing body of pathoanatomic evidence shows that AA deposition in the brain parenchyma may not be the primary causative factor in AD (39Y42). In particular, it was shown that NFTs are formed independently of AA deposition. For example, neurofibrillary pathology appears first in the medial temporal lobe and then spreads to neocortical areas (Fig. 1), whereas AA pathology appears first in the neocortex and later spreads to the medial temporal lobe (13, 43, 44). Furthermore, the clinical status of AD patients correlates better with the NFT burden than with AA deposition (6, 7, 13, 15, 16, 41). Another genetic factor provoking the more common late-onset form of AD is related to the polymorphism of apolipoprotein E (ApoE), a protein participating in the metabolism of lipids. There are 3 isoforms of ApoE, namely, ?2, ?3, and ?4. The inheritance of at least one ?4 allele of the ApoE gene is a powerful genetic risk factor for AD that determines up to 70% of AD cases (8, 45Y48). In contrast, the more common ApoE3 isoform and especially the rare ApoE2 isoform are likely to protect against AD. According to the amyloid hypothesis, ApoE4 disturbs the intrabrain AA metabolism. However, the mechanism underlying this disturbance remains obscure. As emphasized in a recent review by Kim et al (47, p 287), ‘‘although there have been numerous studies attempting to elucidate the underlying mechanism for this increased risk, how apoE4 influences AD onset and progression has yet to be proven.’’

Nongenetic (Environmental) Risk Factors Mounting evidence demonstrates that cerebrovascular disorders may represent serious risk factors for AD. Similarly to genetically determined cases, AD provoked by vascular disorders develops very slowly. This slow progression complicates the understanding of connection between vascular disorders and AD. For example, there is a positive correlation between the risk for AD and midlife hypertension; however, no correlation is observed when the onset of hypertension occurs in old age (49Y51). The same is true for diabetes mellitus, which leads to cerebrovascular pathology. There is a positive correlation between AD and middle-age onset diabetes but no correlation between AD and late-onset diabetes (52, 53). People with chronic brain ischemia or atherosclerosis, as well people who survived strokes in middle age, also have a higher risk for AD (54Y57). Another nongenetic risk factor is brain trauma (58Y60). For example, retired boxers and football players are at high risk for AD-like pathology (61, 62). Because of its slow development, clinical manifestations of AD provoked by brain trauma usually appear only at an advanced age. However, severe head injuries at a young age may lead to early-onset AD, as detailed in the case described by Rudelli et al (63). This case deserves a detailed description because it demonstrates that pathologic processes underlying AD do not develop in the brain areas adjacent to the primary site of injury

194

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

but, rather, in the structures involved in memory formation and storage. Rudelli et al (63) described a 22-year-old steelworker who sustained a severe head injury accompanied by loss of consciousness. About 2 years later, after a substantial recovery of his cognitive functions, he was able to return to his job. Six years later, he displayed symptoms of progressive dementia and, 8 years later, he died. The examination of his brain demonstrated signs of a contusion necrosis in the left frontotemporal and right superior temporal areas, whereas the microscopic study diagnosed bilateral AD pathology (i.e. amyloid plaques and NFTs) in the hippocampus and associated areas of the frontal, temporal, parietal, and occipital cortices. His family history revealed no evidence that this case of early-onset AD could be attributed to genetic factors. Data on the spatiotemporal progression of AD pathology provoked by brain trauma are very scarce. To clarify this point, I wrote to one of the leading experts in the field, Professor Kurt Jellinger (Institute of Clinical Neurobiology, Vienna, Austria) and asked whether this progression is dependent on the location of the brain injury. Here is his answer: ‘‘Unfortunately, from our material, I cannot definitely answer your question. However, we did not find a relationship between location of brain injury and progression of AD pathology. On the other hand, we did not see any significant deviation of AD pathology versus classical forms’’ (reproduced with permission). The data described in this section indicate that AD cases caused by genetic and nongenetic risk factors share at least 2 common features. In both, the disease is characterized by a very long latency and similar patterns of spatiotemporal progression of AA and neurofibrillary pathologies. Furthermore, genetic and nongenetic factors may contribute synergistically to disease progression. For example, cerebrovascular disorders and brain trauma accelerate the onset of AD in the carriers of the ApoE4 isoform (59, 64, 65). These observations suggest that factors such as a brain injury, insufficient cerebral circulation, hyperinsulinemia, and others are unlikely to be a direct cause of AD pathology, as proposed by many authors (49, 51Y53, 55, 57). It is more likely that these factors serve as triggers initiating the pathologic process, which then develops according to its intrinsic mechanism. The existence of 2 independent groups of risk factors for AD (i.e. genetic and nongenetic) represents another unsolved puzzle of AD: why do these 2 groups of apparently dissimilar risk factors provoke the development of a similar pathology? I will attempt to answer this question in the next section.

THE BBB AND THE IMMUNE PRIVILEGE OF THE BRAIN Numerous morphologic and biochemical studies have demonstrated that AD is typically associated with disruption of the BBB (45, 54, 56, 66Y72). A common property of all nongenetic risk factors described above is that they cause BBB disruption (54, 56, 71, 72). The same is likely to apply to genetic factors because, in addition to AA deposition in the brain, they provoke cerebral amyloid angiopathy (CAA), which is characterized by AA deposition on the walls of cerebral blood vessels (8, 45, 70Y74). Recent studies demonstrated Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

that ApoE4-dependent AD is associated with capillary CAA but not with CAA restricted to larger vessels (73, 74). Capillary CAA, in turn, compromises BBB integrity. Blood-brain barrier degradation has also been found in numerous experiments on transgenic rodents (71, 72, 75). For example, BBB dysfunction was detected in triple-transgenic mice suffering from both amyloid and neurofibrillary pathologies (76). Therefore, one potential explanation for why genetic and nongenetic risk factors provoke similar patterns of AD pathology is that they all cause BBB dysfunction. The BBB is formed by cerebral capillary endothelial cells. To some extent, it acts as a physical barrier formed by the tight junctions between adjacent endothelial cells (77Y80). The presence of tight junctions distinguishes the vasculature of the brain from that in most other organs. It is commonly accepted that the BBB maintains the stable environment required for normal functioning of neuronal circuits. However, the BBB is an imperfect obstacle for penetration of small water-soluble molecules. For example, a large number of CNS-effective drugs easily permeate the BBB. In contrast, the BBB is mostly impermeable to macromolecules, including immunoglobulins (Igs). In addition, the BBB plays a major role in restricting the penetration of immunocompetent cells into the brain (77, 79, 81). Several studies indicated that only activated T lymphocytes can penetrate the intact BBB, as in the case of multiple sclerosis, an autoimmune disease of the nervous system (81). These data suggest that one of the most critical functions of the BBB is the isolation of the brain from the organism’s immune system. Combined with other mechanisms (the absence of a lymphatic system, low level of major histocompatibility complex molecule expression, local production of immunosuppressive factors, and others), the BBB provides immune privilege of the brain (82Y85). One function of the brain immune privilege widely discussed in literature is the protection of the brain from inflammation (82Y85). This function is particularly important for the organ whose cells lack the ability to regenerate. However, studies of immune privilege in other organs, such as the testis, suggest that protection from inflammation may not be the only function of the brain’s isolation from the immune system. The final differentiation of spermatozoa in the testis includes an expression of unique surface proteins that are required for fertilization. An important role in this process belongs to epigenetic mechanisms (86). Because the final differentiation of spermatozoa occurs at puberty, that is, long after immune self-tolerance is acquired during the perinatal period, these newly synthesized proteins exposed on the spermatozoa surface are recognized by the immune system as ‘‘nonself’’ antigens. This is thought to be a major reason why the spermatozoa are protected from the organism’s immune system. A significant role in this protection is played by the blood-testis barrier (82, 87, 88). Damage to this barrier leads to an autoimmune reaction against spermatozoa and, eventually, to male infertility. Based on this example, it is conceivable that the BBB isolates the brain from the immune system not only to protect it from inflammation but also because brain neurons express a subset of specific proteins after immune self-tolerance has been established. It may be further suggested that the expression

Alzheimer Disease and Memory

of these proteins occurs in the process of memory formation. Indeed, memorizing new facts and events means that the entering signals produce definite changes within the brain (memory engrams). Earlier, Pen˜a de Ortiz and I (89) hypothesized that these changes include the synthesis of proteins recognized by the immune system as ‘‘nonself’’ antigens. I will further elaborate on this hypothesis after a brief discussion of the cellular mechanisms of memory.

CELLULAR MECHANISMS OF MEMORY FORMATION AND STORAGE The Synaptic Plasticity Hypothesis According to the most commonly accepted view, traces of memory are stored within the brain through modifications in the strength of synaptic connections resulting in the formation of new patterns of neural activity. This hypothesis, suggested by Hebb (90), remains the prevalent hypothesis of memory today and guides most experimental efforts in the field. It is presumed that short-term memory is based on functional changes in synapses, whereas the consolidation of long-term memory requires protein synthesis and structural modifications in synaptic connections (91Y93). The general acceptance of the Synaptic Plasticity Hypothesis (SPH) seems to be rooted in the so-called connectionist concept, which is the main paradigm of modern neuroscience. According to this concept, all functions of the nervous system, including the highest cognitive functions of the brain, are performed at the level of neural networks. The complexity of these functions is determined merely by the complexity of networks, whereas single neurons are considered to serve as simple network elements whose role is reduced to the generation of electrical potentials and transmitting signals to other neurons. Comprehensive critique of the pure connectionist concept, as well as arguments that the higher functions of the brain may be performed primarily at the intraneuronal level, can be found elsewhere (94Y97). Clearly, the pure connectionist concept leaves little room to consider any mechanisms for memory formation and storage aside from modifications in strength of synaptic connections. Most of the data on the molecular mechanisms of memory formation and consolidation were obtained in behavioral assays using different variations of Pavlovian conditioning. Another major experimental approach to this problem is to study the electrophysiologic phenomenon known as longterm potentiation (LTP), which represents a prolonged (up to 24 hours) facilitation of synaptic transmission caused by a brief high-frequency (up to 100 Hz) stimulation of intrahippocampal pathways (98). Although LTP is evoked by an artificial electrical stimulation of input pathways with frequencies that rarely take place in the nervous system, this phenomenon is usually regarded as an experimental model of synaptic plasticity and long-term memory. Figure 2 is a simplified diagram of the proposed molecular reactions leading to the structural changes in synaptic connections formed by glutamatergic input fibers on the dendritic spines of hippocampal pyramidal cells (91Y93). Glutamate released by the endings of input fibers activates the

Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

195

Arshavsky

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

The SPH and AD Pathology According to the SPH, selective memory impairments observed at the preclinical stage of AD result primarily from disturbances in synaptic functions, especially synaptic plasticity. Therefore, the SPH predicts that the extracellular deposition of AA should lead to diminished LTP in the hippocampus and other brain structures that are involved in memory formation (100). For a long time, experimental verification of this prediction was not possible because AD is unique to humans and spontaneous AD-like pathology does not exist in animals. The situation has changed with the generation of genetically modified rodents that express human mutant APP and/or presenilins. These animals, which display amyloid plaque formation and reduced learning abilities, were used as a model to study hippocampal LTP during AD-like pathology (Fig. 3). Unexpectedly, the results obtained in these studies were contradictory. Several authors found reduced or abolished LTP in the

FIGURE 2. A simplified scheme of putative molecular processes leading to modifications in the strength of synaptic connections of hippocampal pyramidal neurons. AC, adenylate cyclase; CaM, calmodulin; cAMP, cyclic adenosine monophosphate; CBP, CREBbinding protein; CRE, cAMP-dependent element; CREB, cAMPresponsive element binding protein; Glu, glutamate; GluR, glutamate receptor; MAPK, mitogen-activated protein kinases; PKA, cAMP-dependent protein kinase.

glutamate receptors, which results in the Ca2+ entry into the cell. Increased intracellular Ca2+ triggers the cascade of reactions leading to the activation of several protein kinases, including the cAMP-dependent protein kinase (PKA) and the mitogen-activated protein kinases. Activated protein kinases translocate into the nucleus, where they phosphorylate the transcription factor CREB (cAMP-responsive element binding protein). Binding of CREB to the cAMP-dependent element (CRE) on the DNA leads to the transcriptional activation of ‘‘early regulatory genes.’’ This process requires the participation of the transcriptional coactivator CREB-binding protein. The products of early regulatory genes are also transcription factors regulating the expression of downstream genes. The final stages of this gene cascade expression remain mostly unknown. It is commonly suggested that they include the synthesis of proteins involved in the structural modifications of synaptic connections (Fig. 2). However, experimental data supporting this suggestion are sparse. On the other hand, the set of transcription factors involved in memory formation (e.g. CREB) does not overlap with the factors involved in the embryonic development of the brain (99). This opens the possibility that not all proteins synthesized in the process of memory formation are related to synaptic modifications.

196

FIGURE 3. Long-term potentiation (LTP) in hippocampal slices obtained from transgenic mice expressing human mutant amyloid precursor protein (APP). (A) Reduced LTP recorded in slices taken from 5- to 7-month-old APP/Ld/2 mice (open circles). The arrow indicates the beginning of high-frequency stimulation. Figure 5C from Moechars et al (102) (Reproduced with permission, Copyright 1999, the American Society for Biochemistry and Molecular Biology). (B) Enhanced LTP recorded in slices taken from 10-month-old B6.152H double-transgenic mice (filled circles). Picrotoxin (100 Kmol/L) was added to block F-aminobutyric acid A (GABAA) receptors. Figure 1A from Poirier et al (105) (Reproduced with permission. Copyright 2010, Elsevier). EPSP, excitatory postsynaptic potential. Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

hippocampal slices from these mice, which is consistent with the SPH predictions (Fig. 3A) (101Y103). However, this observation was not replicated by others who documented normal or increased LTP in mutant animals displaying the AD-like pathology (Fig. 3B) (104Y106). Ambiguous results were also obtained in studies of mice expressing different isoforms of human ApoE (107). Animals expressing the ApoE4 isoform were characterized by reduced LTP. Because this isoform is associated with an increased risk of AD, this observation supports the prediction of the SPH. However, LTP was also reduced in animals expressing ApoE2, which is not consistent with the SPH prediction because this ApoE isoform is associated with a decreased risk of AD. Thus, the studies of genetically modified animals did not provide definitive evidence that amyloid plaque formation leads to reduced synaptic plasticity, at least as judged from a decreased LTP.

The Permanence of Memory and the Genomic Hypothesis An unbiased analysis of the SPH suggests that, despite its general acceptance, it faces significant difficulties in explaining multiple data obtained in animal and human studies. Here, I will touch on the central aspect of this problem, the remarkable persistence of memory. Detailed critical analysis of the SPH can be found elsewhere (89, 108). One of the most important features of memory is its permanence. Everyone is aware of memory persistence from personal experience. Under the influence of accidental associations, we can remember facts, events, poems, music, and so on, that have not been recalled for decades. Our inability to recall certain facts and events does not necessarily mean that the corresponding information has been deleted from the brain. Facts and events that seemed to be entirely forgotten could be retrieved in a state of hypnosis (109) or in response to electric stimulation of the temporal lobe made for diagnostic purposes before or during brain surgery (110, 111). Cases of spontaneous memory recovery after prolonged retrograde amnesia caused by head trauma were described as well. To illustrate this point, I provide a detailed account of one of these cases (112). The patient (M.M.) was involved in a car accident that resulted in retrograde amnesia. He did not know who he was and failed to recognize his relatives, although he did not suffer from other neuropsychologic impairments, including anterograde amnesia. Recovery from retrograde amnesia occurred quite suddenly 1 month after the accident while he was playing tennis. He made a mistake and was struck by the awareness that he had made the same mistake in another match, years before. This triggered the recall of all the details of that match and then, in a matter of minutes, all of his memories came flooding back. Memories from his past life were now vivid and clearly loaded with emotional attributes. Taken together, these observations suggest that forgetting often results from the inaccessibility of memory traces for recollection even though they continue to be stored in the brain. The most impressive evidence for the ability of the human brain to acquire and permanently store enormous amounts of information has been obtained in studies of people

Alzheimer Disease and Memory

with an extraordinary memory (113Y115). For example, the recently deceased individual, Kim Peek, remembered by heart all approximately 9,000 books that he had read throughout his lifetime, including the first books that had been read to him by his parents when he was 18 months old. As emphasized by the authors who described Peek’s case, understanding the mechanism of memory must include an understanding of this remarkable phenomenon (113). One manifestation of memory persistence is its resistance to uncontrolled synaptic activities. The most impressive evidence that the mechanism of memory storage is protected from ongoing synaptic activity is the fact that memory can survive repeated epileptic seizures, which are accompanied by bursts of synchronous high-frequency discharges of neurons throughout wide areas of the cerebral cortex. If plasticity is an inherent property of synapses, each epileptic attack should lead to a chaotic perturbation of synaptic connections and, therefore, to the destruction of memory engrams. However, this is not the case. In fact, seizures lead only to the loss of memory relating to events immediately preceding an epileptic attack and do not affect consolidated memory (116, 117). The case of the patient known in neurophysiologic literature as H.M. clearly illustrates this point. At the age of 27, H.M. underwent a bilateral resection of his medial temporal lobe, which led to complete anterograde amnesia (17). This case played a pivotal role in understanding the physiology of memory because it showed that memory has a distinct anatomical localization in the brain (18). But I would like to emphasize another aspect of this case. H.M. was operated on to prevent strong epileptic fits that originated within the medial temporal lobe. The first small attacks appeared at the age of 10 and worsened over the years. For several years preceding the operation, severe seizures recurred about once a week and were accompanied by generalized convulsions. Studies of the memory of H.M. demonstrated that these strong epileptic seizures, which involved large areas of his brain and lasted for 17 years, did not produce a radical distortion of his memory. Not only did H.M. remember premorbid events but also many facts and events that occurred after the beginning of strong recurring epileptic fits. For example, he remembered his surgeon, Dr. Scoville, whom he had met a couple of years before the operation (17). The latter example shows that memories consolidated between the fits were not destroyed by subsequent epileptic discharges. The only way to explain such memory resistance to epileptic fits within the framework of the SPH is to assume that the synaptic connections that are modified in the process of memory formation are transformed from a plastic to a rigid state. However, this assumption runs into another difficulty. If that were the case, paroxysmal neural activity arising during epileptic seizures should also transform synaptic connections from a plastic to a rigid state, ultimately exhausting the brain’s capacity for learning. Yet, even epilepsy beginning in childhood does not impair the human ability to memorize new facts and events, as well as to acquire and store vast amounts of general and professional knowledge. The ability of memory to survive recurrent epileptic fits is the most impressive but not the only evidence illustrating the resistance of memory to synaptic activity. Another fact is its

Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

197

Arshavsky

resistance to constant spontaneous activity that is characteristic of neurons at higher levels of the brain. According to the SPH, this activity should introduce minimal but incessant distortions to newly formed synaptic connections and eventually to memory erosions. Again, we can assume that the synaptic connections modified in the process of memory formation are transformed from a plastic to a rigid state. But, in this case, spontaneous inputs would also transform synapses into a rigid state and, therefore, block the brain’s ability to learn. The latter difficulty of the SPH was first recognized by Eccles (118, p 57), who wrote, ‘‘The simple concept that disuse leads to regression of spine synapses and excess usage to hypertrophy can be criticized because I almost all cells I are discharging continuously. One can imagine therefore that there would be overall hypertrophy of all synapses under such conditions and hence no possibility of any hypertrophic change. Evidently, frequent synaptic excitation could hardly provide a satisfactory explanation of synaptic changes involved in learning.’’ Memory is resistant not only to uncontrolled synaptic inputs but also to various adverse events leading to strong suppression of neural and synaptic activity (general anesthesia, hypothermia, concussion, etc). For illustraclosed head injury at age 19. He remained in a coma for 1 to 2 weeks, followed by a vegetative state lasting for several months and then by a minimally conscious state lasting for 19 years. During this period, he was unable to communicate using gestures or verbal output and magnetic resonance imaging showed widespread cerebral atrophy. After 19 years in a minimally conscious state, the patient spoke his first word, ‘‘mom,’’ which was followed by the recovery of his consciousness during a period of several days. Remarkably, many of his memories were preserved. He recognized his relatives, regained the ability to speak and communicated efficiently with his family and doctors, could read single words, and so on. Taken together, these facts argue that permanent memory storage is unlikely to be based on such an unreliable foundation as synaptic plasticity. Memory has to be based on more conservative mechanisms protected from ongoing synaptic activity, pathologic epileptic activity, and other adverse factors. Crick (120) and Davis and Squire (121) independently proposed that permanent memory is coded at the genomic level in alterations of chromosomal DNA, the only truly stable molecule in the organism. Later, Holliday (122) hypothesized that the mechanism of long-term memory storage is based on DNA methylation. The idea that epigenetic mechanisms underlie memory formation and consolidation recently received significant experimental support (123Y128). Before describing these experimental results, I will briefly outline the role of epigenetic mechanisms in gene transcription (129, 130). In nondividing cells, DNA molecules are arranged in a 3-dimensional structure, chromatin. The basic elements of chromatin are nucleosomes (NS in Fig. 4) composed of DNA segments wrapped around histone octamers containing 2 copies of histones H2A, H2B, H3, and H4. Within this structure, only certain parts of the DNA molecule can be transcribed, whereas other parts are tightly packed and inaccessible by transcription factors and RNA polymerase. Gene expression in cells is controlled by epigenetic modifications of both DNA and associated histones. Histone acetylation renders gene promoters

198

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

accessible for the transcriptional machinery (Fig. 4A). Histone methylation can lead to either gene activation or repression, depending on which specific residues are methylated. Finally, histone phosphorylation is usually associated with an increase in gene transcription. Methylation of the DNA cytosines maintains chromatin in the condensed state (Fig. 4B), whereas DNA demethylation is a major epigenetic factor initiating gene transcription (Fig. 4A). Unlike the dynamic effects of histone modifications, the effects of DNA methylation/ demethylation are long lasting. Complex interactions between histone modifications and DNA methylation determine many aspects of cell-specific transcription programs. As a result, epigenetic mechanisms participate in the phenotypic differentiation of genotypically identical cells in multicellular organisms (131Y133). During embryogenesis, characteristic changes in the chromatin structure involved in determining the differentiation status of a cell may be induced by the transient action of specific signal molecules at definite periods of development. These changes in the chromatin structure, particularly those determined by changes in the DNA methylation, can remain stable over time and across cell divisions. Accordingly, the DNA methylation pattern is typically stable in terminally differentiated cells and the level of methyltransferase activity in these cells is low. Studies conducted during the past 20 years, particularly in the last decade, suggest that epigenetic mechanisms play an important role in the nervous system not only during the embryonic period but also throughout the entire lifetime. In contrast to most postmitotic somatic cells, certain types of brain neurons maintain large concentrations of DNA methyltransferases. This phenomenon, initially observed in mice (134, 135), was replicated in postmortem studies of fresh human brain

FIGURE 4. (A) Active state of chromatin characterized by histone acetylation and DNA demethylation. (B) Condensed state of chromatin characterized by DNA cytosine methylation and histone deacetylation. NS, nucleosome. Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

tissue (136). High levels of DNA methylation and DNA methyltransferase were found in the anterior and lateral temporal lobe of the neocortex throughout the entire life span. Studies completed during the past decade showed that epigenetic mechanisms play an important role in memory formation and consolidation. They revealed that memory formation is accompanied by multiple chromatin modifications in rodent hippocampal neurons, summarized as follows. 1. Histone Acetylation. The formation of contextual fear memory is accompanied by the acetylation of histone H3 (137), the latent inhibition training that blocked the formation of contextual fear memory by the acetylation of histone H4 (137), and the formation of spatial memory by a specific increase in H2B and H4 acetylation (138). These findings suggest that different types of memory are associated with different patterns of histone acetylation and, therefore, with activation of different genes. Histone acetylation is catalyzed by histone acetyltransferases that can act as coactivators of transcription factors; for example, a transcriptional coactivator of CREB, CREB-binding protein (Fig. 2), is a histone acetyltransferase (139). The elimination of the histone acetyltransferase activity and overexpression of histone deacetylases impair memory formation (139, 140). In contrast, the inhibition or focal deletion of histone deacetylases results in the facilitation of memory formation (137, 141). 2. Histone Phosphorylation. The formation of contextual fear memory is accompanied by the phosphorylation of histone H3 in area CA1 of the hippocampus. Both histone phosphorylation and histone acetylation are mediated by the mitogen-activated protein kinase signaling system (142). 3. Histone Methylation. The formation of contextual fear memory is associated with the methylation of histone H3 in hippocampal area CA1. The deletion of the H3K4-specific methyltransferase gene leads to a deficit in memory formation (143). 4. DNA Methylation. Fear conditioning is associated with rapid methylation and demethylation of different DNA loci within hippocampal neurons. The inhibition of DNA methyltransferases blocks memory formation (144). Taken together, these observations suggest that the cascade of reactions shown in Figure 2 leads to complex modifications of chromatin within hippocampal pyramidal neurons. Importantly, both histone acetylation (145) and DNA methylation (146) were documented not only in the hippocampus, which is involved in primary memory formation, but also in the prefrontal cortex, which is the ultimate memory repository. DNA methylation associated with contextual fear conditioning of mice persisted in neurons of the prefrontal cortex for at least 30 days, the maximum period tested by Miller et al (146). This study further revealed that DNA methylation plays a critical role in the long-term memory storage. An infusion of DNA methyltransferase inhibitors into the prefrontal cortex on the 29th day after conditioning impaired memory retrieval when tested the following day. Similarly, conditioned place-preference memory was destroyed by injecting DNA methyltransferase inhibitors in the prelimbic prefrontal cortex of mice (147).

Alzheimer Disease and Memory

It is conceivable that memory consolidation in humans is also associated with epigenetic chromatin modifications. As previously mentioned, postmortem analysis of the brain tissue demonstrated that both DNA methylation and DNA methyltransferase expression continue in the anterior and lateral temporal lobes of the neocortex throughout the entire life span (136). Robust DNA methyltransferase expression was found in layers III and V, where pyramidal neurons forming shortand long-distance cortico-cortical connections are localized. These neurons seem to be involved in persistent memory consolidation and are most vulnerable to AD pathology. Another postmortem study showed that DNA methylation patterns in neurons of the prefrontal cortex are much more variable than DNA methylation patterns in non-neuronal cells (148). A possible interpretation of such a difference is that high DNA methylation variability in neurons is associated with their involvement in memory consolidation and storage.

THE AUTOIMMUNE HYPOTHESIS OF AD: TESTING AND PERSPECTIVES The data described in the previous section show that input signals leading to memory formation and consolidation result in substantial chromatin modifications in memoryassociated neurons. This, in turn, is expected to activate the expression of certain genes and inactivate others. Therefore, it may be suggested that chromatin modulations associated with memory formation is accompanied not only by altered production of already existing proteins but also by the synthesis of specific proteins that were not present in these neurons before memory formation. For simplicity, I will call them ‘‘memory-specific proteins.’’ (I refrain from discussing any potential functions of presumed newly synthesized proteins in memory formation, consolidation, or retrieval. Human declarative memory is a cognitive function of the brain tightly connected to consciousness. The molecular mechanisms underlying higher cognitive functions are unknown, as are the molecular mechanisms of consciousness. I should stress, however, that this gap in our knowledge does not affect the overall logic of this review.) Because declarative memory is connected to consciousness and, therefore, begins to form after the establishment of immune self-tolerance, these memoryspecific proteins are expected to be recognized by the immune system as ‘‘nonself’’ antigens. As previously mentioned, this could be one of the critical reasons why the brain is an immuneprivileged organ, isolated from the immune system by the BBB. The next logical step is to assume that impairments of the BBB caused by either genetic or nongenetic AD risk factors allow the immune system to attack the neurons involved in memory formation and storage. The slow development of AD suggests that antigen concentration on the neuron surface is low and the autoimmune reaction is not strong enough for immediate neuron killing. But it may provoke memory-associated neurons to be particularly sensitive to AA and neurofibrillary pathologies, which are present to some extent in the brains of all elderly people (16). Neuron degeneration progressing in the brain not fully protected by the BBB leads to nonspecific brain inflammation. As previously mentioned, this inflammation results in complete breakdown of brain functions at later

Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

199

Arshavsky

stages of AD. This presumed sequence of events leading to AD is shown in Figure 5. This hypothesis suggests that AD starts with BBB impairment and a subsequent autoimmune reaction against neurons involved in memory formation. Although a significant body of evidence generally consistent with this hypothesis already exists in the literature, these data are sometimes contradictory among individual investigations and often allow multiple interpretations. Let us consider the BBB first. As previously described, BBB impairment is one of the characteristic manifestations of AD. According to many authors, CAA and BBB impairment are the primary events in AD pathogenesis (45, 56, 66, 67, 70Y72, 149). However, data on BBB status during AD were mainly obtained either in biochemical studies on patients displaying prominent AD symptoms or in postmortem studies. Therefore, the cause/effect relationship between BBB impairment and the development of AD remains elusive. Many investigators believe that a breach in the BBB is not the primary cause but, rather, a secondary result of the AD pathology. The hypothesis that AD starts as an impairment of the BBB seems to be supported by results obtained in experiments on transgenic mice. In mice expressing human mutant APP, BBB impairments preceded both cerebral amyloid plaque formation and memory defects (150, 151). These BBB impairments were caused by AA deposition on vessel walls (152). Similarly, the expression of human APOE4 favored the formation of CAA versus parenchymal amyloid plaques (153). However, these results should be treated cautiously because the AD-like pathology in transgenic mice does not precisely phenocopy AD in humans. The mechanism by which BBB damage could lead to AD also remains a subject of debate. The prevalent view is that defects in the BBB lead to the development of AD because of intrabrain penetration of blood-borne AA or/and impaired AA clearance (56, 71, 72, 154). This view is in full agreement with the amyloid hypothesis of AD. On the other hand, recent studies have shown that the progression of medial temporal lobe atrophy in AD patients is associated with increased BBB permeability and a higher cerebrospinal fluid/ serum Ig ratio, whereas AA transport is not influenced by BBB alteration (68, 155). Typical features of an adaptive immune response have also been found in the brains of AD patients, particularly in the medial temporal lobe. These features include the following: 1) An increased content of parenchymal Igs and Igpositive neurons (156Y158). The most detailed study was performed by D’Andrea (158), who hypothesized that AD is an autoimmune disease arising because of BBB impairments. However, in a subsequent study, D’Andrea and colleagues changed this interpretation and argued that BBB destruction leads to the entry of blood-borne AA into the brain where it binds selectively to neurons involved in memory formation (154). 2) Significant increases in numbers of T lymphocytes (20, 159Y163). 3) The activation of microglia (20, 161, 164, 165). Unlike neurons and macroglia, microglia have a mesodermal origin. Microglial progenitors derive from primitive macrophages that colonize the nervous system during the embryonic

200

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

FIGURE 5. The putative sequence of events leading to Alzheimer disease (AD) caused by the inheritance of the ?4 allele of the ApoE gene. The primary event is the A-amyloid (AA) deposition at the walls of cerebral capillaries (i.e. cerebral amyloid angiopathy), leading to blood-brain barrier (BBB) impairment. The AA deposits are represented by red ellipses; unaffected tight junctions between endothelial cells are marked by small arrows. The BBB impairment allows the immune system to attack specifically the neurons that are involved in memory formation and consolidation (purple neurons). This autoimmune attack makes memory-associated neurons vulnerable to AA deposition and the formation of neurofibrillary tangles (black bars). Four successive stages of neuron degeneration are shown from bottom to top. Eventually, progressive neuron degeneration initiates nonspecific brain inflammation, leading to the breakdown of all cognitive functions at later stages of AD. Modified Figure 1 from Arshavsky (5) (Reproduced with permission, Copyright 2010, IOS Press).

period (166, 167). In the adult brain, activated microglia may mediate the immune cellYspecific functions, including the functions of antigen-presenting cells (168, 169). 4) The expression of major histocompatibility complex class II molecules by reactive microglia (20, 159, 160, 164, 165). Overall, these findings are consistent with the hypothesis on the autoimmune origin of AD. These data were mainly obtained in postmortem studies, however, and do not allow for a definitive conclusion on the cause/effect relationship between immune reactions and AD. According to most common interpretation, the features of immune reactions found in the brains of AD patients have a secondary origin and display the development of inflammatory processes initiated by the formation of amyloid plaques and neuron death. Understanding the cause/effect relationship between BBB impairment, immune reactions in the brain, and AD pathology can be of fundamental importance for elucidating the primary mechanisms underlying this disease and, therefore, for the development of therapeutic approaches to prevent or delay its clinical stage. One productive approach to advancing this direction is to create a database of people who are carriers Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014

of at least one ?4 allele of the ApoE gene and, therefore, belong to the high-risk group. If any of these people dies from nonYAD-related causes before any AD symptom becomes evident, his or her brain would be examined to test 2 major predictions of the autoimmune hypothesis. One analysis would address the presence of capillary CAA and other manifestations of BBB dysfunction, and the other would seek evidence for signs of an immune reaction within the structures of the medial temporal lobe involved in memory formation. Obtaining consent from patients and family members should be feasible because there is a high probability that their relatives may also belong to the risk group. These studies would help establish whether BBB dysfunction and autoimmunity against the memory-associated neurons precede or follow the appearance of both AA and neurofibrillary pathologies. Should the hypothesis that the BBB impairment serves as the initial trigger of AD pathology be confirmed, it can radically change the strategy of fighting AD. Major therapeutic efforts would be focused on developing approaches to maintain BBB integrity in people belonging to AD risk groups rather than on treating these people when it is too late. Recent achievements in understanding the molecular mechanisms of BBB formation and maintenance suggest that such approaches are likely to be feasible. The formation of tight junctions between the endothelial cells of cerebral capillaries during embryonic development is induced by signals originating within the neural tissue (77). At least one of the signal systems involved was identified as the canonical Wnt/A-catenin pathway (78, 80, 170Y172). Wnt ligands (primarily Wnt7a/7b) secreted by brain precursor cells, mainly astrocytes, bind to the Frizzled/LRP (lipoprotein receptorYrelated protein) complexes on the surface of cerebrovascular endothelial cells (Fig. 6). This leads to the stabilization of intracellular A-catenin, which enters the nucleus to activate the expression of claudins, the membrane proteins that function as the primary seal for the tight junctions between endothelial cells. The results demonstrating the role of the Wnt7a/7b ligands in BBB formation in rodents were confirmed in experiments on human pluripotent stem cells codifferentiated with neural cells (173). The latest data suggest that the canonical Wnt/A-catenin signal system not only induces BBB development but also participates in maintaining its barrier functions in adult organisms, at least in the retina and cerebellum (174). The identification of the signal system inducing BBB formation and its subsequent stabilization may contribute to the development of therapeutic approaches to maintaining BBB integrity.

CONCLUSIONS By the beginning of the 21st century, medicine has achieved great successes that have significantly extended the average life span, at least in developed countries. Even for a complex medical condition such as cancer, today’s impressive breakthroughs in experimental and clinical oncology lead us to expect significant progress in the near future. Combined with advances in other fields of medicine, this will lead to an inevitable increase in human life span. As a result, AD may become the most challenging medical problem of this century. The number of AD patients in the United States is expected

Alzheimer Disease and Memory

FIGURE 6. The role of the canonical Wnt/A-catenin signaling system in blood-brain barrier (BBB) formation. FZD, Frizzled receptor; LRP5/6, lipoprotein receptorYrelated proteins 5 and 6; GSK3, glycogen synthase kinase 3; TCF, T cell factorYdependent transcription; JAM, junctional adhesion molecule; TJ, tight junction. Modified Figure 1 from Polakis (175) (Reproduced with permission, Copyright 2008, the Rockefeller University Press).

to increase from approximately 5 million to approximately 7.7 million by 2030 and to 11 to 16 million by 2050 (1). This prospect, along with an essential lack of progress in the treatment of AD, is extremely alarming. Clearly, this field is in urgent need of new ideas. In this review, I attempted to answer 2 puzzling questions related to the etiology of AD: why memory-associated neurons are selectively vulnerable to AD pathology, and why AD is provoked by apparently dissimilar genetic and nongenetic risk factors. These attempts have led me to argue that AD may have an autoimmune origin. According to this hypothesis, the primary cause of AD is a dysfunction of the BBB that allows the organism’s immune system to attack the neurons involved in memory formation, making them vulnerable to AA and neurofibrillary pathologies. The autoimmune hypothesis of AD is testable and, if confirmed, should radically change the whole tenor of the field. Significant effort should be concentrated on the development of therapeutic approaches that would improve the integrity of the ailing BBB. Recent achievements in understanding the molecular mechanisms of BBB formation are likely to facilitate the success of these efforts. ACKNOWLEDGMENTS I thank Dr. Vadim Arshavsky (Duke University, Durham, NC) for numerous stimulating discussions and critical suggestions on the manuscript, Dr. Allen Selverston (University of California, San Diego, CA) for his critical reading of the manuscript, Dr. Kurt Jellinger (Institute of Clinical Neurobiology, Vienna, Austria) for kindly sharing his clinical expertise, and Dr. Alexander Yarilin (Institute of Immunology, Moscow, Russia) for kindly sharing his expertise in immunology. REFERENCES 1. Brookmeyer R, Evans DA, Hebert L, et al. National estimates of the prevalence of Alzheimer’s disease in the United States. Alzheimer Dement 2011;7:61Y73

Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

201

Arshavsky 2. Huang Y, Mucke L. Alzheimer mechanisms and therapeutic strategies. Cell 2012;148:1204Y22 3. Zotova E, Bharambe V, Cheaveau M, et al. Inflammatory components in human Alzheimer’s disease and after active amyloid-A42 immunization. Brain 2013;136:2677Y96 4. Arshavsky YI. Alzheimer’s disease, brain immune privilege and memory: A hypothesis. J Neural Transm 2006;113:1697Y707 5. Arshavsky YI. Why Alzheimer’s disease starts with a memory impairment: Neurophysiological insight. J Alzheimer Dis 2010;20:5Y16 6. Braak H, Braak E, Bohl J, et al. Evolution of Alzheimer’s diseaseYrelated cortical lesions. J Neural Transm 1998;54(Suppl):97Y106 7. Braak H, Del Tredici K, Schultz C, et al. Vulnerability of select neuronal types to Alzheimer’s disease. Ann NY Acad Sci 2000;924:53Y61 8. Selkoe DJ. Alzheimer’s disease: Genes, proteins, and therapy. Physiol Rev 2001;81:741Y66 9. Sperling RA, Dickerson BC, Pihlajamaki M, et al. Functional alterations in memory networks in early Alzheimer’s disease. Neuromolecular Med 2010;12:27Y43 10. Gallagher M, Koh MT. Episodic memory on the path to Alzheimer’s disease. Curr Opin Neurobiol 2011;21:929Y34 11. Rapoport SI, Nelson PT. Biomarkers and evolution in Alzheimer disease. Prog Neurobiol 2011;95:510Y13 12. Braak H, Del Tredici K. Evolutional aspects of Alzheimer’s disease pathogenesis. J Alzheimer Dis 2013;33(Suppl 1):S155Y61 13. Braak H, Braak E. Frequency of stages of Alzheimer-related lesions in different age categories. Neurobiol Aging 1997;18:351Y57 14. Smith AD. Imaging the progression of Alzheimer pathology through the brain. Proc Natl Acad Sci USA 2002;99:4135Y37 15. Giannakopoulos P, Herrmann FR, Bussie`re T, et al. Tangle and neuron numbers, but not amyloid load, predict cognitive status in Alzheimer’s disease. Neurology 2003;60:1495Y500 16. Nelson PT, Alafuzoff I, Bigio EH, et al. Correlation of Alzheimer’s disease neuropathologic changes with cognitive status: A review of the literature. J Neuropathol Exp Neurol 2012;71:362Y81 17. Milner B, Squire LR, Kandel ER. Cognitive neuroscience and the study of memory. Neuron 1998;20:445Y68 18. Squire LR, Wixted JT. The cognitive neuroscience of human memory since H.M. Annu Rev Neurosci 2011;34:259Y88 19. Wyss-Coray T. Inflammation in Alzheimer disease: Driving force, bystander or beneficial response? Nat Med 2006;12:1005Y15 20. Zotova E, Nicoll JA, Kalaria R, et al. Inflammation in Alzheimer’s disease: Relevance to pathogenesis and therapy. Alzheimer Res Ther 2010;2:1 21. Braak H, Alafuzoff I, Arzberger T, et al. Staging of Alzheimer diseaseY associated neurofibrillary pathology using paraffin sections and immunocytochemistry. Acta Neuropathol 2006;112:389Y404 22. Stranahan AM, Mattson MP. Selective vulnerability of neurons in layer II of the entorhinal cortex during aging and Alzheimer’s disease. Neural Plast 2010;2010:108190 23. Nelson PT, Head E, Schmitt FA, et al. Alzheimer’s disease is not ‘‘brain aging’’: Neuropathological, genetic, and epidemiological human studies. Acta Neuropathol 2011;121:571Y87 24. Goate A, Hardy JA. Twenty years of Alzheimer’s diseaseYcausing mutations. J Neurochem 2012;120(Suppl 1):3Y8 25. Bateman RJ, Xiong C, Benzinger TL, et al. Clinical and biomarker changes in dominantly inherited Alzheimer’s disease. N Engl J Med 2011; 367:795Y804 26. Selkoe DJ. Preventing Alzheimer’s disease. Science 2012;337:1488Y92 27. Hardy JA, Higgins GA. Alzheimer’s disease: The amyloid cascade hypothesis. Science 1992;256:184Y85 28. Games D. Alzheimer-type neuropathology in transgenic mice overexpressing V717F beta-amyloid precursor protein. Nature 1995;373: 523Y27 29. Hsiao K, Chapman P, Nilsen S, et al. Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science 1996; 274:99Y102 30. McGowan E, Eriksen J, Hutton M. A decade of modeling Alzheimer’s disease in transgenic mice. Trends Genet 2006;22:281Y89 31. Morgan D, Diamond DM, Gottschall PE, et al. A beta peptide vaccination prevents memory loss in an animal model of Alzheimer’s disease. Nature 2000;408:982Y85

202

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014 32. Boche D, Denham N, Holmes C, et al. Neuropathology after active Abeta42 immunotherapy: Implications for Alzheimer’s disease pathogenesis. Acta Neuropathol 2010;120:369Y84 33. Delrieu J, Ousset PJ, Caillaud C. Clinical trials in Alzheimer’s disease: Immunotherapy approaches. J Neurochem 2012;120(Suppl 1):186Y93 34. Golde TE, Schneider LS, Koo EH. Anti-aA therapeutics in Alzheimer’s disease: The need for a paradigm shift. Neuron 2011;69:203Y13 35. Orgogozo JM, Gilman S, Dartigues JF, et al. Subacute meningoencephalitis in a subset of patients with AD after Abeta42 immunization. Neurology 2003;61:46Y54 36. Panza F, Frisardi V, Imbimbo BP, et al. Anti-A-amyloid immunotherapy for Alzheimer’s disease: Focus on bapineuzumab. Curr Alzheimer Res 2011;8:808Y17 37. Sperling R, Salloway S, Brooks DJ, et al. Amyloid-related imaging abnormalities in patients with Alzheimer’s disease treated with bapineuzumab: A retrospective analysis. Lancet Neurol 2012;11:241Y49 38. Moreth J, Mavoungou C, Schindowski K. Passive anti-amyloid immunotherapy in Alzheimer’s disease: What are the most promising targets? Immun Ageing 2013;10:18 39. Bancher C, Jellinger K, Lassmann H, et al. Correlations between mental state and quantitative neuropathology in the Vienna Longitudinal Study on Dementia. Eur Arch Psychiatry Clin Neurosci 1996;246:137Y46 40. Lee HG, Zhu X, Castellani RJ, et al. Amyloid-beta in Alzheimer disease: The null versus the alternate hypotheses. J Pharmacol Exp Ther 2007;321:823Y29 41. Josephs KA, Whitwell JL, Ahmed Z, et al. Beta-amyloid burden is not associated with rates of brain atrophy. Ann Neurol 2008;63:204Y12 42. Pimplikar SW, Nixon RA, Robakis NK, et al. Amyloid-independent mechanisms in Alzheimer’s disease pathogenesis. J Neurosci 2010;30: 14946Y54 43. Price JL, Morris JC. Tangles and plaques in nondemented aging and ‘‘preclinical’’ Alzheimer’s disease. Ann Neurol 1999;45:358Y68 44. Thal DR, Ru¨b U, Orantes M, et al. Phases of A beta-deposition in the human brain and its relevance for the development of AD. Neurology 2002;58:1791Y800 45. Donahue JE, Johanson CE. Apolipoprotein E, amyloid-beta, and blood-brain barrier permeability in Alzheimer disease. J Neuropathol Exp Neurol 2008;67:261Y70 46. Bu G. Apolipoprotein E and its receptors in Alzheimer’s disease: Pathways, pathogenesis and therapy. Nat Rev Neurosci 2009;10:333Y44 47. Kim J, Basak JM, Holtzman DM. The role of apolipoprotein E in Alzheimer’s disease. Neuron 2009;63:287Y303 48. Verghese PB, Castellano JM, Holtzman DM. Apolipoprotein E in Alzheimer’s disease and other neurological disorders. Lancet Neurol 2011;10:241Y52 49. Qiu C, Winblad B, Fratiglioni L. The age-dependent relation of blood pressure to cognitive function and dementia. Lancet Neurol 2005;48: 487Y99 50. Skoog I, Gustafson D. Update on hypertension and Alzheimer’s disease. Neurol Res 2006;28:605Y11 51. Kennelly SP, Lawlor BA, Kenny RA. Blood pressure and the risk for dementia: A double-edged sword. Ageing Res Rev 2009;8:61Y70 52. Schnaider Beeri M, Goldbourt U, Silverman JM, et al. Diabetes mellitus in midlife and the risk of dementia three decades later. Neurology 2004; 63:1902Y7 53. Xu W, Qiu C, Gatz M, et al. Mid- and late-life diabetes in relation to the risk of dementia: A population-based twin study. Diabetes 2009;58:71Y77 54. Jellinger KA. Alzheimer disease and cerebrovascular pathology: An update. J Neural Transm 2002;109:813Y36 55. De la Torre JC. Alzheimer’s disease is a vasocognopathy: A new term to describe its nature. Neurol Res 2004;26:517Y24 56. Pluta R, Amek MU. Brain ischemia and ischemic blood-brain barrier as etiological factors in sporadic Alzheimer’s disease. Neuropsychiatr Dis Treat 2008;4:855Y64 57. Zhang X, Le W. Pathological role of hypoxia in Alzheimer’s disease. Exp Neurol 2010;223:299Y303 58. Rasmusson DX, Brandt J, Martin DB, et al. Head injury as a risk factor in Alzheimer’s disease. Brain Inj 1995;9:213Y19 59. Plassman BL, Havlik RJ, Steffens DC, et al. Documented head injury in early adulthood and risk of Alzheimer’s disease and other dementias. Neurology 2000;55:1158Y66

Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014 60. Jellinger KA. Head injury and dementia. Curr Opin Neurol 2004;17: 719Y23 61. Tokuda T, Ikeda S, Yanagisawa N, et al. Re-examination of ex-boxers’ brains using immunohistochemistry with antibodies to amyloid beta-protein and tau protein. Acta Neuropathol 1991;82:280Y85 62. Guskiewicz KM, Marshall SW, Bailes J, et al. Association between recurrent concussion and late-life cognitive impairment in retired professional football players. Neurosurgery 2005;57:719Y26 63. Rudelli R, Strom JO, Welch PT, et al. Posttraumatic premature Alzheimer’s disease. Neuropathologic findings and pathogenetic considerations. Arch Neurol 1982;3:570Y75 64. Nicoll JA, Roberts GW, Graham DI. Apolipoprotein E. Epsilon 4 allele is associated with deposition of amyloid beta-protein following head injury. Nat Med 1995;1:135Y37 65. Irie F, Fitzpatrick AL, Lopez OL, et al. Enhanced risk for Alzheimer disease in persons with type 2 diabetes and APOE epsilon4: The Cardiovascular Health Study. Arch Neurol 2008;65:89Y93 66. Hardy JA, Mann DM, Wester P, et al. An integrative hypothesis concerning the pathogenesis and progression of Alzheimer’s disease. Neurobiol Aging 1986;7:489Y502 67. Desai BS, Monahan AJ, Carvey PM, et al. Blood-brain barrier pathology in Alzheimer’s and Parkinson’s disease: Implications for drug therapy. Cell Transplant 2007;16:285Y99 68. Matsumoto Y, Yanase D, Noguchi-Shinohara M, et al. Blood-brain barrier permeability correlates with medial temporal lobe atrophy but not with amyloid-beta protein transport across the blood-brain barrier in Alzheimer’s disease. Dement Geriatr Cogn Disord 2007;23:241Y45 69. Kumar-Singh S. Cerebral amyloid angiopathy: Pathogenetic mechanisms and link to dense amyloid plaques. Genes Brain Behav 2008; 7(Suppl 1):67Y82 70. Jeynes B, Provias J. The case for blood-brain barrier dysfunction in the pathogenesis of Alzheimer’s disease. J Neurosci Res 2011;89:22Y28 71. Zlokovic BV. Neurovascular pathways to neurodegeneration in Alzheimer’s disease and other disorders. Nat Rev Neurosci 2011;12:723Y38 72. Zlokovic BV. Cerebrovascular effects of apolipoprotein E: Implications for Alzheimer disease. JAMA Neurol 2013;70:440Y44 73. Thal DR, Papassotiropoulos A, Saido TC, et al. Capillary cerebral amyloid angiopathy identifies a distinct APOE epsilon4Yassociated subtype of sporadic Alzheimer’s disease. Acta Neuropathol 2010;120: 169Y83 74. Attems J, Jellinger K, Thal DR, et al. Review: Sporadic cerebral amyloid angiopathy. Neuropathol Appl Neurobiol 2011;37:75Y93 75. Nicolazzo JA, Mehta DC. Transport of drugs across the blood-brain barrier in Alzheimer’s disease. Ther Deliv 2010;1:595Y611 76. Mehta DC, Short JL, Nicolazzo JA. Reduced CNS exposure of memantine in a triple transgenic mouse model of Alzheimer’s disease assessed using a novel LC-MS technique. J Pharm Biomed Anal 2013; 85C:198Y206 77. Abbott NJ, Patabendige AA, Dolman DE, et al. Structure and function of the blood-brain barrier. Neurobiol Dis 2010;37:13Y25 78. Liebner S, Czupalla CJ, Wolburg H. Current concepts of blood-brain barrier development. Int J Dev Biol 2011;55:467Y76 79. Daneman R. The blood-brain barrier in health and disease. Ann Neurol 2012;72:648Y72 80. Siegenthaler JA, Sohet F, Daneman R. ‘‘Sealing off the CNS’’: Cellular and molecular regulation of blood-brain barrier genesis. Curr Opin Neurobiol 2013;23:1057Y64 81. Engelhardt B. The blood-central nervous system barriers actively control immune cell entry into the central nervous system. Curr Pharm Des 2008;14:1555Y65 82. Ferguson TA, Green DR, Griffith TS. Cell death and immune privilege. Int Rev Immunol 2002;21:153Y72 83. Pachter JS, de Vries HE, Fabry Z. The blood-brain barrier and its role in immune privilege in the central nervous system. J Neuropathol Exp Neurol 2003;62:593Y604 84. Wilson EH, Weninger W, Hunter CA. Trafficking of immune cells in the central nervous system. J Clin Invest 2010;120:1368Y79 85. Muldoon LL, Alvarez JI, Begley DJ, et al. Immunologic privilege in the central nervous system and the blood-brain barrier. J Cereb Blood Flow Metab 2013;33:13Y21 86. Trasler JM. Epigenetics in spermatogenesis. Mol Cell Endocrinol 2009; 306:33Y36

Alzheimer Disease and Memory 87. Fijak M, Bhushan S, Meinhardt A. Immunoprivileged sites: The testis. Methods Mol Biol 2011;677:459Y70 88. Li N, Wang T, Han D. Structural, cellular and molecular aspects of immune privilege in the testis. Front Immunol 2012;3:152 89. Pen˜a de Ortiz S, Arshavsky YI. DNA recombination as a possible mechanism in declarative memory: A hypothesis. J Neurosci Res 2001; 63:72Y81 90. Hebb DO. The Organization of Behavior. New York, NY: Wiley, 1949 91. Alberini CM. Transcription factors in long-term memory and synaptic plasticity. Physiol Rev 2009;89:121Y45 92. Byrne JH, LaBar KS, LeDoux JE, et al. Learning and memory: Basic mechanisms. In: Byrne JH, Roberts JL, eds. From Molecules to Networks: An Introduction to Cellular and Molecular Neuroscience, 2nd ed. Boston, MA: Academic Press/Elsevier, 2009:539Y608 93. Mayford M, Siegelbaum SA, Kandel ER. Synapses and memory storage. Cold Spring Harb Perspect Biol 2012;4:a005751 94. Arshavsky YI. Role of individual neurons and neural networks in cognitive functioning of the brain: A new insight. Brain Cogn 2001;46: 414Y28 95. Arshavsky YI. ‘‘Scientific roots’’ of dualism in neuroscience. Prog Neurobiol 2006;79:190Y204 96. Arshavsky YI. Two functions of early language experience. Brain Res Rev 2009;60:327Y40 97. Hawkins J, Blakeslee S. On Intelligence. New York, NY: An Owl Book, 2004 98. Bliss TV, LLmo T. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J Physiol (Lond) 1973;232:331Y56 99. Thiel G, ed. Transcription Factors in the Nervous System. Weinheim, Germany: Wiley-VCH Verlag, 2006 100. Walsh DM, Klyubin I, Fadeeva JV, et al. Amyloid-beta oligomers: Their production, toxicity and therapeutic inhibition. Biochem Soc Trans 2002;30:552Y57 101. Chapman PF, White GL, Jones MW, et al. Impaired synaptic plasticity and learning in aged amyloid precursor protein transgenic mice. Nat Neurosci 1999;2:271Y76 102. Moechars D, Dewachter I, Lorent K, et al. Early phenotypic changes in transgenic mice that overexpress different mutants of amyloid precursor protein in brain. J Biol Chem 1999;274:6483Y92 103. Ma T, Hoeffer CA, Capetillo-Zarate E, et al. Dysregulation of the mTOR pathway mediates impairment of synaptic plasticity in a mouse model of Alzheimer’s disease. PLoS One 2010;5:e12845 104. Jolas T, Zhang XS, Zhang Q, et al. Long-term potentiation is increased in the CA1 area of the hippocampus of APP(swe/ind) CRND8 mice. Neurobiol Dis 2002;11:394Y409 105. Poirier R, Veltman I, Pflimlin MC, et al. Enhanced dentate gyrus synaptic plasticity but reduced neurogenesis in a mouse model of amyloidosis. Neurobiol Dis 2010;40:386Y93 106. Volianskis A, KLstner R, MLlgaard M, et al. Episodic memory deficits are not related to altered glutamatergic synaptic transmission and plasticity in the CA1 hippocampus of the APPswe/PS1CE9-deleted transgenic mice model of A-amyloidosis. Neurobiol Aging 2010;31: 1173Y87 107. Trommer BL, Shah C, Yun SH, et al. ApoE isoform affects LTP in human targeted replacement mice. Neuroreport 2004;15:2655Y58 108. Arshavsky YI. ‘‘The seven sins’’ of the Hebbian synapse: Can the hypothesis of synaptic plasticity explain long-term memory consolidation? Prog Neurobiol 2006;80:99Y113 109. Cheek DB, LeCron LM. Clinical Hypnotherapy. New York, NY: Grune & Stratton, 1968 110. Penfield W. Some mechanisms of consciousness discovered during electrical stimulation of the brain. Proc Nat Acad Sci USA 1958;44:51Y66 111. Gloor P. Experiential phenomena of temporal lobe epilepsy. Facts and hypotheses. Brain 1990;113:1673Y94 112. Lucchelli F, Muggia S, Spinnler H. The ‘‘Petites Madeleines’’ phenomenon in two amnesic patients. Sudden recovery of forgotten memories. Brain 1995;118:167Y83 113. Treffert DA, Christensen DD. Inside the mind of a savant. Sci Am 2005; 293:108Y13 114. Parker ES, Cahill L, McGaugh JL. A case of unusual autobiographical remembering. Neurocase 2006;12:35Y49

Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

203

Arshavsky 115. Leport AK, Mattfeld AT, Dickinson-Anson H, et al. Behavioral and neuroanatomical investigation of Highly Superior Autobiographical Memory (HSAM). Neurobiol Learn Mem 2012;98:78Y92 116. Thompson PJ. Memory function in patients with epilepsy. Adv Neurol 1991;55:369Y84 117. Leritz EC, Grande LJ, Bauer RM. Temporal lobe epilepsy as a model to understand human memory: The distinction between explicit and implicit memory. Epilepsy Behav 2006;9:1Y13 118. Eccles JC. Possible synaptic mechanism subserving learning. In: Karcmar AG, Eccles JC, eds. Brain and Human Behavior. New York, NY: Springer-Verlag, 1972:39Y61 119. Voss HU, Ulu0 AM, Dyke JP, et al. Possible axonal regrowth in late recovery from the minimally conscious state. J Clin Invest 2006;116: 2005Y11 120. Crick F. Memory and molecular turnover. Nature 1984;312:101 121. Davis HP, Squire LR. Protein synthesis and memory: A review. Psychol Bull 1984;96:518Y59 122. Holliday R. Is there an epigenetic component in long-term memory? J Theor Biol 1999;200:339Y41 123. Levenson JM, Sweatt JD. Epigenetic mechanisms in memory formation. Nat Rev Neurosci 2005;6:108Y18 124. Roth TL, Roth ED, Sweatt JD. Epigenetic regulation of genes in learning and memory. Essays Biochem 2010;48:263Y74 125. Day JJ, Sweatt JD. Cognitive neuroepigenetics: A role for epigenetic mechanisms in learning and memory. Neurobiol Learn Mem 2011;96:2Y12 126. Molfese DL. Advancing neuroscience through epigenetics: Molecular mechanisms of learning and memory. Dev Neuropsychol 2011;36: 810Y27 127. Yu NK, Baek SH, Kaang BK. DNA methylationYmediated control of learning and memory. Mol Brain 2011;4:5 128. Zovkic IB, Guzman-Karlsson MC, Sweatt JD. Epigenetic regulation of memory formation and maintenance. Learn Mem 2013;20:61Y74 129. Gibney ER, Nolan CM. Epigenetics and gene expression. Heredity (Edinb) 2010;105:4Y13 130. Bannister AJ, Kouzarides T. Regulation of chromatin by histone modifications. Cell Res 2011;21:381Y95 131. Turek-Plewa J, Jagodzicski PP. The role of mammalian DNA methyltransferases in the regulation of gene expression. Cell Mol Biol Lett 2005;10:631Y47 132. Meshorer E. Chromatin in embryonic stem cell neuronal differentiation. Histol Histopathol 2007;22:311Y19 133. Zaidi SK, Young DW, Montecino M, et al. Architectural epigenetics: Mitotic retention of mammalian transcriptional regulatory information. Mol Cell Biol 2010;30:4758Y66 134. Brooks PJ, Marietta C, Goldman D. DNA mismatch repair and DNA methylation in adult brain neurons. J Neurosci 1996;16:939Y45 135. Feng J, Zhou Y, Campbell SL, et al. Dnmt1 and Dnmt3a maintain DNA methylation and regulate synaptic function in adult forebrain neurons. Nat Neurosci 2010;13:423Y30 136. Siegmund KD, Connor CM, Campan M, et al. DNA methylation in the human cerebral cortex is dynamically regulated throughout the life span and involves differentiated neurons. PLoS One 2007;2:e895 137. Levenson JM, O’Riordan KJ, Brown KD, et al. Regulation of histone acetylation during memory formation in the hippocampus. J Biol Chem 2004;279:40545Y59 138. Bousiges O, Vasconcelos AP, Neidl R, et al. Spatial memory consolidation is associated with induction of several lysine-acetyltransferase (histone acetyltransferase) expression levels and H2B/H4 acetylationY dependent transcriptional events in the rat hippocampus. Neuropsychopharmacology 2010;35:2521Y37 139. Korzus E, Rosenfeld MG, Mayford M. CBP histone acetyltransferase activity is a critical component of memory consolidation. Neuron 2004; 42:961Y72 140. Guan JS, Haggarty SJ, Giacometti E. HDAC2 negatively regulates memory formation and synaptic plasticity. Nature 2009;459:55Y60 141. McQuown SC, Barrett RM, Matheos DP, et al. HDAC3 is a critical negative regulator of long-term memory formation. J Neurosci 2011;31: 764Y74 142. Chwang WB, O’Riordan KJ, Levenson JM, et al. ERK/MAPK regulates hippocampal histone phosphorylation following contextual fear conditioning. Learn Mem 2006;1:322Y28

204

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014 143. Gupta S, Kim SY, Artis S, et al. Histone methylation regulates memory formation. J Neurosci 2010;30:3589Y99 144. Miller CA, Sweatt JD. Covalent modification of DNA regulates memory formation. Neuron 2007;53:857Y69 145. Bredy TW, Wu H, Crego C, et al. Histone modifications around individual BDNF gene promoters in prefrontal cortex are associated with extinction of conditioned fear. Learn Mem 2007;14:268Y76 146. Miller CA, Gavin CF, White JA, et al. Cortical DNA methylation maintains remote memory. Nat Neurosci 2010;13:664Y66 147. Han J, Li Y, Wang D, et al. Effect of 5-aza-2-deoxycytidine microinjecting into hippocampus and prelimbic cortex on acquisition and retrieval of cocaine-induced place preference in C57BL/6 mice. Eur J Pharmacol 2010;642:93Y98 148. Iwamoto K, Bundo M, Ueda J, et al. Neurons show distinctive DNA methylation profile and higher interindividual variations compared with non-neurons. Genome Res 2011;21:688Y96 149. Halliday MR, Pomara N, Sagare AP, et al. Relationship between cyclophilin a levels and matrix metalloproteinase 9 activity in cerebrospinal fluid of cognitively normal apolipoprotein e4 carriers and blood-brain barrier breakdown. JAMA Neurol 2013;70:1198Y200 150. Ujiie M, Dickstein DL, Carlow DA, et al. Blood-brain barrier permeability precedes senile plaque formation in an Alzheimer disease model. Microcirculation 2003;10:463Y70 151. Biron KE, Dickstein DL, Gopaul R, et al. Amyloid triggers extensive cerebral angiogenesis causing blood brain barrier permeability and hypervascularity in Alzheimer’s disease. PLoS One 2011;6:e23789 152. Kumar-Singh S, Pirici D, McGowan E, et al. Dense-core plaques in Tg2576 and PSAPP mouse models of Alzheimer’s disease are centered on vessel walls. Am J Pathol 2005;167:527Y43 153. Fryer JD, Simmons K, Parsadanian M, et al. Human apolipoprotein E4 alters the amyloid-beta 40:42 ratio and promotes the formation of cerebral amyloid angiopathy in an amyloid precursor protein transgenic model. J Neurosci 2005;25:2803Y10 154. Clifford PM, Zarrabi S, Siu G, et al. Abeta peptides can enter the brain through a defective blood-brain barrier and bind selectively to neurons. Brain Res 2007;1142:223Y36 155. Matsumoto Y, Yanase D, Noguchi-Shinohara M, et al. Cerebrospinal fluid/serum IgG index is correlated with medial temporal lobe atrophy in Alzheimer’s disease. Dement Geriatr Cogn Disord 2008;25:144Y47 156. Licandro A, Ferla S, Tavolato B. Alzheimer’s disease and senile brains: An immunofluorescence study. Rivis Patol Nerv Ment 1983; 104:75Y87 157. Singh VK, Fudenberg HH. Increase of immunoglobulin G3 subclass is related to brain autoantibody in Alzheimer’s disease but not in Down’s syndrome. Autoimmunity 1989;3:95Y101 158. D’Andrea MR. Evidence linking neuronal cell death to autoimmunity in Alzheimer’s disease. Brain Res 2003;982:19Y30 159. Itagaki S, McGeer PL, Akiyama H. Presence of T-cytotoxic suppressor and leucocyte common antigen positive cells in Alzheimer’s disease brain tissue. Neurosci Lett 1988;91:259Y64 160. Rogers J, Luber-Narod J, Styren SD, et al. Expression of immune system-associated antigens by cells of the human central nervous system: Relationship to the pathology of Alzheimer’s disease. Neurobiol Aging 1988;9:339Y49 161. Singh VK. Neuroautoimmunity: Pathogenic implications for Alzheimer’s disease. Gerontology 1997;43:79Y94 162. Togo T, Akiyama H, Iseki E, et al. Occurrence of T cells in the brain of Alzheimer’s disease and other neurological diseases. J Neuroimmunol 2002;124:83Y92 163. Town T, Tan J, Flavell RA, et al. T-cells in Alzheimer’s disease. Neuromolecular Med 2005;7:255Y64 164. Perlmutter LS, Scott SA, Barron E, et al. MHC class IIYpositive microglia in human brain: Association with Alzheimer lesions. J Neurosci Res 1992;33:549Y58 165. Gordon MN. Microglia and immune activation in Alzheimer’s disease. J Fla Med Assoc 1993;80:267Y70 166. Chan WY, Kohsaka S, Rezaie P. The origin and cell lineage of microglia: New concepts. Brain Res Rev 2007;53:344Y54 167. Ginhoux F, Greter M, Leboeuf M, et al. Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science 2010; 330:841Y45

Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

J Neuropathol Exp Neurol  Volume 73, Number 3, March 2014 168. Yang I, Han SJ, Kaur G, et al. The role of microglia in central nervous system immunity and glioma immunology. J Clin Neurosci 2010;17:6Y10 169. Sosa RA, Forsthuber TG. The critical role of antigen-presentationYinduced cytokine crosstalk in the central nervous system in multiple sclerosis and experimental autoimmune encephalomyelitis. J Interferon Cytokine Res 2011;31:753Y68 170. Liebner S, Corada M, Bangsow T, et al. Wnt/beta-catenin signaling controls development of the blood-brain barrier. J Cell Biol 2008;183:409Y17 171. Daneman R, Agalliu D, Zhou L, et al. Wnt/beta-catenin signaling is required for CNS, but not non-CNS, angiogenesis. Proc Natl Acad Sci USA 2009;106:641Y46

Alzheimer Disease and Memory 172. Stenman JM, Rajagopal J, Carroll TJ, et al. Canonical Wnt signaling regulates organ-specific assembly and differentiation of CNS vasculature. Science 2008;322:1247Y50 173. Lippmann ES, Azarin SM, Kay JE, et al. Derivation of blood-brain barrier endothelial cells from human pluripotent stem cells. Nat Biotechnol 2012; 30:783Y91 174. Wang Y, Rattner A, Zhou Y, et al. Norrin/Frizzled4 signaling in retinal vascular development and blood brain barrier plasticity. Cell 2012;151: 1332Y44 175. Polakis P. Formation of the blood-brain barrier: Wnt signaling seals the deal. J Cell Biol 2008;183:371Y73

Ó 2014 American Association of Neuropathologists, Inc.

Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

205