23 May 2009. Cholesterol plays the villain in Alzheimer disease as an enabler of the production of toxic amyloid-β (Aβ) peptides, but it may have a second role as well, according to a paper out in the May 20 Journal of Neuroscience. Anna Colell and Jose Fernandez-Checa at the Institute de Investigaciones Biomediques de Barcelona in Spain use two mouse models of cholesterol accumulation to show that excess cholesterol in brain mitochondria depletes the antioxidant glutathione (GSH), which renders the cells vulnerable to Aβ toxicity. The results provide a novel link among cholesterol, oxidative stress, and intracellular Aβ, and suggest that replacing mitochondrial GSH may bolster cells’ defenses against Aβ toxicity.
High cholesterol in mid-life is a risk factor for AD, and studies have shown that cholesterol can affect the production of Aβ from its precursor by modulating the activity of both the β- and γ-secretase enzymes. Colell and Fernandez-Checa were interested in another aspect of cholesterol, namely its effects on mitochondria. They had previously described that mitochondrial GSH, by modulating generation of reactive oxygen species (ROS), is a key factor in the cellular susceptibility to apoptotic stimuli including TNF and Fas. Mitochondria do not make GSH, but instead rely on transport of the compound from the cytosol. Colell and Fernandez-Checa had found that this transport is highly sensitive to changes in membrane fluidity, and that increases of mitochondrial membrane cholesterol result in depleted GSH and enhanced oxidative stress in liver cells (Mari et al., 2006). Emerging evidence of the intracellular presence of Aβ and data pointing to mitochondria as a source of oxidative stress in response to Aβ led them to ask whether mitochondrial cholesterol might also regulate Aβ neurotoxicity.
To probe that question, lead author Anna Fernandez, along with Laura Llacuna, made use of two mouse models of altered cholesterol metabolism. The first mice overexpress a truncated, active form of the transcription factor sterol regulatory element binding protein-2 (SREBP-2), and make excess cholesterol. The second had a knockout of the NPC1 gene, a model of Niemann-Pick type C disease where a loss of normal cholesterol transport results in lipid accumulation and progressive neurodegeneration. In both cases, brain mitochondria isolated from the mice had twice or higher the cholesterol and half as much GSH compared to normal mice, the investigators found. When the researchers treated isolated mitochondria with Aβ, the accumulation of reactive oxygen species was twice that of controls, and mitochondrial permeabilization increased. The sensitivity translated to enhanced toxicity in whole cells, where even low concentrations of Aβ42 (5 nM and less) caused more cell death in neurons from the SREBP-2 mice than from normal mice.
The results suggest that mitochondrial GSH helps fend off oxidative stress in response to Aβ. To show this more directly, the investigators used chemical treatment of mitochondria to deplete GSH, and looked at the response to Aβ42. Whether the depletion was done in vitro by exposing isolated mitochondria to ethacrynic acid or in vivo, by infusing rats with buthionine sulfoxamine, an inhibitor of GSH synthesis, the treated mitochondria had lower GSH, produced more ROS, and were more permeable in response to Aβ. Likewise, selective mitochondrial GSH depletion by an inhibitor of import sensitized a neuronal cell line (SH-SY5Y) and three glial-derived cell lines to Aβ toxicity, suggesting that the effect was specific to mitochondrial GSH, but not to neurons.
To look at the response to Aβ in vivo, the researchers treated the SREBP-2 or control mice with an intracerebroventricular infusion of human Aβ42. By all measures, the transgenic mice showed a more dramatic inflammatory response to Aβ42 than the controls, displaying elevated microglia and astrocyte activation, and expression of proinflammatory cytokines. In addition, the transgenic animals showed more oxidative damage (lipid peroxidation and accumulation of oxidized proteins), synaptic loss, and neuronal death with markers of apoptosis. All these endpoints were reversed by co-infusion of GSH ethyl ester, a cell-permeable form of GSH, which significantly increased the pool of mitochondrial GSH in the mice. Intraperitoneal infusion of the GSH ethyl ester was also effective.
To see if mitochondrial GSH depletion might occur in AD, the researchers looked at APP/PS1 transgenic mice, which accumulate Aβ from early life on. There, they found that cholesterol levels in brain mitochondria were elevated and GSH decreased in 10-month-old mice, but not in younger mice. SREBP-2 mRNA and protein levels were increased in mice from four months of age on, leading the authors to speculate that Aβ might itself regulate cholesterol levels in some way.
Does the same thing happen in humans? In regard to mitochondrial cholesterol accumulation, Colell and Fernandez-Checa wrote in an e-mail to ARF, “Although, to our knowledge, the specific pool of mitochondrial cholesterol has not been analyzed, there are different reports describing increased cholesterol in brain tissue samples of AD patients (Cutler et al., 2004; Marí et al., 2006; Bandaru et al., 2007). These studies show that cholesterol accumulates specifically in vulnerable brain regions and correlates with the disease severity. Moreover, increased expression of steroidogenic acute regulatory protein (StAR) has been shown in the hippocampal neurons of AD-affected patients (Webber et al., 2006). Thus, given the role of StAR in the mitochondrial transport of cholesterol, these data strongly suggest that mitochondrial cholesterol accumulation may actually occur in patients with AD. However, further studies are required to confirm this increase.”
Despite epidemiological evidence that cholesterol-lowering drugs can stave off AD, clinical trials of statins have not shown positive effects. The new study points to another, related target for treatment. “The important point in the study is that given the key role of mitochondrial GSH in modulating Aβ neurotoxicity, this specific pool of GSH should be targeted for therapy,” say Fernandez-Checa and Colell. “This is critical because just increasing the total cellular GSH pool may not necessarily result in boosting the mitochondrial GSH due to the impairment in the transport into mitochondria imposed by cholesterol. Thus, the combination of lowering cholesterol and the use of cell permeable GSH precursors such as GSH ethyl ester may be of relevance in the treatment of the disease.”—Pat McCaffrey.
Fernandez A, Llacuna L, Fernandez-Checa JC, Colell A. Mitochondrial cholesterol loading exacerbates amyloid beta peptide-induced inflammation and neurotoxicity. J. Neurosci. 20 May 2009; 29(20):6394-6405. Abstract