Innate Immunity Autotoxicity and Degenerative Neurologies

Super Memory Formula

Natural Dementia Cure and Treatment

Get Instant Access

PATRICK L. McGEER, KOJI YASOJIMA AND EDITH G. McGEER

The lesions of established Alzheimer's disease (AD) are characterized by the presence of a broad spectrum of inflammatory molecules (for review, see Neuroinflammation Working Group, 2000). They include complement proteins and their regulators, inflammatory cytokines, acute-phase reactants, and numerous proteases and protease inhibitors. Neurons, astrocytes and microglia participate in their production. Several of these inflammatory products are known to be toxic to neurons, providing a rational basis for the hypothesis that neuroinflammation is a major contributing factor to the pathogenesis of AD (McGeer and Rogers, 1992).

This hypothesis has been greatly strengthened by reports that polymorphisms in promoter and other untranslated regions of the inflammatory cytokine genes IL-1a, IL-1p, IL-6 and TNFa significantly enhance the odds ratios for contracting AD (Bagli et al., 2000a,b; Bhojak et al., 2000; Collins et al., in press; Du et al., 2000; Grimaldi et al., 2000; Nicoll et al., 2000; Papassotiropoulos et al., 1999; Rebeck, in press; see other chapters, this volume). These data indicate that the level of expression of inflammatory cytokines influences the onset of AD. It has been additionally strengthened by further epidemiological studies supporting earlier findings (McGeer et al., 1996) that chronic use of non-steroidal antiinflammatory drugs (NSAIDs) reduces the risk of contracting AD. Stewart et al. (1997), in the Baltimore longitudinal study, found that patients who were using NSAIDs for more than 2 years had the risk of AD reduced by 60% (Figure 31.1). Veld et al. (2000), in the Rotterdam aging study, after verifying NSAID use by prescription records, found that the reduction in AD following more than 2 years' use of NSAIDs was 80%, even higher than that reported from the Baltimore study. The sparing was close to that originally found in our study of rheumatoid arthritics, who had presumably been taking NSAIDs and other inflammatory drugs for many years (McGeer et al., 1990) (Figure 31.1).

Rheumatoid arthritics (McGeeret al., 1990)

Rheumatoid arthritics (McGeeret al., 1990)

Figure 31.1. Odds ratios for the development of AD noted in three studies of populations taking NSAIDs as compared to that in age-matched controls (Stewart et al., 1997; Veld et al., 2000; McGeer et al., 1990). See McGeer et al. (1996) for data from other epidemiological studies

Thus, there is a strong correlation in AD between post mortem pathology showing neuroinflammatory damage, genetic polymorphisms influencing inflammatory cytokine upregulation, and epidemiological studies revealing sparing of disease amongst those using antiinflammatory drugs.

The absence of antibodies and minimal presence of lymphocytes indicate that the elements of a classical autoimmune disease are absent in AD tissue. The neuroinflammation must then be viewed as evidence of an innate immune response which has self-damaging effects. We define this phenomenon as 'autotoxicity' in order to distinguish it from 'autoimmunity'.

We describe here some of the characteristics of autotoxicity as found in AD tissue. Among the most important are the complement system and activated microglia. The tangles and plaques of AD are marked by the opsonizing complement components, C4d and C3d, and dystrophic neurites in AD brain are immunostained for the membrane attack complex (C5b-9) (Neuroinflammation Working Group, 2000). This lytic molecule is designed to destroy foreign bacteria and viruses. There are multiple protective molecules which defend host cells against self-attack. However, if the concentration of the membrane attack complex exceeds the levels of defensive proteins in host tissue, then self-attack can occur in a process called bystander lysis. It is the smoking gun of autodestruction in AD, since dystrophic neurites are observed immunostained for this complex.

It is of great interest to know which cells are responsible for generating the complement proteins and how they become activated. Neurons (Terai et al., 1997), astrocytes and microglia (Neuroinflammation Working Group, 2000)

and even endothelial cells (Klegeris et al., 2000) are producers of complement components, so there are many potential sources for the activated complement fragments in brain. Neurons appear to be the most abundant generators. As far as activation is concerned, p-amyloid protein and the pentraxins, amyloid P and C-reactive protein, have all been shown to activate the classical pathway in vitro (Neuroinflammation Working Group, 2000). All are associated with AD senile plaques. We have recently shown that neurons are also the major producers of amyloid P and C-reactive protein (Yasojima et al., in press), and it has long been assumed that p-amyloid protein comes from neurons. Thus, neurons are the source of both the complement proteins and the activators that ultimately result in self-attack by the membrane attack complex.

Microglia constitute a few percent of the glia of brain and are normally in a quiescent state. When activated by an insult or injury to the brain, they change their shape and upregulate a large number of proteins, such as the complement receptor CD11b and the major histocompatibility class II glycoprotein HLA-DR. Activated microglia also produce massive amounts of oxygen radicals and other materials which may, in themselves, damage host cells. In vitro, cultured microglia secrete materials that are directly toxic to neurons (Klegeris and McGeer, 2000; Giulian et al., 1996). Thus, the complement system and activated microglia are both potential sources of neuronal destruction.

In order to estimate the intensity of inflammatory reactions, we have measured the levels of the mRNAs for complement proteins (Yasojima et al., 1999), two microglial markers (HLA-DR and CD11b) and the pentraxins, C-reactive protein and serum amyloid P, in normal and AD brain, as well as in osteoarthritic joints, atherosclerotic plaques, infarcted hearts and adjacent normal tissue. We have also measured levels in the liver and other peripheral organs. To obtain such data, total RNA is extracted from post mortem tissue, and the relative levels of mRNAs determined by RT-PCR. Messenger RNAs have been found to be surprisingly stable so that useful data can be obtained from post mortem tissue (Yasojima et al., 1999; Johnson et al., 1986; Morrison and Griffin, 1981). The results indicate an innate inflammatory reaction in pathological material from each of these diseases.

In AD brain, the regional upregulation is highly related to the degree of pathological involvement. In areas with a heavy burden of plaques and tangles, such as the entorhinal cortex, hippocampus and temporal cortex, there are very large increases in the mRNAs, while only small increases are observed in areas such as the caudate or cerebellum, with little pathological involvement (Yasojima et al., 1999). Figure 31.2A illustrates this for complement C1q and C9, Figure 31.2B for HLA-DR and CD11b, and Figure 31.2C for C-reactive protein and amyloid P. It was long thought that these last two were only synthesized in the liver but we have now shown that

A. Complement mRNAs

"o

>

60

<

s

40

>

Is

PS

CD Alzheimer-Clq

□ Control-Clq ^ Alzheimer-C9

CD Alzheimer-Clq

□ Control-Clq ^ Alzheimer-C9

Hippocampus Midtemporal Cerebellum

Caudate

C. C-Reactive protein (CRP) and serum amyloid P (SAP)

'148

Hippocampus Midtemporal Cerebellum

Liver

Activated microglial markers

□ Control-CDllb

Hippocampus Midtemporal Cerebellum Caudate

Activated microglial markers

□ Control-CDllb

Hippocampus Midtemporal Cerebellum Caudate

Figure 31.2. (A) Relative levels (mean+SEM) of the mRNAs for the complement proteins Clq and C9 in AD and control hippocampus, midtemporal gyrus, caudate and cerebellum. Data on all the complement mRNAs in 11 regions of AD and control brains have been published (Yasojima et al., 1999). (B) Relative levels of the mRNAs for the markers of activated microglia CD lib and HLA-DR in AD and control hippocampus, midtemporal gyrus, caudate and cerebellum. (C). Relative levels (mean+SEM) of the mRNAs for C-reactive protein (CRP) and serum amyloid P (SAP) in AD and control hippocampus, midtemporal gyrus, cerebellum and liver the mRNAs exist in brain and are upregulated in AD in pathologically affected regions (Yasojima et al., in press). Figure 31.2C also illustrates that the levels of the mRNAs for C-reactive protein and serum amyloid P were unchanged in AD liver as compared to controls. There was also no difference in the mRNA levels for either pentraxin in AD heart, spleen or kidney, as compared to the corresponding control tissues (Yasojima et al., in press).

Upregulation of the mRNAs for these inflammatory markers is also found in osteoarthritic joints, atherosclerotic plaques and infarcted hearts (Figure 31.3). The initial causes of pathology are quite different—plaques and tangles in AD, anoxia in heart disease, too much fat in atherosclerosis, and mechanical grinding in osteoarthritis. But the secondary process is common to all. The innate immune system may be playing a role in all of these conditions, and the level of its involvement can be measured by the general technique described. However, judging from the mRNA levels of these markers (Figure 31.4), the degree of inflammation is as large or larger in AD hippocampus than in any of these peripheral conditions (Figure 31.4). The mRNA levels for these markers in liver, spleen and kidney were unaffected in any of these diseases, emphasizing the local nature of the innate immune response (Yasojima et al., 1998, 1999).

The pentraxins and complement are ancient host defense mechanisms that can trace their lineage back at least as far as the horseshoe crab. They are key components of the innate immune system. The widespread ability of mammalian tissue to generate these proteins is consistent with this role. By contrast, the adaptive immune system, which is an invention of vertebrates, relies on the cloning of lymphocytes. Generation of cells is restricted to peripheral lymphatic organs. Thus, in any inflammatory process, the role of the innate immune system must be carefully evaluated to determine the extent to which it is involved.

Agents which reduce the inflammatory burden should be beneficial not only in AD but in cardiovascular disease, arthritis and other diseases in which chronic inflammation is prominently associated with the pathology. Cyclo-oxygenase (COX) inhibitors (NSAIDs) are a known class that reduces inflammation by lowering prostaglandin production. The epidemiological data reported to date (McGeer et al., 1990, 1996; Stewart et al., 1997; Veld et al., 2000; Figure 31.1) are based on the use of traditional NSAIDs, which are

2 B6

H lnflamed-C1q

B Normal-C1q

E2 Inflamed-C9

Arteries

Heart

A. Complement proteins and CRP

Joints

H lnflamed:HLA-DR

H Normal:HLA-DR

0 lnflamed-CD11b

B. Activated microglial markers

Arteries

Heart

Joints

Figure 31.3. (A) Relative levels of the mRNAs for Clq and C9 in osteoarthritic joints, atherosclerotic plaques and infarcted heart tissue as compared with those in normal joints, normal artery and normal heart tissue. (B) Relative levels of the mRNAs for HLA-DR and CD lib in osteoarthritic joints, atherosclerotic plaques and infarcted heart tissue as compared with those in normal joints, normal artery and normal heart tissue. Data for CRP mRNA levels are also given for normal artery and atherosclerotic plaques. All tissues were post mortem except for the joint tissue, which was surgically removed because of intractable pain

Figure 31.4. Comparison of the levels of the mRNAs for C1q, C9, HLA-DR and CD11b in AD hippocampus, osteoarthritic joints, atherosclerotic plaques and infarcted heart tissue

either inhibitors of COX-1 or mixed inhibitors of COX-1 and COX-2. Considerable interest has been shown in the possible use of selective COX-2 inhibitors because of their lesser gastrointestinal effects. However, COX-2 inhibitors may not be useful because this enzyme shows a high constitutive expression in neurons (McGeer, 2000). Animal experiments suggest that COX-2 may be performing adaptive functions associated with normal neurons and protective functions associated with stressed neurons. COX-1 is found in activated microglia. MacKenzie and Munoz (1998), in assessing the level of microglial activation in post mortem human brain, found that patients taking traditional NSAIDs had significantly fewer activated microglia. This occurred in those with and without tangles and plaques, indicating that such NSAIDs generally reduce the degree of microglial activation. Thus, COX-1 inhibitors may be superior to COX-2 inhibitors for the treatment of AD.

NSAIDs inhibit cyclooxygenase and thereby interfere with the production of prostaglandins. Prostaglandins are known to be inflammatory mediators, although far from the most powerful ones. NSAIDs therefore strike at fringe players in this whole autotoxic and inflammatory scheme. Analysis of lesion-associated inflammatory molecules in AD and other inflammatory conditions suggests more promising therapeutic targets. These include inhibitors of complement production, release and activation, blockers of inflammatory cytokines and anaphylotoxins, and diminishers of microglial activation.

Was this article helpful?

0 0
Unraveling Alzheimers Disease

Unraveling Alzheimers Disease

I leave absolutely nothing out! Everything that I learned about Alzheimer’s I share with you. This is the most comprehensive report on Alzheimer’s you will ever read. No stone is left unturned in this comprehensive report.

Get My Free Ebook


Post a comment