Alzforum: On the Relationship between Inflammation, Apolipoprotein E and Alzheimer's disease


	Current Papers
	ARF Recommends
	Milestone Papers
	Search All Papers
	Search Comments


	Research News
	Drug News
	Conference News


	AD Hypotheses
		SWAN Knowledge Base
		Current Hypotheses
		Hypothesis Factory
	Forums
		Live Discussions
		Virtual Conferences
		Interviews
	Enabling Technologies
		Workshops
		Research Tools
	Compendia
		AlzGene
		AlzRisk
		Antibodies
		Mutations
		Research Models
	Resources
		Bulletin Boards
		Conference Calendar
		Grants
		Jobs


	Overview
	Diagnosis/Genetics
	Research
	News
	Profiles
	Clinics


	Companies
	Tutorial
	Drugs in Clinical Trials


	About Alzheimer's
		FAQs
	Diagnosis
		Clinical Guidelines
		Tests
		Brain Banks
	Treatment
		Drugs and Therapies
	Caregiving
		Patient Care
		Support Directory
		AD Experiences
	Expert Opinion
		Ask the Expert


	Member Directory
	Researcher Profiles
	Institutes and Labs


	Mission
	ARF Team
	ARF Awards
	Advisory Board
	Sponsors
	Partnerships
	Fan Mail
	Support Us

Home: Research: Forums

On the Relationship between Inflammation, Apolipoprotein E and Alzheimer's disease

Keith A. Crutcher
Department of Neurosurgery
University of Cincinnati
crutchka@email.uc.edu
http://crutcherlab.med.uc.edu

Overview

Apolipoprotein E (apoE) exhibits an isoform-specific association with Alzheimer's disease (AD) such that the E4 isoform segregates with a higher risk of the disease. A similar isoform association of apoE exists with dementia pugilistica and there may be an association between E4, herpes infection and the risk of AD. Several hypotheses have been put forward to explain the association of apoE with AD based on an indirect role for apoE in the pathological cascade. An alternative hypothesis is that this association arises from the fact that these conditions are associated with inflammation and that apoE, perhaps produced by microglia, contributes to neurodegeneration through direct neurotoxic effects. This hypothesis is based on several lines of evidence, some of which is reviewed in this poster.

Considerations

1) Peptides derived from the receptor binding domain of apoE are toxic to neurons (Figures 1&2)

Figure 1: Diagram of the region of apoE from which synthetic peptides were derived for studies of neurotoxicity
The sequence from human apoE of amino acids 130-169 is shown. This spans position 158 (indicated by the red R), which is occupied by cysteine in the E2 isoform. Otherwise, the sequence is identical for all three common apoE isoforms. The receptor binding domain is thought to primarily involve the sequence from 130 to approximately 151. A heparin binding domain is included within this region. Two different synthetic apoE peptides that have been used for toxicity studies are tandem duplicates of either the sequence from 141-148 (the short tandem peptide or S-Tan) or of the sequence from 141-155 (the long tandem peptide or L-Tan). A third synthetic peptide includes the monomeric sequence from 130 to 169. More detailed information regarding these peptides can be found in references 7 and 9.

Figure 2: Dose-Dependent Toxic Effects of ApoE Peptides Tested on Chick Sympathetic Neurons
Cultures of chick sympathetic neurons were exposed to increasing concentrations of three different synthetic peptides derived from the receptor binding domain of apoE (see Figure 1). The cells were exposed to peptide E(141-149)², E(141-155)² or E130-169 for 20 hours. This graph shows the number of surviving cells as revealed by vital dye staining. Dose-response curves of peptides showed half-maximal toxic concentrations ranging from 1-4 µM. Similar dose-response curves were obtained with cultures of dissociated chick cortical neurons and rat hippocampal neurons following exposure to peptide E(141-149)². The full description of these results can be found in Tolar et al., J. Neurosci. 17:5678-5686, 1997.

2) The N-terminal 22 kDa thrombin-cleavage product of apoE (truncated apoE) is neurotoxic with the E4 isoform being more toxic than the E3 isoform (Figure 3)

Figure 3: Isoform-specific neurotoxicity of truncated apoE
Chick sympathetic neurons were exposed to different concentrations of the 22 kDa N-terminal domain of apoE derived through thrombin cleavage of full-length apoE3 or apoE4 (expressed in HEK cells). The truncated E4 isoform was significantly more toxic than the truncated E3 isoform under the conditions used here. These results suggest that truncated apoE can be neurotoxic and that there is an isoform difference in neurotoxicity that parallels the apoE isoform-associated risk of Alzheimer's disease. Full details of these studies can be found in reference 30.

3) Full-length apoE exhibits neurotoxicity, again with the E4 isoform being more toxic than the E3 isoform, and this toxicity is correlated with the appearance of truncated apoE in the media (Figures 4&5)

Figure 4: Full-length apoE exhibits isoform-specific neurotoxicity
Chick sympathetic neurons were exposed to different concentrations of recombinant full-length human apoE3 or apoE4 (expressed in HEK cells). Toxicity was observed with both isoforms but the E4 isoform was effective at lower concentrations. The toxicity was associated with the appearance of a lower molecular weight band (approx. 22 kDa) corresponding to truncated apoE (shown in figure 5).

Figure 5: ApoE gives rise to a 22 kDa N-terminal portion in vitro
The medium from the experiment shown in figure 4 was collected and probed for the presence of apoE species using Western blotting. Lane 1 shows the medium before exposure to the cells and contained 0.5 µM of purified apoE3, which shows up at the appropriate molecular weight (34 kDa). Lane 2 demonstrates the staining associated with medium collected from cells that had been exposed to 6 µM of the same apoE3 preparation shown in lane 1. In addition to the full-length apoE there is another band running at an approximate molecular weight of 22 kDa, corresponding to truncated apoE. Lane 3 shows medium collected from cels that had been exposed to purified apoE4 at a concentration of 2 µM. Again, there is evidence for the appearance of truncated apoE that was not in the initial preparation. These results suggest that the toxicity observed with full-length apoE may be due to the generation of truncated apoE. More complete description of these results is found in reference 28.

4) Truncated apoE is present in the brain and CSF (Figure 6)

Figure 6: Truncated apoE is present in human brain tissue
This Western blot shows apoE immunoreactivity within tissue samples collected from AD and non-AD postmortem brains. The immunoreactive bands are identified with the monoclonal antibody 1D7, which recognizes the receptor-binding domain (Figure 1) of apoE. In addition to the full-length apoE, there is another relatively prominent band running with an apparent molecular weight of 22 kDa in all of the samples, suggesting that this truncated apoE may normally be produced in the brain. There is no obvious difference in the staining intensity of the bands on this Western blot, either between AD and control samples or as a function of apoE genotype for the AD samples. However, these samples were not standardized for protein content and the determination of whether there are disease-related or genotype-related differences in the level of apoE or truncated apoE will require further studies.

5) Protease inhibitors provide significant protection against the neurotoxic effects of apoE and inhibit the production of truncated apoE. These results will be presented at the Society for Neuroscience meeting (reference 29).

6) Medium collected from cultures of neonatal rat glial cells (including astrocytes and microglia) contain low molecular weight apoE fragments, including what appears to be truncated apoE. These findings will be presented at the Society for Neuroscience meeting (reference 21).

How are these observations related?

AD appears to involve inflammation, including the presence of activated microglia, which have been shown to produce neurotoxic substances. The epidemiological evidence for an isoform-specific role of ApoE in neurodegeneration is supported by in vitro studies documenting neurotoxic effects of apoE and the possible role of truncated apoE in such toxicity. Microglia-derived apoE copuld participant in this inflammatory response and contribute to the neurite degeneration and neuronal death that occurs in Alzheimer's disease and other diseases where inflammation is involved.

Evidence that AD involves an inflammatory process:

1) History of non-steroidal anti-inflammatory drug use confers protection against AD [3, 4, 34, 44, 49]

2) Numerous markers of inflammation are upregulated in AD brain [11, 12, 32, 33, 45]

3) Activated microglia are present in the AD brain, often in association with senile plaques [35, 53, 54, 56]

Epidemiologic evidence for a role of apoE in neurodegeneration:

1) ApoE shows an isoform-association with AD such that the E4 allele confers a greater risk [ 1 , 5, 6, 8, 15, 24, 25, 43, 47, 51]

2) ApoE also shows a similar isoform-association with pugilistic dementia and outcome after head injury [20 , 31]

3) Latent herpes simplex infection and apoE4 represent compounding risk factors for AD [18, 19]

In vitro evidence that apoE can cause neurotoxicity:

1) Synthetic peptides derived from the receptor binding domain of apoE are toxic to lymphocytes and neurons [7, 9]

2) 22 kDa N-terminal (truncated) apoE exhibits isoform-specific neurotoxicity [30]

3) Full-length apoE also gives rise to isoform-specific neurotoxicity [28]

Evidence that apoE neurotoxicity involves truncated apoE:

1) Truncated apoE is present in the brain and CSF [10, 30]

2) The in vitro toxicity of full-length apoE is accompanied by the appearance of truncated apoE [28, 29]

3) Protease inhibitors reduce production of truncated apoE and neurotoxicity [29]

Evidence that CNS glial cells, including microglia, produce neurotoxic substances, including apoE:

1) Senile plaque components, such as amyloid, can activate microglia [13, 14, 36, 48]

2) ApoE is produced in astrocytes and microglia [21, 37a, 40, 46, 54, 55]

Hypotheses regarding the role of apolipoprotein E in Alzheimer's Disease

Several different hypotheses have been proposed to explain the isoform-specific association of apoE with the risk of Alzheimer's disease. For the most part, these hypotheses rely on the assumption that amyloid is the primary neurotoxic agent in the disease and/or that apoE3 provides a positive function that is performed less well by the E4 isoform. Some representative examples of these hypotheses are as follows:

1) ApoE, through isoform-specific binding to amyloid, affects the accumulation, clearance, aggregation and/or toxicity of amyloid [22, 23, 26, 42, 43, 51, 57]

2) ApoE plays a role in stabilizing microtubules through isoform-specific interactions with tau or other microtubule-associated proteins [2, 17, 27, 50, 52]

3) ApoE exhibits isoform-specific effects on neurite outgrowth, thereby affecting the extent to which compensatory neuronal sprouting can occur in response to injury [16, 38, 39, 41]

4) ApoE may serve as an antioxidant, with the E4 isoform being less effective [37]

Hypothesis

An alternative hypothesis is that ApoE plays a direct role in neurodegeneration through the generation of neurotoxic truncated apoE or other apoE fragments. The greater toxicity of truncated apoE4 as compared with truncated apoE3 confers the isoform-specific risk for Alzheimer's disease and other disorders in which inflammation may play a role. Activated microglia are a likely source of the apoE in such conditions. Therefore, AD, pugilistic dementia and herpes encephalitis may all represent conditions in which there is greater risk of neurodegeneration due to inflammation and the apoE isoform-associated risk is possibly due to differences in the potency and/or amount of neurotoxic truncated apoE or related fragments produced by glia, especially microglia.

Predictions

Regardless of the cellular source of the apoE, this hypothesis predicts that truncated apoE should be produced in, and associated with, areas of neuropathology in AD. A further prediction is that inheritance of the E4 allele might increase the risk of morbidity in conditions where chronic inflammation occurs. It is possible, for example, that macrophages, which produce apoE, play a comparable role in inflammatory conditions outside the CNS.

Acknowledgements

Portions of the work described here have been supported by the NIH, the Alzheimer's Association and the Mayfield Clinic. The following individuals have contributed significantly to the work reviewed here as well as providing critical input to the hypothesis: Dr. Marcos A. Marques, Dr. Martin Tolar, Dr. Judith A.K. Harmony, and Shubhada Joshi. The HEK cell lines used to produce apoE were a generous gift from Dr. M.-J. LaDu, University of Chicago. The assistance of Dave Pettigrew in the preparation of this poster is gratefully acknowledged.

References