7 November 2008. If a gene allele clearly raises risk for a devastating disease, it might seem a no-brainer to assume that the variant protein contributes to pathogenesis, and that correcting this rogue factor could help restore normality. But in the case of ApoE4 and Alzheimer disease, accumulating evidence—most recently from Peter Reinhart and colleagues at Wyeth Research in Princeton, New Jersey—suggests a more complex scenario. In this week’s Journal of Neuroscience, the Wyeth group reports that mice expressing the human ApoE4 allele have less total ApoE than those expressing the other two human variants. Furthermore, astrocytes seem to degrade ApoE4 more quickly than they do ApoE3, an isoform known to promote neurite outgrowth and synapse formation. “My feeling is that ApoE4 itself is not doing anything toxic,” lead author David Riddell told ARF. “Potentially the reason ApoE4 is associated with AD is that it's not as effective at repairing neuronal connections as ApoE3.” The findings raise intrigue as to whether ApoE-targeted AD treatments might fare better by replacing lost ApoE protein rather than target its biological activity.
Riddell and colleagues measured ApoE protein levels in 20-week-old mice expressing human ApoE2, E3, or E4 in place of the endogenous murine ApoE protein (Knouff et al., 1999). In these targeted replacement (TR) mice, amounts of ApoE protein in the frontal cortex and hippocampus varied in a genotype-dependent fashion. ApoE levels were highest in E2/2 animals, 12-24 percent lower in E3/3 animals, and 26-39 percent lower in the E4/4 group, consistent with the relative AD risk conferred by these ApoE alleles (E2 being linked to lowest and E4 to highest AD risk). The researchers found similar trends when assaying for ApoE protein in mouse plasma and cerebrospinal fluid (CSF) samples, and in measurements of secreted ApoE in culture medium from human astrocytoma cell lines and primary astrocytes cultured from the ApoE TR mice.
Using an ApoE4-specific immunoassay, they determined that ApoE4 represents 30-40 percent of the total ApoE pool in brain, CSF, and plasma of E3/4 heterozygote mice, which have lower levels of total ApoE than do E3/3 mice. They also showed that the per-allele amount of ApoE3 was similar in E3/4 and E3/3 mice, as was the amount of ApoE4 per allele in E3/4 and E4/4 animals. The authors took these data as evidence that reduced E4 levels account for the lower amounts of total ApoE protein in E3/4 compared with E3/3 mice. Levels of ApoE mRNA, however, did not significantly differ in cortical and hippocampal samples across E2/2, E3/3, and E4/4 mice, implicating post-translational mechanisms in the differing ApoE levels observed between the various genotypes.
To get at a possible mechanism, the researchers investigated whether differential degradation and secretion of the ApoE isoforms could account for the alterations in ApoE protein levels. Sure enough, in pulse-chase experiments with E3/3 and E4/4 primary astrocytes, ApoE degradation was higher in E4/4 cultures, with about 43 percent of E4 degraded after a 120-minute chase, compared with 22 percent for E3. The half-life of cell-associated E3 was nearly double that of E4 (96 minutes vs. 49 minutes).
“This paper and these mice are building the hypothesis that the problem with ApoE4 is not something that E4 is doing,” said Bill Rebeck of Georgetown University Medical Center in Washington, DC. “It's that E4 is being degraded, and so there's just less ApoE around.” In contrast to ApoE-based therapy built on the premise that “inheritance of ApoE4 is something bad, so I need to get rid of the ApoE4,” the new data suggest that treatments should instead focus on replenishing the lost ApoE, he said. “This really says it might be a levels question.”
But amounts may only be part of the picture. Last year, Carol Colton and Mike Vitek of Duke University Medical Center, Durham, North Carolina, also showed that ApoE protein levels in E4/4 TR mice appear to be significantly lower than in their E3/3 counterparts (Vitek et al., 2007). That study compared macrophage-mediated innate immune responses in E3/3 and E4/4 TR mice, as well as E3/0 heterozygotes that express levels of total ApoE protein equivalent to those found in E4/4 mice. The researchers found that pro-inflammatory cytokine secretion in the E3/0 heterozygotes was greater than that of E3/3 but much lower than that of E4/4 mice, suggesting that pound for pound, ApoE4 allows a stronger inflammatory response than does ApoE3. “This genetic experiment clearly shows that the effect of the APOE4 gene allele is complex and depends on both the absolute levels of ApoE protein and on an independent effect of the ApoE4 protein,” they wrote in an e-mail to ARF (see full comment below).—Esther Landhuis.
Riddell DR, Zhou H, Atchison K, Warwick HK, Atkinson PJ, Jefferson J, Xu L, Aschmies S, Kirksey Y, Hu Y, Wagner E, Parratt A, Xu J, Li Z, Zaleska MM, Jacobsen SJ, Pangalos MN, Reinhart PH. Impact of Apolipoprotein E (ApoE) Polymorphism on Brain ApoE Levels. J. Neurosci. 5 Nov 2008; 28: 11445-11453. Abstract