"This paper highlights the selective vulnerability of the ApoE4 genotype to disease-related changes in Aβ homeostasis," said Gary Landreth, Case Western Reserve University, Cleveland, Ohio. Landreth was not involved in the study.

ABCA1 is a cell-membrane lipid pump that transports cholesterol and other phospholipids out of the cell and onto apolipoproteins such as ApoA-I and ApoE. Some genomewide association studies have tied single-nucleotide polymorphisms in Abca1 to AD (see Wollmer et al., 2003), though others have not seen the association (see ARF related news story on Li et al., 2004).

In mice, Koldamova's group and others have shown that deleting the Abca1 gene exacerbates Aβ deposition in transgenic lines (see ARF related news story), while Abca1 hemizygosity worsens memory problems (see Lefterov et al., 2009). What's more, Abca1 overexpression prevents the AD phenotype in APP transgenic mice (see ARF related news story). These results apply only to mice that express endogenous ApoE; therefore, Koldamova and colleagues wanted to test what happens in mice that express either human ApoE3 or ApoE4. "Our prediction was that deficiency of one Abca1 copy would be equally damaging to both ApoE genotypes," said Koldamova. "This appears not to be the case."

First authors Nicholas Fitz and Andrea Cronican crossed APP-PS1 transgenic mice with human ApoE3 (ARF related news story) or ApoE4 (ARF related news story)-targeted replacement mice, and then bred progeny with Abca1+/- animals. To characterize the APP-PS1/ApoE/Abca1+/- offspring, the researchers tested cognition and measured plaque levels in the brain, as well as various factors in the brain’s interstitial fluid (ISF) and plasma.

To their surprise, the authors found that halving the normal amount of ABCA1 only affected mice with ApoE4, but not those with ApoE3. APP-PS1/E4 mice lacking a copy of Abca1 performed worse than homozygous controls in the radial water maze test. Immunohistochemical staining revealed more numerous compact and diffuse plaques in brains of E4 mice compared to E3 controls. The pattern continued in the interstitial fluid, where shortage of ABCA1 both elevated ISF Aβ42 and Aβ40 and slowed its clearance in ApoE4 mice only. ApoE and ApoA-I levels also fell in soluble brain fractions and in plasma of E4 mice. Finally, levels of high-density lipoproteins and Aβ42 and Aβ40 fell in the plasma of ApoE4 mice, which correlated with more brain plaques.

Why would an ABCA1 deficiency affect ApoE4 carriers more profoundly than ApoE3 mice? It may have to do with how much ApoE there is in the animals, said Koldamova. "ApoE4 mice make less of the protein, so a deficiency of ABCA1 hits them harder than the E3 genotype," she said. If ABCA1 is not around to lipidate apolipoproteins, they become unstable and are metabolized. Because ApoE3 carriers have more of the protein, they may be more resilient to such losses. It is possible that the same differential outcomes would hold true in humans, she added, but that has not been examined yet.

"As we develop therapeutics that target ABCA1 and ApoE, we need to understand whether there are differential effects of ApoE4 versus ApoE3," said David Holtzman, Washington University in St. Louis, Missouri. Holtzman found the ApoE4-specific effects unexpected. Researchers will want to know if a person’s ApoE genotype causes a particular response to liver X receptor (LXR) or retinoic X receptor (RXR) agonists, which elevate ABCA1 and ApoE levels and are currently under investigation as AD treatments, he added (see ARF related news story).

The study hinted that there may be more to exacerbated amyloid pathology than loss of ApoE. In particular, the authors also found that lower plasma high-density lipoproteins (HDLs) correlated with higher plaque levels in the brain in the ApoE4 mice. However, both Landreth and Holtzman agreed that those results were correlational, and that more evidence would be needed to show causation.

Landreth noted that Koldamova and colleagues measured ApoA-I, a component of HDLs, in the brain as well, although the protein is thought to be sequestered in the periphery. Brain ApoA-I fell in E4 Abca1 hemizygotes. "That's a potentially important finding," said Landreth. "If brain levels of ApoA-I can be modulated, then you've got another player in Aβ homeostasis that we didn't think was active."

Koldamova's group will next test the effects of LXR and RXR agonists in Abca1 hemizygotes to see how mice with different ApoE genotypes respond. She also plans to examine the response to other brain insults, such as neurotrauma. Since ApoE carries cholesterol, which is important for synaptic plasticity, axonal regeneration, and cell death, a drop in ABCA1 could compromise injury responses for people with ApoE4.—Gwyneth Dickey Zakaib.

Reference:
Fitz NF, Cronican AA, Saleem M, Fauq AH, Chapman R, Lefterov I, Koldamova R. Abca1 Deficiency Affects Alzheimer’s Disease-Like Phenotype in Human ApoE4 But Not in ApoE3-Targeted Replacement Mice J Neurosci 2012 Sept 19; 32(38):13125–13136. Abstract