Non-Neuronal Cells in Alzheimer's Disease

The following sections are included:

• Foreword

• List of Contributors

• Contents

Changes in the structure and function of the blood-brain barrier (BBB) have been proposed to underlie the findings in aluminium (Al) toxicity and Alzheimer's disease (AD). If Al does have a causal role in AD, that role might be mediated through the effects of the metal ion on the BBB. Different approaches usually taken in the study of a neurotoxin and of a clinical syndrome have led to different characterizations that present difficulties in direct comparisons of BBB function between the two conditions. Nevertheless, as reviewed here, Al and Alzheimer's disease seem associated with similar profiles regarding changes in BBB integrity and transmembrane diffusion, but not with cerebral blood flow. Saturable transport systems are selectively altered in each condition, but no single transport system has been studied in both conditions. Future studies could allow more distinct comparisons if experimental designs for studying one condition take into account the findings and approaches toward study of the other condition. In particular, studies with Al need to consider parameters such as genetics, metal speciation, aging, and hormonal effects, whereas the BBB in AD needs to be more completely characterized with regard to transmembrane diffusion and saturable transport systems.

Biochemical markers for Alzheimer's disease (AD) would be of great value, both to improve the clinical diagnostic accuracy in scientific studies and drug trials, and to increase the knowledge of the pathogenesis of the disorder. Analysis of cerebrospinal fluid (CSF) gives the most direct opportunity to study biochemical changes in the brain in the clinically relevant stages of AD. Since lumbar puncture is a gentle procedure, with minimal risk for discomfort or complications, it can easily be included as part of scientific studies. Biochemical markers for AD should reflect the degeneration of the neurones and their synapses, and the increased number of senile plaques (SPs) and neurofibrillary tangles (NFTs) compared with that found in non-demented individuals of the same age. However, since the neuropathological findings in AD have low specificity — they also occur in other disorders and in normal aging — there is little hope of ever finding a single biochemical marker specific for AD. Therefore, an approach is presented where several CSF biochemical markers are used together. A single biochemical marker is applied to represent a pathological process (such as neuronal degeneration), while a set of biochemical markers is applied to diagnose AD. These CSF biochemical markers are divided into "positive" markers, intended to identify the underlying pathological processes in AD, and "negative" markers, intended to identify pathological processes that are unrelated to AD, such as cerebrovascular and ischemic lesions. The list of potential positive CSF-biochemical markers includes βA4 protein and amyloid precursor protein (APP) for amyloid deposition, hyper-phosphorylated tau proteins for neurofibrillary tangles, neuron-specific enolase for neuronal degeneration, and Chromogranin A and synaptotagmin synaptic degeneration. Potential negative CSF-biochemical markers are sulfatide for white-matter lesions (leukoaraiosis), and the CSF/serum albumin ratio for intracerebral small-blood vessel arteriolosclerosis.

Several studies suggest that the microviscosity of cell membranes obtained from patients with Alzheimer's disease (AD) differs from that of control subjects. Because metals had previously been implicated in the etiology of AD, it was postulated that metals would hold the key with regard to the difference in microviscosity. Metal ions have been found to affect the microviscosity of membranes in vivo as well as in vitro. In order to explain these effects, which are often paradoxical, the following hypothesis is put forward: the effect of a specific metal on the microviscosity of a membrane depends on the resultant effect that the metal has on both the hydrophilic surface domain and the internal lipid domain of the membrane. On the membrane surface, metal cations can bind to the phosphate groups of the phospholipids. Cations such as calcium and zinc may bind to more than one phospholipid, drawing them closer together, thereby increasing the membrane microviscosity. Conversely, in the case of cations such as aluminium, each ion would bind to only one phospholipid, leaving residual positive charges. By repulsion the phospholipids would move away from each other, thereby decreasing the microviscosity of the membrane. Metal concentrations in vivo have also been found to influence the fatty acid composition of membranes, increasing or decreasing the degree of unsaturation; the higher the degree of unsaturation, the less viscous a membrane becomes. Additionally, the microviscosity of membranes may be indirectly affected by metals such as iron or copper, which initiate, and aluminium, which potentiates, lipid peroxidation.

Alzheimer's disease (AD) and cellular aging are intimately interwoven entities. In patients with AD, the erythrocyte aging process seems to be disturbed. This aging process revolves around structural and functional changes in band 3 (or AE1), the erythroid member of the anion exchanger (AE) gene family. AE1 and at least two other AE proteins are also present in human neurons in the central nervous system. The neuronal expression level of one or more AE proteins increases with age, with an additional increase observable in degenerating neurons in brain areas affected by AD. The AD-related changes in AE processing in neurons bear many similarities with aging-related processing of AE1 in erythrocytes. A better understanding of the physiological function of AE proteins in the human brain, and the changes in these functions during normal neuronal aging, may be instrumental in understanding the causes leading to neuronal death and degeneration.

The following sections are included:

• INTRODUCTION

• PLATELET CHANGES IN AD

• ERYTHROCYTE CHANGES IN AD

• CONCLUSIONS AND FUTURE DIRECTIONS

• REFERENCES

Study of the function and metabolism of APP in peripheral systems such as platelets and kidney may shed light on understanding the mechanism of amyloid deposition in Alzheimer's disease. High levels of APP are expressed in both tissues. APP exists as both full-length and soluble forms in platelets and is processed in a similar way as established in nucleated cells. Platelet APP is secreted upon activation and may be involved in the regulation of coagulation or wound healing. The function of APP in the kidney is unknown.

Alzheimer's disease is considered an age-related disorder and not a form of accelerated aging. However, because of the absence of direct experimental evidence, this point remains controversial. At least three different groups have investigated the growth characteristics of cultured skin fibroblast lines from FAD patients and sex-age-matched controls at different passages preceding the onset of senescence. We have recently performed a kinetic study of the replicative capacity at different culture times by [³H]-thymidine incorporation and crystal violet staining. Data showed no significant difference between the two groups at any studied passage. The life-span of the two types of cultures was also comparable. All these results provide direct experimental evidence that Alzheimer's disease is not simply a form of accelerated aging.

Alzheimer's disease (AD) affects millions of people worldwide, and therefore is of critical importance as a public health issue. Fundamental insight into the underlying molecular mechanisms leading to brain cell death and formation of senile plaques has recently been obtained from our laboratory [Hensley et al. (1994), Proc. Nat. Acad. Sci. USA 91:3270–3274; Butterfield et al. (1994), Biochem. Biophys. Res. Commun. 200:7100–715]. Previous research employing electron paramagnetic resonance spectroscopy described alterations in the physical state of cytoskeletal proteins in non-neuronal erythrocyte membranes in AD, pharmacologic modulation of cytoskeletal protein-protein interactions by promising AD therapeutic agents, and choline transport anomolies in AD red cells. This paper reviews these findings of membrane alterations related to AD.

Mitochondria appear to play an important role in aging and diseases of aging.

Mitochondria are critical subcellular organelles, which carry out a number of functions. These include the generation of available energy from oxidative metabolism, generation of a transmitochondrial potential (electromotive force), sequestration of calcium under pathological conditions, and generation of free radicals because of inevitable thermodynamic constraints during electron transport.

Mitochondria and mitochondrial damage has been proposed to play a role in aging and degeneration, both by excess free-radical generation and by impairment of the ability of cells to maintain and restore homeostasis when a challenge overcomes their ability to mobilize the necessary energy. Evidence for mitochondrial damage exists in a number of degenerative diseases associated with aging. Current evidence indicates the existence of an inherent abnormality in mitochondria in AD. Recent data raise the possibility of a genetically determined alteration in the Krebs tricarboxylic acid component, α-ketoglutarate dehydrogenase, in a subgroup of AD patients.

To explore the role of glia in cerebral amyloidosis, highly purified microglia and astrocyte cultures from human fetal brain were probed for the presence of amyloid precursor protein (APP) with immuno-cytochemistry, Western blots and Northern blots. APP mRNA was detected in both microglia and astrocytes. In neither cell type did APP mRNA levels appear significantly affected by endotoxin, interleukin-1b or tumor necrosis factor. Microglia and astrocytes were immunoreactive with a panel of antibodies recognizing epitopes spanning APP. Western blots with antibodies to carboxy-terminus (APPc) and amino-terminus (22C11) demonstrated that astrocytes and microglia contained full length APP (110-116 kd). APP was more readily detected in astrocytes than microglia, where relatively more of the APP had a lower molecular weight consistent with protein lacking post-translational modification. In astrocytes, higher molecular weight peptides predominated; in addition, there were lower molecular weight peptides of about 70, 30 and 15 kd. The 30 and 15 kd peptides reacted with APPc, but not 22C11. The results suggest that human glia produce APP and that astrocytes may contribute to the production of potentially amyloidogenic fragments of APP. Although the evidence does not suggest that microglia are major producers of Ab, they may nevertheless be involved in processing APP or reacting to tissue amyloid deposits.

Neuroglia comprise the largest class of non-neuronal or non-excitable cells in brain, and are some 10 times more numerous than neurons. They are a heterogenous population of two classes, microglia and macroglia. All types of glia are critical elements in brain inflammatory reactions, producing factors that are capable of inducing inflammation. In this paper, we review markers of inflammation that have been found in the Alzheimer's disease (AD) brain, and the relationship of glia to the expression and function of these markers.

Microglial cells are clearly associated with the pathological lesions of Alzheimer's disease. They occur in and around the β amyloid protein of senile plaques and cerebral blood vessels and close to extracellular neurofibrillary tangles. They are present as amoeboid microglia and probably perform a reactive role(s). They may play a part in the production and organization of fibrillary amyloid, the phagocytosis of neural and neuronal debris, participate in acute phase responses, and complement activation and presentation of antigen or release of neurotoxins. As yet a definitive role remains to be determined. However, modulation of potentially harmful effects of microglial action may yield therapeutic rewards.

Many studies have yielded conflicting results regarding the toxicity of Aβ, the peptide which is the principal component of senile plaques in brains of patients with Alzheimer's disease. Using in vitro and in vivo models, we have studied the role of glial cells and extracellular matrix molecules in mediating the effects of Aβ. Glial cells respond to Aβ substrate by accumulating and depositing chondroitin sulfate proteoglycans (CSPGs) which are inhibitory to neurite outgrowth. CSPGs are present around the senile plaque core, an area with both dystrophic neurites and a general decrease in normal neurites. We suggest that CSPG may contribute to the pathology by leading to regenerative failure of neurites surrounding senile plaques.

A variety of inflammatory proteins has been identified, particularly in the vicinity of β-amyloid plaques, in the brain of patients with Alzheimer's disease. Current findings indicate that these molecules are involved in a number of key steps in the pathological cascade. Activation products of classical complement pathway Clq, C4 and C3, but not the alternative pathway factor properdin, have been demonstrated in diffuse and classical plaques. In addition, complement receptors (CR3 and CR4) bearing cells are found within classical plaques. There are contradictory findings about the question whether or not the late complement proteins that can form the membrane attack complex, are present in Alzheimer's disease brains.

In this chapter we review the findings indicating that complement activation in amyloid plaques does not proceed further than C3. The idea is discussed that in Alzheimer's disease, complement does not function as an inflammatory mediator through membrane attack complex formation, but through the action of early complement activation products.

The following sections are included:

• INTRODUCTION

• NEURONAL Aβ DEPOSITS

• NON-NEURONAL Aβ DEPOSITS

• SUMMARY

• ACKNOWLEDGEMENTS

• REFERENCES

Instrumental neutron activation analysis was used to determine levels of up to 10 minor and trace elements in samples of kidney, liver, heart, lung and spleen taken from 24 Alzheimer's disease (AD) and 16 age-matched control subjects from the Central Kentucky region. In addition, particle-induced gamma-ray emission and particle-induced X-ray emission were utilized to determine 14 elements in bone samples from 11 AD and 17 age-matched control subjects. AD-control imbalances were observed and are discussed with reference to elemental imbalances observed in neural tissues. Other recent literature reports relating to elemental concentrations in non-neural AD tissues and body fluids are reviewed.

Index