21 December 2011. Apolipoprotein E4 (ApoE4), the Goliath of late-onset Alzheimer's disease genes (see AlzGene top results), is an enigma. Its protein clearly plays a role in AD and could be a therapeutic target, but scientists are not sure whether they should raise or lower it in the brain. While some studies suggest that more ApoE, regardless of isoform, would be beneficial for plaque clearance, a new paper by David Holtzman and colleagues at the Washington University School of Medicine in St. Louis, Missouri, provides strong evidence that less ApoE is better. They report that, in mice, halving the genetic dose of either human ApoE3 or ApoE4 results in fewer amyloid plaques and less microglial activation than have transgenic mice expressing two copies of the genes. The results, published December 7 in the Journal of Neuroscience, imply that ApoE-lowering treatments have a place among proposed AD therapies.
"Decreasing ApoE resulted in a massive decrease in amyloid-β pathology and inflammation," said Holtzman. "It's the first study to show how the dosage of human ApoE affects amyloid-related pathology," he added. Previous studies found a similar dose-dependent relationship in mice expressing their own ApoE gene (see Bales et al., 1997 and Holtzman et al., 2000).
First author Jungsu Kim and colleagues crossed APPswe/PS1(L166P) mice with knock-in mouse models that replace the endogenous ApoE gene with human sequences for ApoE3 or ApoE4. The researchers designed the crosses to get APP transgenic mice with either one or two copies of the human genes. Compared to mice with two copies, haploinsufficient mice, with only one human ApoE gene, had about half the amount of ApoE mRNA in the cortex. They also had less than half the insoluble Aβ40 and Aβ42, a dramatic reduction in plaques and fibrillar Aβ, and 95 percent less microglial activation. This dose-dependent effect was true for both ApoE4, the strongest risk allele for AD, and ApoE3, considered the allele with normal risk. Holtzman said they found a similar result for the ApoE2 allele, which is thought to be protective against Alzheimer's, but these data are not published yet. While mice that carried the ApoE4 allele had far greater Aβ accumulation than did those carrying the ApoE3 allele, decreasing the dose of either allele cut down on pathology.
The results of this study go against the theory that raising ApoE levels would aid in plaque clearance. Researchers previously observed that transgenic mice carrying the human ApoE4 allele have less ApoE and more Aβ deposition than mice expressing ApoE3 (see Bales et al., 2009). Humans homozygous for the ApoE4 also have less plasma ApoE than those with other alleles (see Gupta et al., 2011). Those observations led to a widespread hypothesis that dropping ApoE levels, of any allele, leads to more pathology and that boosting ApoE, regardless of isoform, might be a reasonable therapeutic strategy.
The current paper directly tests that hypothesis and finds that the opposite is true, said Yadong Huang, University of California, San Francisco. "The therapeutic implication is quite significant—it changes the direction of the ApoE as a target," he told ARF. That was Holtzman’s contention as well. "If you want to develop a treatment to affect ApoE and decrease pathology, this study suggests you'd want to decrease ApoE levels, assuming it didn't cause some other problem in the brain," he told Alzforum. His group did not notice anything unusual in the mice with just one copy, he said, but that would need further exploration. Huang noted that the lowered protein level in these mice is a lifelong change. Further tests are necessary to verify that an ApoE-dampening treatment administered in adulthood would have the same effect, he added.
Holtzman said that the current results fit with the possibility that ApoE impairs the clearance of Aβ (see ARF Webinar and ARF related news story on Castellano et al., 2011). "ApoE isoforms appear to regulate the clearance of Aβ from the brain, and they do it differentially," Holtzman told ARF. While all isoforms seem to slow clearance, he said, ApoE4 seems to be the slowest, which may be why it confers the most genetic risk. "ApoE is probably also directly affecting Aβ aggregation itself," he added, though it is not yet clear how.
For some scientists, the results were unexpected. "It's a bit surprising that decreasing ApoE provided this much of a dramatic effect," said Steven Paul, Weill Cornell Medical College, New York City. Paul was not involved in the study. "There's a lot of evidence that increasing lipidated ApoE in models is beneficial," he told Alzforum. Holtzman agrees. He said raising the amount of lipidated ApoEs is another story entirely. Compounds that hike the amount of cholesterol and phospholipids present in each ApoE-containing lipoprotein particle (raising ApoE's "lipidation state") reportedly lessen plaque load in transgenic mice (see ARF related news story on Jiang et al., 2008). These compounds both increase expression of ApoE, and simultaneously stimulate production of molecules such as ABCA1 (ATP-binding cassette A1, which loads ApoE with lipids) that alter the protein's lipidation state. "When ApoE is more lipidated in the brain, it appears to either increase Aβ clearance or decrease its aggregation," Holtzman told Alzforum. "If one increases ApoE levels but also increases its lipidation state, that is very different from just increasing ApoE."
For therapeutic purposes, altering the lipidation state of ApoE with drugs, such as nuclear liver X receptor agonists, could be just as important, and a potentially more viable therapeutic strategy than altering the level of the protein alone, said Gary Landreth, Case Western Reserve University in Cleveland, Ohio. "The biologically relevant species are the ApoE-based high-density lipoproteins (HDLs)," he said. Simultaneously increasing both ApoE expression and its lipidation could clear plaques just as well as lowering ApoE, he contended.
Holtzman's next steps will be to raise the level of ApoE in the brain (without altering the lipidation state) to see if, as the current experiment predicts, Aβ levels go up. He will also look to repeat his results in other animal models and work to figure out the mechanisms behind how ApoE lowering in turn alters Aβ deposition. Given the genetic importance of ApoE in AD, "This puzzle is unequivocally the most important riddle to solve in the field of Alzheimer's disease," said Paul.
The ApoE puzzle has implications for other disorders as well. Its E2 allele, while protective against AD, was reported to heighten the risk of sporadic Parkinson's disease, frontotemporal dementia, and to cause earlier onset of Huntington's disease. In addition, a recent paper by Conceicao Bettencourt, University of the Azores in Ponta Delgada, Portugal, and colleagues, published in the December 12 Archives of Neurology, reports that the ApoE2 allele lowers the onset age for Machado-Joseph Disease (MJD), the most common form of spinocerebellar ataxia. People inheriting a nucleotide expansion in ataxin 3, which causes the disease, and who had an ApoE2 allele got the disease, on average, five years earlier than those carrying ApoE3 or ApoE4 alleles, and had a five times higher risk of developing the disease before the age of 39. The authors pointed out that ApoE2 has a 50- to 100-fold weaker affinity for low-density lipoprotein receptors in astrocytes than ApoE3 and ApoE4. This "could contribute to the altered homeostasis of cholesterol in the brain, which may ultimately be associated with the earlier manifestation of MJD in ε2 allele carriers," they wrote.—Gwyneth Dickey Zakaib.
Kim J, Jiang H, Park S, Eltorai AEM, Stewart FR, Yoon H, Basak JM, Finn MB, Holtzman D. Haploinsufficiency of Human APOE Reduces Amyloid Deposition in a Mouse Model of Amyloid-β Amyloidosis. The Journal of Neuroscience, 2011 December 10; 31(49):18007–12. Abstract
Bettencourt C, Raposo M, Kazachkova N, Cymbron T, Santos C, Kay T, Vasconcelos J, Maciel P, Donis KC, Saraiva-Pereira ML, Jardim LB, Sequeiros J, Lima M. The APOE 2 Allele Increases the Risk of Earlier Age at Onset in Machado-Joseph Disease. Arch Neurol. 2011 December;68(12):1580-3. Abstract