6 December 2008. One thing stood out at this year’s Society for Neuroscience annual meeting, held in Washington, DC, 15-19 November: researchers are not shy about putting new therapies through preclinical paces. While some of those potential therapies clearly aim to reduce what many consider the essence of AD pathology—Aβ and tau—other strategies, including those targeting inflammation, transcriptional control, and calcium toxicity, clearly added some spice to the mix.
One therapy that has been shown to lower brain Aβ in animals is polyphenol-rich grape seed extract (see ARF related news story). This palatable concoction may also help reduce tau toxicity, according to a presentation by Lap Ho from Giulio Pasinetti’s group at Mount Sinai Medical School, New York. Ho showed that a grape seed extract (GSE) helped rescue tau toxicity in various mouse models. In Tg2576 mice, phospho-tau241 immunoreactivity disappears and tau no longer forms fibrils when the animals take the extract with their food. In fly and mouse models of tauopathy, GSE effects were more distinct. Flies expressing the R406W mutant form of tau, which causes an inherited form of frontotemporal dementia, have a deformed eye phenotype; that is, their eyes are smaller and distorted. The extract partially rescued this defect, and the animals scored higher in visual screening tests. In the JNPL3 mouse, which expresses P301L mutant tau that causes frontotemporal dementia in other families, 150 mg/Kg/day of the extract (equivalent to about 800 mg/day in humans) rescued motor dysfunction and reduced fatality rates, Ho reported at SfN.
It appears that polyphenols in the extract may behave as general inhibitors of protein fibrillization. Ho showed that, in vitro, the grape seed extract reduced aggregation of a synthetic tau hexapeptide (VQIVYK), which serves to model tau aggregation. The GSE reduced thioflavin fluorescence resulting from aggregation of the hexapeptide, and results from photo-induced cross-linking of unmodified proteins (also known as PICUP) assays showed that the extract prevented peptide oligomerization. In a different presentation, Pasinetti lab collaborator David Teplow, University of California at Los Angeles, reported similar effects on Aβ. In thioflavin fluorescence and EM studies, Teplow found that the grape seed extract MegaNatural-AZ, a commercially available nutraceutical, prevented Aβ fibrillization, while PICUP assays also revealed reduced oligomerization. This work appeared in the 21 November Journal of Biological Chemistry (see Ono et al., 2008). Both Teplow and Ho used circular dichroism spectrometry to demonstrate that the extract prevented formation of β-sheet structures in Aβ and the tau hexapeptide, respectively, and that it eliminated paired helical formations of tau. The GSE even broke apart pre-formed tau aggregates, increasing their dissociation within 15 minutes and in a dose-dependent manner, reported Ho. Teplow showed that MegaNatural-AZ prevented cell death as judged by lactate dehydrogenase release and MTT (3-[4, 5-dimethylthiazol-2-yl]2,5-diphenyl-tetrazolium bromide) reduction, a measure of cell metabolic activity.
In addition to their anti-amyloidogenic potential, polyphenols also have anti-inflammatory activity, and this complicates the interpretation of their in-vivo effects (see ARF related news story). In fact, Yan-Jiang Wang, Flinders University, Adelaide, Australia, reported that GSE reduces not only Aβ but also inflammatory responses when given to three-month-old double transgenic mice (APPSwe/PS1dE9). At 12 months, Aβ deposits were down 45 percent in GSE-treated animals and microgliosis was down 70 percent, suggesting that at least some of the effects of polyphenols may be due to anti-inflammatory activity.
Anti-inflammatory agents have disappointed in AD clinical trials, though they continue to be screened in clinical and preclinical trials. One anti-inflammatory that has been considered is triflusal, which prevents activation of the inflammatory transcription factor NF-κB. Lidia Sereno from Teresa Gomez-Isla’s laboratory at the Universitat Autònoma de Barcelona, Spain, reported that when given to 10-month-old Tg2576 mice for three months, triflusal dampened both glial activation in the brain and expression of the proinflammatory markers IL-1β and TNFα. This happened alongside increased expression of brain-derived neurotrophic factor and c-fos, which is activated during learning. In behavioral tests, the animals performed better than untreated transgenic controls in a Morris water maze and contextualfear conditioning tests of learning and memory. Treated mice also had significantly reduced plaque load as determined by thioflavin S staining. The findings suggest that triflusal might have some potential as an AD therapeutic. Alas, a recent clinical trial run by Gomez-Isla and colleagues was halted prematurely due to recruitment problems (see ARF related news story and Gomez-Isla et al., 2008).
Non-steroidal anti-inflammatories have been given a fair shake in clinical trials, but one that has not drawn much attention is tolfenamic acid, currently approved in Europe for migraine. Lina Adwan from the University of Rhode Island, Kingston, reported that the drug might be worth studying as a potential AD therapeutic as well, though not because of its NSAID activity. Tolfenamic acid apparently reduces activity of the transcription factor Sp1 in pancreatic tumors. Because Sp1 has been implicated in regulation of APP expression, Adwan and colleagues examined the effect of the compound on APP levels. They found that in wild-type mice, tolfenamic acid reduced APP expression in the cerebral cortex and, slightly but statistically significantly, reduced Aβ40 levels in the brain as well. Treated guinea pigs showed a similar trend. It remains to be seen how this drug might work in AD mouse models and exactly how it affects Sp1 activity. “Indications are that it may activate a protease that degrades the transcription factor,” said Adwan.
Another signaling molecule that came in for scrutiny was macrophage colony stimulating factor (MCSF). Researchers led by Tony Wyss-Coray at Stanford University, Palo Alto, California, reported last year that MCSF was one of a panel of 18 blood markers that has diagnostic potential for AD (see ARF related news story). In Washington, Jian Luo, from Wyss-Coray’s lab, reported that in an independent analysis performed since then, MCSF plasma levels were significantly lower in AD patients (n = 85) than age-matched controls (n = 127). To see if increasing MCSF might help ameliorate AD pathology, Luo and colleagues administered this protein intraperitoneally to seven- to eight-month-old APP transgenic mice (Thy1-hAPP). Ten weeks later, the treated mice performed much better in the Morris water maze. Similarly, older mice (16-18 months) treated three times a week for four weeks performed better than controls. Curiously, these improvements occurred without any change in brain levels of soluble or insoluble Aβ, or in microglial activation.
To further probe the actions of MCSF, Luo and colleagues turned to the kainic acid excitotoxicity model. When administered 24 hours before kainite, MCSF almost completely blocked neuronal loss, measured five days later in the CA1 region of the hippocampus, and prevented loss of calbindin, used as a marker of neurodegeneration. Luo showed that kainate turned up activation of microglia, as judged by increased immunoreactivity to microglial markers CD68 and CD11b, but that this could be prevented by addition of MCSF. The numbers of microglia, on the other hand, were unchanged (determined by Iba1 immunoreactivity). Luo concluded by suggesting that MCSF might improve learning and memory performance in APP transgenic mice because it counteracts the excitotoxic effects of Aβ, making MCSF a potential future therapeutic for AD.
A further cytotoxin that has been linked to neurodegeneration in AD models is calcium (see related ARF live discussion). A poster by Dong Liu, from Mark Mattson’s lab at the NIA, described a novel way to reduce calcium toxicity by wedging open potassium ATP (K-ATP) channels. The researchers previously reported that diazoxide, a K-ATP channel opener, can protect against neuronal death caused by ischemia and apoptosis (see Liu et al., 2003). In Washington, Liu showed that diazoxide can improve learning and memory when chronically administered to 3xTg mice expressing human mutant APP, presenilin, and tau. The researchers began diazoxide treatment by adding the channel opener to the drinking water when the mice were four months old, and at 12 months tested their behavior. In both the Morris water maze test of learning and memory and in an open field test of spontaneous locomotor activity, the treated animals outperformed controls. Postmortem examination showed that treated animals had a dramatic reduction in Aβ levels in the hippocampus and, to a lesser degree, in the cortex. Levels of phosphorylated tau were also lower in treated animals.
Liu said that it is not clear how diazoxide works because it binds to both mitochondrial and cell membrane K-ATP channels. However, the fact that it hyperpolarizes cells and reduces calcium influx in hippocampal neurons suggests that dampening calcium cytotoxicity may be a crucial aspect. In the treated 3xTg mice, diazoxide also reduced oxygen consumption and increased cerebral blood flow, which might also be important factors in reducing pathology.
However tantalizing some of these approaches might seem, it should be kept in mind that promising preclinical strategies often fail miserably when applied to real-life disease. Trials of immunotherapies, vitamins, anti-inflammatories, statins, and most recently gingko biloba (see ARF related news story) are testament to the difficulties in translational research. As always, the real proof of the pudding will be in the eating.—Tom Fagan.