With statins doing wonders for Alzheimer’s mice and apolipoprotein E as the top risk gene, links between cholesterol and Alzheimer’s presumably run deep. Much prior research has explored how cholesterol influences amyloid-β, the peptide that clogs the brains of Alzheimer’s disease (AD) patients. A study in this week’s Journal of Neuroscience proposes a mechanism for how amyloid-β, in turn, can regulate cholesterol homeostasis. From experiments in cultured rat neurons, Elena Posse de Chaves and colleagues at the University of Alberta, Canada, propose that Aβ42 oligomers promote cell death by blocking cleavage of a transcription factor that drives cholesterol synthesis. The team traced the cholesterol problem to a reduction in protein prenylation, and showed that treating neurons with intermediates of the cholesterol pathway can relieve Aβ-induced toxicity. The findings suggest that tampering with cholesterol production to prevent or treat AD may be even more complicated than previously believed.

The idea for this study came from the observation that neurons exposed to Aβ develop cholesterol transport defects closely resembling those seen in Niemann-Pick type C disease. “That’s when we started thinking Aβ may have an important role in cholesterol trafficking,” de Chaves told Alzforum. In support of that idea, her lab found that cholesterol synthesis is required for neurons to internalize Aβ in the absence of ApoE (Saavedra et al., 2007), and others reported that Aβ accumulates before cholesterol levels go up in APP/PS1 transgenic mice (ARF related news story on Fernández et al., 2009).

To see if Aβ triggers cholesterol trafficking problems, first author Amany Mohamed and colleagues prepared Aβ42 oligomers using a published protocol (Dahlgren et al., 2002), and added them at 20 μM to cultures of rat basal forebrain neurons. Confocal microscopy of cells stained with filipin—a fluorescent tag for non-esterified cholesterol—suggested cholesterol levels may be higher in neurons cultured with Aβ than in those without. But when they measured cholesterol more quantitatively, the scientists found that Aβ-treated cells did not have more of it—they just could not get it to “go where it’s supposed to go,” de Chaves said. Aβ42 oligomers slowed retrograde transport in the neurons, causing cholesterol to get stuck in late endosomes and at the Golgi apparatus.

When the researchers looked at cholesterol synthesis, they found it was down in neurons exposed to Aβ. Probing the mechanism, they focused on sterol regulatory element-binding protein (SREBP-2), because this transcription factor activates genes of the mevalonate pathway of cholesterol synthesis. Furthermore, SREBP-2 activation requires proper trafficking of the protein from the endoplasmic reticulum to the Golgi, where it gets cleaved to release the DNA-binding fragment. Indeed, Aβ42 did seem to inhibit SREBP-2 cleavage, as determined by Western blots showing full-length and cleaved forms.

The pathway activated by SREBP-2 produces not only cholesterol, but also isoprenoids, which are post-translationally added to proteins in a process known as prenylation. By inhibiting the mevalonate pathway, the authors figured, Aβ42 might also curb prenylation—and sure enough, they found this to be true for Rab proteins, which regulate vesicular trafficking, in cultured rat neurons and in brains of TgCRND8 mice. Some question whether the Aβ-induced inhibition SREBP-2 is robust enough to account for the prenylation effects (see Ben Wolozin comment below). Nevertheless, in cultured neurons, the scientists were able to offset Aβ42’s ill effects on cell metabolism and survival by supplying geranylgeranyl pyrophosphate (GGPP), an intermediate in the mevalonate pathway.

By suggesting reduction of prenylation as a possible means by which Aβ triggers cell death, the new data seem to fit with a prior study showing that blocking isoprenoid production leads to buildup of intracellular Aβ (ARF related news story on Cole et al., 2005). At the same time, the current findings appear at odds with other research linking decreased isoprenylation with beneficial outcomes—namely, reduced Aβ production (ARF related news story on Ostrowski et al., 2007) and enhanced α-processing of amyloid precursor protein (ARF related news story on Pedrini et al., 2005). The studies used different methods and looked at different subcellular compartments, suggesting that “more work is required to put the whole prenylation story together,” Sam Gandy of Mount Sinai School of Medicine, New York, wrote in an e-mail to Alzforum (see full comment below). Furthermore, proteins have different sensitivities to reduced prenylation, said De Chaves, noting that her team only saw effects in neurons that accumulate Aβ. “In the brain, not all cells may react the same way to Aβ,” she said.

The present finding sheds light on, yet also complicates, the situation with statins, which, despite causing remarkable improvements in AD mice, have yet to translate into benefits for people (see ARF related news story). “Our work indicates that statins and Aβ will have a synergistic effect because they both inhibit cholesterol synthesis,” de Chaves said. “If statins are provided early in the process of AD, they could decrease production of Aβ. But if they are given after Aβ has built up, you will have two agents causing the same effect—decreased cholesterol synthesis—with the additional impact of Aβ not only on cholesterol but on the whole [mevalonate] pathway. The implications of that could be detrimental.”—Esther Landhuis


  1. This manuscript presents some novel effects of oligomeric Aβ on neuronal biology. Mohamed et al. show that Aβ inhibits isoprenylation, and suggest that it does so by impairing cholesterol metabolism and SREBP-2 cleavage. The observations presented in the manuscript are quite convincing; however, I disagree with the interpretations.

    The authors observe that cholesterol accumulates in the presence of Aβ despite treatment with pravastatin. They use this observation to pursue putative effects of Aβ on SREBP-2. Unfortunately, I'd caution that this assumption defies logic. If cholesterol accumulates despite pravastatin treatment, then the mechanism cannot be SREBP-mediated cholesterol synthesis, because pravastatin inhibits the rate-limiting step, which is HMG-CoA reductase, and this is what is controlled by SREBP. The authors then show that Aβ inhibits SREBP-2 by about 60 percent, and conclude that this accounts for the effects on isoprenylation. Again, I suspect this is not accurate because HMG-CoA reductase must be inhibited by more than 90 percent before one sees significant deficits in isoprenylation (this is because the affinity of HMG-CoA for its substrates is so strong).

    Thus, while the observations are convincing and likely true, the interpretation is likely to be different than suggested in the manuscript. Unfortunately, I don't know the biology of SREBP-2 well enough to know whether it directly inhibits production of the isoprenyl synthetic enzymes/conjugases, such as geranylgeranyl synthetase. Such inhibition would better explain the observed phenomena. Hopefully, future studies will clarify the mechanism.

    View all comments by Benjamin Wolozin
  2. The cholesterol/ApoE/Aβ connection is one of the most challenging areas in AD research. Studies of statins led to very dramatic benefits in mice that, so far, have not been translatable to humans. Elucidation of statins' effects led many labs, including our own, to investigate the isoprenoid pathway, where we implicated Rho kinase as a tonic physiological antagonist of the α-secretase pathway.

    This new paper looks at a different step in the pathway. This group demonstrates that oligomeric Aβ inhibits protein prenylation (e.g., farnesylation, geranylgeranylation) in cultured neurons. This isn't immediately reconcilable with our earlier studies, where farnesylation or geranylgeranylation inhibitors seemed to be a good thing. To be fair, however, the new paper asks a question that is not really predicated on our results, and it is conceivable that both are right. In addition to using different protocols, our groups may be targeting different subcellular compartments. More work is required to put the whole prenylation story together.

    Still, in this area, the more important questions relate to differential ApoE isoform action and the field's vexing inability to translate robust Aβ-lowering effects of statins into a meaningful therapeutic or prophylaxis. One is tempted to wonder whether a statin prevention trial might be worthwhile, but many colleagues in the area are convinced that there are enough statin data to exclude much possibility of success.

    View all comments by Sam Gandy
  3. We would like to respond to Dr. Wolozin on his disagreement with the interpretations of our results. His views focus mainly on cholesterol synthesis, when, in fact, our work suggests that changes in cholesterol synthesis are not responsible for the “cholesterol sequestration” phenotype observed in neurons challenged with Aβ during the experimental window. Although the finding that Aβ inhibited cholesterol synthesis seemed paradoxical to the intensive filipin staining, it is not unprecedented since the drug U18666A is a potent inhibitor of cholesterol synthesis and induces a similar pattern of cholesterol sequestration. Our rationale for examining SREBP-2 as the target for Aβ came from the observations that, although both Aβ and pravastatin significantly reduced cholesterol synthesis, pravastatin (at the concentration used in our study) did not cause cholesterol sequestration, nor did it cause apoptosis.

    Moreover, in agreement with Dr. Wolozin’s concepts on HMGCoA and prenylation, we did not observe any significant change in protein prenylation in neurons treated with pravastatin. This suggests that the conditions of pravastatin treatment were insufficient to achieve enough HMGCoA inhibition, even though we did not measure HMGCoA reductase activity or levels. At no point in our work have we claimed that HMGCoA reductase was involved in the effects of Aβ, and we would not be confident to directly extrapolate levels of HMGCoA reductase from SREBP-2 levels; thus, we do not understand the remark about accuracy that has been made. Finally, as Dr. Wolozin suspected, SREBP-2 regulates many other enzymes of the mevalonate pathway including farnesyl diphosphate synthase, which catalyzes formation of farnesyl-PP; therefore, inhibition of SREBP-2 can be expected to have a higher impact on prenylation. We believe that our work presents strong evidence that SREBP-2 is a target of Aβ. We have identified one pathway affected by the decrease of nuclear SREBP-2, but we expect that other targets of SREBP-2 could also play important roles.


    . Combined analysis of oligonucleotide microarray data from transgenic and knockout mice identifies direct SREBP target genes. Proc Natl Acad Sci U S A. 2003 Oct 14;100(21):12027-32. PubMed.

    . SREBPs: activators of the complete program of cholesterol and fatty acid synthesis in the liver. J Clin Invest. 2002 May;109(9):1125-31. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. High Cholesterol Leaves Mitochondria Defenseless Against Aβ
  2. Statins and AD—What Role Isoprenoids?
  3. What Role Brain Statins—Sparing Isoprenoids from the Rac?
  4. Statins Boost α-Secretase, but Not Through Cholesterol
  5. Philadelphia: All Is Not Well with the Statin Story

Paper Citations

  1. . Internalization of beta-amyloid peptide by primary neurons in the absence of apolipoprotein E. J Biol Chem. 2007 Dec 7;282(49):35722-32. PubMed.
  2. . Mitochondrial cholesterol loading exacerbates amyloid beta peptide-induced inflammation and neurotoxicity. J Neurosci. 2009 May 20;29(20):6394-405. PubMed.
  3. . Oligomeric and fibrillar species of amyloid-beta peptides differentially affect neuronal viability. J Biol Chem. 2002 Aug 30;277(35):32046-53. PubMed.
  4. . Statins cause intracellular accumulation of amyloid precursor protein, beta-secretase-cleaved fragments, and amyloid beta-peptide via an isoprenoid-dependent mechanism. J Biol Chem. 2005 May 13;280(19):18755-70. PubMed.
  5. . Statins reduce amyloid-beta production through inhibition of protein isoprenylation. J Biol Chem. 2007 Sep 14;282(37):26832-44. PubMed.
  6. . Modulation of statin-activated shedding of Alzheimer APP ectodomain by ROCK. PLoS Med. 2005 Jan;2(1):e18. PubMed.

Other Citations

  1. TgCRND8 mice

Further Reading


  1. . Statins reduce amyloid-beta production through inhibition of protein isoprenylation. J Biol Chem. 2007 Sep 14;282(37):26832-44. PubMed.
  2. . Modulation of statin-activated shedding of Alzheimer APP ectodomain by ROCK. PLoS Med. 2005 Jan;2(1):e18. PubMed.
  3. . Statins cause intracellular accumulation of amyloid precursor protein, beta-secretase-cleaved fragments, and amyloid beta-peptide via an isoprenoid-dependent mechanism. J Biol Chem. 2005 May 13;280(19):18755-70. PubMed.

Primary Papers

  1. . β-amyloid inhibits protein prenylation and induces cholesterol sequestration by impairing SREBP-2 cleavage. J Neurosci. 2012 May 9;32(19):6490-500. PubMed.