This is Part 3 of a three-part series. See also Part 1 and Part 2.
5 July 2010. The role of ApoE receptors was high on the agenda at “ApoE, ApoE Receptors and Neurodegeneration,” a one-day symposium held 7 June 2010 at Washington University, St. Louis, Missouri. These cell-surface proteins enable a flurry of ApoE activity, much of it totally independent of amyloid-β pathology, from enhancing synaptic modeling and plasticity to activating microglia (see Part 1 and Part 2 of this series). But whenever AD is discussed, it is not long before Aβ comes up, and the same was true at this meeting as well, which featured a useful summary of where the science of amyloid and ApoE currently stand.
It is well established that apolipoprotein E binds to Aβ in vitro, colocalizes with it in vivo, and profoundly affects Aβ structure, levels, and toxicity in mouse models. Cognitively normal people are more likely to amass amyloid-β in the brain as they age if they carry ApoE4, whereas ApoE2 carriers show not such propensity (see Reiman et al., 2009; Morris et al., 2010). And yet, how ApoE affects Aβ accumulation in real life and over time is far from trivial a question to answer. Over the last few years, David Holtzman and colleagues at Washington University, St. Louis, have tackled this issue by developing a microdialysis procedure to study the comings and goings of Aβ in the brains of AD model mice. With this system they found that ApoE extends the half-life of Aβ in the extracellular space in PDAPP mouse models. Joe Castellano, a graduate student in the Holtzman lab, found that in 18-month-old mice, the half-life of Aβ is around an hour in an ApoE4 background, about half that in mice expressing ApoE2, and 40 minutes in mice expressing ApoE3. This means that ApoE stays around about twice as long in mice expressing the AD risk allele. Amyloid pathology is worse in the E4 mice, suggesting that the longer half-life of Aβ in these animals could result from the peptide simply sticking to plaques. However, the half-life of Aβ is also about 60 minutes in young ApoE4 mice that have no plaques yet. “This is pretty clear evidence that differential Aβ deposition in these animals occurs because of soluble Aβ clearance problems,” said Holtzman. He thought this was an important finding to keep in mind, but added that much work needs to be done to figure out the cellular mechanisms. One avenue of research currently underway in Holtzman’s lab focuses on the nexus among ApoE, its receptors, and Aβ clearance (see ARF related news story).
Daniel Michaelson, Tel Aviv University, Israel, also presented evidence that ApoE4 extends Aβ’s half-life. In a “hot topic” poster abstract chosen for oral presentation, he described how the neprilysin inhibitor thiorphan can help reveal ApoE-dependent Aβ dynamics and effects on neurodegeneration. In ApoE targeted replacement (TR) mice, which carry human ApoE isoforms in lieu of their own, both total and oligomeric Aβ accumulated in hippocampal neurons when neprilysin was inhibited in an ApoE4 background, but not in an ApoE3 background. This fits with Holtzman’s finding that Aβ stays around longer in the presence of ApoE4. Michaelson, collaborating with Eliezer Masliah at the University of California, San Diego, turned to the electron microscope to view the consequences of this long-lived Aβ for cells. With immuno-EM, Michaelson saw that Aβ and oligomers of Aβ were found in mitochondria, but mostly colocalized with ApoE in enlarged lysosomes. Michaelson suggested that Aβ becomes toxic to cells by targeting both the mitochondria and the lysosomal system, resulting in increased apoptosis and cell death. He believes that the LDL receptor-related protein 1 (LRP1) may mediate the effects of ApoE because LRP1 levels rise in ApoE4 TR mice treated with neprilysin. (For more on LRP1, see Part 2 of this series.)
For his part, Holtzman agreed that apolipoprotein receptors might be involved in ApoE/Aβ dynamics. Knocking out one of them, the LDL receptor (LDLR), boosts ApoE in the cerebrospinal and interstitial fluids in the brains of PDAPP mice (see Fryer et al., 2005), while overexpressing LDLR in the same transgenic mice ablates the apolipoprotein, probably due to clearance mediated by the LDLR, noted Holtzman (see also ARF related news story on Kim et al., 2009). Overexpression of the LDLR also curtails plaque pathology and limits soluble Aβ as measured by microdialysis. His lab has shown that boosting the LDL receptor can reduce both levels of ApoE in the brain and plaque pathology, Holtzman said. In the poster session, Jacob Basak from the lab showed that astrocytes from LDLR overexpressing mice clear Aβ added to the culture medium much faster than do wild-type astrocytes. That data suggest that enhanced clearance of Aβ by glial cells may explain why LDLR overexpressing mice have less amyloid pathology, though it is still not clear if ApoE is part of the mechanism.
One way ApoE might extend the half-life of Aβ is by affecting its aggregation. In his hot-topic presentation, Tadafumi Hashimoto, who works in Brad Hyman’s lab at Massachusetts General Hospital, Charlestown, described a “split luciferase” assay system to measure Aβ interactions. This system expresses the N-terminal half of the luciferase on one Aβ construct, and the C-terminal half on another. When Aβ molecules aggregate, luciferase is reconstituted and emits light. Hashimoto found that ApoE4 enhances formation of high-molecular-weight Aβ aggregates in HEK293 cells, whereas ApoE2 facilitates formation of lower-molecular-weight species. Furthermore, the lipidation status of ApoE is crucial for its effect on Aβ aggregation. While recombinant ApoE had no effect on in vitro Aβ aggregation, affinity-purified, lipidated ApoE enhanced the formation of Aβ oligomers. The work also suggests that ApoE might extend Aβ’s half-life by promoting aggregation.
One other potential nexus between ApoE and Aβ clearance is the blood-brain barrier (BBB). The barrier may be compromised in AD patients and is known to leak in ApoE knockout mice, but just how ApoE isoforms affect the BBB is unknown, said Kazuchika Nishitsuji, National Center for Geriatrics and Gerontology in Aichi, Japan. In his hot-topic talk, Nishitsuji described how he followed the distribution of the large-molecular-weight dye Evans Blue in the brain after injecting it into the peritoneal cavity as a way of assessing the BBB. He reported that leakage was greater in ApoE4 knock-in animals than in ApoE3 animals. In a cell model of the BBB that co-cultures astrocytes, endothelial cells, and pericytes, electrical resistance across endothelial cells was lower when astrocytes expressed ApoE4 rather than ApoE3. The lower resistance suggests that the endothelial cells were more permeable. How ApoE4 makes these glia more permeable is not clear, but Nishitsuji noted that phosphorylation of occludin was up. Occludin is one of the proteins found in endothelial cell tight junctions, which help hold the BBB together. It also remains to be seen how ApoE4 isoforms might affect endothelial cells and the BBB in an AD mouse model.
All told, this meeting demonstrated that the biology that surrounds ApoE, ApoE receptors, and AD pathology is complex. Many people might like to know how this knowledge could be put to practical use, as in a therapeutic that could prevent or slow Aβ pathology in ApoE4-positive patients. There was little to get excited about on that front. It was mentioned that statins increase LDLR in the periphery and have been investigated as potential AD therapeutics with little success, perhaps because they do not readily penetrate the brain. Holtzman noted one new therapeutic aimed at boosting LDLR levels—an antibody to proprotein convertase subtilisin/kexin type 9, or PCSK9, which degrades LDLR. Amgen is currently putting that human monoclonal antibody through its paces in a Phase 1 clinical trial in hopes to use the antibody to treat high plasma cholesterol. But it is too soon to tell whether the antibody, or better yet, a small molecule inhibitor of the protease, could be beneficial in AD.—Tom Fagan.
This is Part 3 of a three-part series. See also Part 1 and Part 2.