Aside from its crystal-clear status as top risk gene for late-onset Alzheimer’s disease (LOAD), much mystery still shrouds apolipoprotein E. Three new papers investigate the protein through disparate lenses, shedding light on its role in neurodegenerative disease at the level of networks, cells, and genetics. One study solidifies the view that seniors carrying one ApoE4 allele, which raises their LOAD risk three- to fourfold, can lose connectivity in functional networks without (or before) cognitive decline or brain atrophy. Clifford Jack of Mayo Clinic, Rochester, Minnesota, led that research, published May 9 in the Archives of Neurology online. In the May 11 Journal of Neuroscience, a report by Michael Heneka, University of Bonn, Germany, and colleagues shows that ApoE is critical for Aβ clearance by microglia. And, possibly broadening the impact of ApoE on the genetics front, a small case-control study in this month’s print issue of the Archives of Neurology suggests the ApoE locus may be linked to frontotemporal dementia and primary progressive aphasia.

Functional magnetic resonance imaging (fMRI) has uncovered differences between cognitively normal ApoE4 carriers and non-carriers in the default-mode network—a set of brain areas that fire up when the mind is adrift, and tone down during goal-oriented activity. ApoE4 carriers in their late fifties and early sixties show default-mode abnormalities when their minds are resting (ARF related news story on Sheline et al., 2010) and while they focus on cognitive tasks (Fleisher et al., 2005). Even young ApoE4 carriers between the ages 20 to 35 show default-mode connectivity changes (ARF related news story on Filippini et al., 2009), suggesting that the E4 variant somehow causes this network to falter years before initial signs of memory loss.

In the online Archives of Neurology report, first author Mary Machulda and colleagues used task-free fMRI to analyze 56 E4 carriers and an equal number of age-, education-, and gender-matched non-carriers, all at least 63 years old (median age 79) and cognitively normal. The researchers not only examined connectivity in default-mode areas, which lie toward the back of the brain, but also in the salience network. This is a set of frontal structures that activate when the mind is focused and disengage during daydreaming—just the opposite of the default-mode network. In AD patients, the salience network becomes inappropriately active and default-mode activity wanes when the mind is at rest; the opposite happens in people with frontotemporal dementia (FTD).

Relative to non-carriers, ApoE4 carriers in the current study showed increased connectivity in the salience network and decreased connectivity in the default-mode network. The default mode overlaps with the brain areas that become hypometabolic and accumulate amyloid in AD (see ARF related news story on Buckner et al., 2005); hence, the low connectivity in this network makes sense, Jack told ARF. “It’s a direct manifestation of deinnervation in those areas.”

Why connectivity increases in the salience network of E4 carriers is harder to explain. Jack proposed two ideas. According to one theory, the phenomenon might be beneficial—the frontal areas become more interconnected to compensate for early local pathology. Alternatively, the increased frontal connectivity in E4 carriers could reflect their impaired ability to appropriately switch between default-mode and salience activation, leading to a decoupling of these networks. (For more on decoupling, see ARF conference story on multimodal imaging.)

“This is a very solid…paper that shores up our understanding of where and when ApoE4 begins to exert its susceptibility to Alzheimer’s disease,” Michael Greicius of Stanford University, Palo Alto, California, e-mailed to ARF (see full comment below). In his view, the report’s most important finding was the lack of differences in gray matter density between E4 and non-E4 groups. These data suggest that “resting-state functional connectivity may be a more sensitive measure of preclinical disease than structural MRI,” he wrote.

Adam Fleisher of the Banner Alzheimer’s Institute, Phoenix, Arizona, agreed that the findings suggest fMRI measures might prove useful as an AD biomarker. However, “a lot of work has to be done to see if these will be reasonable outcome measures for clinical trials,” he told ARF in a phone interview. “We need longitudinal data to really understand how these networks are changing over time, and in relationship to disease.”

At a mechanistic level, how does ApoE4 cause default-mode areas to lose connectivity? David Holtzman of Washington University in St. Louis, Missouri, thinks it is because a much larger proportion of ApoE4-positive seniors have high brain amyloid, relative to age-matched E4 non-carriers (see Morris et al., 2010). “ApoE4 is driving earlier Aβ deposition,” Holtzman wrote. “This occurs first in the default-mode network, and I would hypothesize that something associated with this process leads to dysfunction,” he wrote in an e-mail to ARF. “Thus, dysfunction in connectivity is seen first there.”

The current paper does not report participants’ brain amyloid levels. However, in Jack’s view, “it’s a certainty that our E4 carriers had, on average, higher Aβ burden than the non-carriers. So, to some extent, the effects we’re seeing are indeed related to increased amyloid deposition in ApoE4 carriers,” he said. However, Jack noted that past fMRI studies uncovering functional connectivity changes in this subgroup only looked at seniors without brain Aβ (Sheline et al., 2010), or examined young people who were highly unlikely to have appreciable brain amyloid (Filippini et al., 2009). These studies indicate that “over and above ApoE4’s effect on increasing amyloid deposition in the brain, E4 has independent effects on functional connectivity,” Jack said.

Given that Aβ contributes to ApoE4’s effect on network connectivity, a key question becomes, Why do E4 allele carriers have higher Aβ burden? The Journal of Neuroscience paper may offer a potential explanation. In this study, Heneka and colleagues reduced brain amyloid levels in AD transgenic mice (APP23) using a compound that binds and activates liver X receptors (LXRs), which regulate inflammation and cholesterol metabolism. The findings jibe with previous work showing that LXR agonists enhance Aβ clearance in cultured cells and AD mouse models (see ARF related news story on Jiang et al., 2008), and that knocking out LXRα or LXRβ intensifies plaque load in APP/PS1 mice (see ARF related news story on Zelcer et al., 2007). The present study went further, demonstrating that “microglial Aβ phagocytosis is under the control of astrocytes,” Heneka said. Led by first author Dick Terwel, the researchers showed that LXR agonist-treated astrocytes encouraged brain phagocytes to gobble up fibrillar Aβ, but not if the astrocytes came from ApoE- or LXRα-deficient mice. They conclude that microglial Aβ clearance relies on astrocytic LXRα activation and astrocytic ApoE.

In collaboration with Philip Verghese in the Holtzman lab, the University of Bonn researchers showed that ApoE lipidation in astrocytes is mediated by LXRα. These data, coupled with previous work suggesting that lipidated ApoE3 binds Aβ better than lipidated E4 (Tokuda et al., 2000), argue that “E4 carriers have lipidated ApoE with decreased affinity toward Aβ, and therefore decreased microglial clearance of pathological Aβ peptides,” Heneka said.

Extending similar reasoning to tau, Jack wondered if people with certain ApoE variants may be less effective than non-carriers at resisting the effects of toxic tau proteins. If true, this may agree with a genotype analysis by Carlo Masullo and colleagues at Catholic University School of Medicine in Rome, Italy. As reported in this month’s Archives of Neurology, first author Davide Seripa and colleagues genotyped 86 people with sporadic FTD (of whom 32 patients had primary progressive aphasia [PPA]), and 99 non-demented controls. The FTD and PPA groups had a three- to fourfold higher proportion of ApoE3/E4 participants compared with the control group. The PPA group also had a 12-fold overrepresentation of the ApoE2/E4 genotype. In addition, the researchers found several polymorphisms at the ApoE promoter region associating with both diseases. Links to ApoE have turned up in a few studies of Parkinson’s disease, as well (Blázquez et al., 2006; Buchanan et al., 2007; see others in PDGene). And for those still hungry for more on ApoE, the New York Academy of Sciences is hosting a one-day conference on May 24.—Esther Landhuis

Comments

  1. These findings from Cliff Jack’s group are of considerable interest for a number of different reasons. First, they confirm the importance and the strength of resting fMRI to detect gene-related (in this case ApoE) brain functional differences. Secondly, they suggest that changes in the default-mode network (DMN), previously observed during pathological aging (Greicius et al., 2004; Sorg et al., 2007), may be particularly useful in detecting individuals at risk of developing neurodegenerative disorders. Finally, paired with our previous reports in younger healthy subjects (Filippini et al., 2009), they support our recent observation that overactivity of brain function in young ApoE ε4-carriers is disproportionately reduced with advancing age even before the onset of measurable memory impairment (Filippini, 2011). Thus, the ApoE genotype seems to have different consequences for brain function depending on age. These age-related changes in brain function may reflect the increased vulnerability of ApoE ε4-carriers to late-life pathology or cognitive decline. However, caution is urged. Indeed, while ApoE is the best-established genetic risk factor associated with Alzheimer’s disease, its mechanism of action has not been clearly understood yet. More research is needed to define the role played by ApoE on modulating brain function.

    References:

    . Default-mode network activity distinguishes Alzheimer's disease from healthy aging: evidence from functional MRI. Proc Natl Acad Sci U S A. 2004 Mar 30;101(13):4637-42. PubMed.

    . Selective changes of resting-state networks in individuals at risk for Alzheimer's disease. Proc Natl Acad Sci U S A. 2007 Nov 20;104(47):18760-5. PubMed.

    . Distinct patterns of brain activity in young carriers of the APOE-epsilon4 allele. Proc Natl Acad Sci U S A. 2009 Apr 28;106(17):7209-14. PubMed.

    . Differential effects of the APOE genotype on brain function across the lifespan. Neuroimage. 2011 Jan 1;54(1):602-10. PubMed.

    View all comments by Nicola Filippini
  2. In this paper, Terwel et al. show that microglial uptake of fibrillar Aβ is reliant upon ApoE. They nicely show that this effect relies upon astrocytic liver X receptors (LXRs), consistent with the idea that LXR agonists act to elevate ApoE-containing HDL levels in the medium. An important finding is that this effect is LXRα dependent. This is significant since this isoform is inducible by PPARγ, and in humans is auto-induced.

    One new observation is that LXR agonist treatment enhances microglia association with plaques in an LXRα-dependent manner, although the basis of this effect is unclear.

    Overall, this paper adds new depth to the idea that LXRs and lipidated ApoE-containing HDLs play important roles in the normal clearance of fibrillar forms of amyloid, and that pharmacological targeting of LXRs is likely to be of therapeutic benefit.

    View all comments by Gary Landreth
  3. This is a very solid paper (over 50 subjects per group) that shores up our understanding of where and when ApoE4 begins to exert its influence on Alzheimer’s disease (AD). Though the sample selection and methods differ to some degree, the findings are largely in keeping with another recent study (Sheline et al., 2010) in which healthy older controls with an E4 allele showed reduced connectivity in default-mode network regions. The study by Sheline et al. examined slightly younger subjects (mean age around 60, compared to around 80 here) who had negative amyloid imaging scans. The current study does not include information on amyloid imaging, but given the advanced age of the group, we can assume that somewhere between 30 and 50 percent of the subjects would have positive amyloid imaging scans.

    Both studies touch on the dorsal anterior cingulate cortex (ACC), a key node in a second network, dubbed the salience network, which appears to exist in a dynamic equilibrium with the default-mode network. In the Sheline study, the dorsal ACC showed more (of the expected) negative correlation to the precuneus in ApoE4 carriers. In this paper by Machulda et al., the dorsal ACC seed showed greater connectivity to several other salience network regions in the ApoE4 carriers. The Machulda paper does not report any between-network results.

    In my view, the most important finding in the current study is a null result rightly highlighted by the authors. Machulda and colleagues did not find any significant differences in gray matter density between E4 carriers and non-carriers, suggesting that resting state functional connectivity may be a more sensitive measure of preclinical disease than structural MRI. Or, stated more carefully, resting-state fMRI is more sensitive to preclinical changes than is a voxelwise measure of brain structure. The authors used voxel-based morphometry and failed to find any significant differences between E4 carriers and non-carriers. They did not report other structural measures such as entorhinal or hippocampal volume which may have detected differences.

    With the exception of the intriguing finding of increased medial temporal lobe and posterior cingulate connectivity in the default-mode network of very young E4 carriers (Filippini et al., 2009), all studies to date suggest that default-mode network connectivity declines in preclinical, MCI, and dementia stages of AD pathology. Resting-state functional connectivity appears to be a viable AD biomarker at the group level and may prove helpful in assessing treatment response in clinical trials. Its interpretability at the single subject level remains to be determined. The inclusion of resting-state fMRI in the second round of the Alzheimer’s disease neuroimaging initiative (ADNI2) will help the field judge the utility of this approach in assessing an individual patient.

    References:

    . APOE4 allele disrupts resting state fMRI connectivity in the absence of amyloid plaques or decreased CSF Aβ42. J Neurosci. 2010 Dec 15;30(50):17035-40. PubMed.

    . Distinct patterns of brain activity in young carriers of the APOE-epsilon4 allele. Proc Natl Acad Sci U S A. 2009 Apr 28;106(17):7209-14. PubMed.

    View all comments by Michael Greicius

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. A Foreshadowing? ApoE4 Disrupts Brain Connectivity in Absence of Aβ
  2. ApoE4 Linked to Default Network Differences in Young Adults
  3. Tracing Alzheimer Disease Back to Source
  4. Miami: Multimodal Imaging, New Way to Test Amyloid Hypothesis
  5. ApoE’s Secret Revealed? Protein Promotes Aβ Degradation
  6. Hearts and Minds: Can Both Benefit From Cholesterol Pathway Fix?

Paper Citations

  1. . APOE4 allele disrupts resting state fMRI connectivity in the absence of amyloid plaques or decreased CSF Aβ42. J Neurosci. 2010 Dec 15;30(50):17035-40. PubMed.
  2. . Identification of Alzheimer disease risk by functional magnetic resonance imaging. Arch Neurol. 2005 Dec;62(12):1881-8. PubMed.
  3. . Distinct patterns of brain activity in young carriers of the APOE-epsilon4 allele. Proc Natl Acad Sci U S A. 2009 Apr 28;106(17):7209-14. PubMed.
  4. . Molecular, structural, and functional characterization of Alzheimer's disease: evidence for a relationship between default activity, amyloid, and memory. J Neurosci. 2005 Aug 24;25(34):7709-17. PubMed.
  5. . APOE predicts amyloid-beta but not tau Alzheimer pathology in cognitively normal aging. Ann Neurol. 2010 Jan;67(1):122-31. PubMed.
  6. . ApoE promotes the proteolytic degradation of Abeta. Neuron. 2008 Jun 12;58(5):681-93. PubMed.
  7. . Attenuation of neuroinflammation and Alzheimer's disease pathology by liver x receptors. Proc Natl Acad Sci U S A. 2007 Jun 19;104(25):10601-6. PubMed.
  8. . Lipidation of apolipoprotein E influences its isoform-specific interaction with Alzheimer's amyloid beta peptides. Biochem J. 2000 Jun 1;348 Pt 2:359-65. PubMed.
  9. . Apolipoprotein E epsilon4 allele in familial and sporadic Parkinson's disease. Neurosci Lett. 2006 Oct 9;406(3):235-9. PubMed.
  10. . Association of APOE with Parkinson disease age-at-onset in women. Neurosci Lett. 2007 Jan 16;411(3):185-8. PubMed.

Other Citations

  1. APP23

External Citations

  1. apolipoprotein E
  2. PDGene

Further Reading

Papers

  1. . ApoE promotes the proteolytic degradation of Abeta. Neuron. 2008 Jun 12;58(5):681-93. PubMed.
  2. . Attenuation of neuroinflammation and Alzheimer's disease pathology by liver x receptors. Proc Natl Acad Sci U S A. 2007 Jun 19;104(25):10601-6. PubMed.
  3. . APOE4 allele disrupts resting state fMRI connectivity in the absence of amyloid plaques or decreased CSF Aβ42. J Neurosci. 2010 Dec 15;30(50):17035-40. PubMed.
  4. . Distinct patterns of brain activity in young carriers of the APOE-epsilon4 allele. Proc Natl Acad Sci U S A. 2009 Apr 28;106(17):7209-14. PubMed.
  5. . Identification of Alzheimer disease risk by functional magnetic resonance imaging. Arch Neurol. 2005 Dec;62(12):1881-8. PubMed.

Primary Papers

  1. . Effect of APOE ε4 status on intrinsic network connectivity in cognitively normal elderly subjects. Arch Neurol. 2011 Sep;68(9):1131-6. PubMed.
  2. . Critical role of astroglial apolipoprotein E and liver X receptor-α expression for microglial Aβ phagocytosis. J Neurosci. 2011 May 11;31(19):7049-59. PubMed.
  3. . The APOE gene locus in frontotemporal dementia and primary progressive aphasia. Arch Neurol. 2011 May;68(5):622-8. PubMed.