22 August 2008. Another preclinical study presented at the International Conference on Alzheimer’s Disease, held 26-31 July in Chicago, capitalized on tau’s microtubule-binding property. This feature has inspired the notion that pathologic forms of tau—such as those that develop into the hallmark AD tangles—destabilize microtubules and thereby cause neurodegeneration. Consistent with this idea, previous work (Michaelis et al., 2004; Zhang et al., 2005) has demonstrated that microtubule-stabilizing agents such as the cancer drug paclitaxel (aka Taxol) can fend off Aβ-mediated neurodegeneration. The Zhang et al. study, led by John Trojanowski at University of Pennsylvania School of Medicine, Philadelphia, also showed that very low doses of paclitaxel could reverse motor deficits characteristic of PrPT44 mice, which overexpress the shortest human tau isoform in the central nervous system.

Cathy Andorfer and her collaborators at Mayo Clinic, Jacksonville, Florida, and Bristol-Myers Squibb, Wallingford, Connecticut, were intrigued by this finding. As were Sigurdsson and colleagues (see Part 1 of this series), these investigators worried that the motor impairments—deriving from predominantly spinal cord expression of transgenic tau in PrPT44 mice—would confound cognitive tests needed to show drug efficacy. Furthermore, taxols penetrate the brain poorly and thus seemed unlikely to reach the clinic as an AD treatment. Researchers at Bristol-Myers Squibb had identified an epothilone (another microtubule-binding cancer drug) that efficiently crosses the blood-brain barrier, and wanted to test this compound in a tauopathy mouse model.

They and Andorfer chose rTg4510 mice, a well-characterized transgenic line that expresses a tetracycline-inducible tau mutant (P301L) primarily in the cortex and hippocampus, and hence lacks a motor phenotype. These mice accumulate abnormal tau in AD-relevant areas as early as 1.5 to two months of age, with mature tangles and measurable cognitive defects by four to five months and gross brain shrinkage by 10 months (see ARF related news story). Andorfer and colleagues initially tested the mice at 2.5 months of age using the Morris water maze to establish baseline memory function, and thereafter began weekly injections of either vehicle or one of two epothilone doses (each 10-100-fold below the dose used in oncology). At 4.5 months, the mice were retested in the water maze and sacrificed for histopathologic analysis a month later.

Mice treated with the lower (1 mg/kg) epothilone dose were able to learn the location of the submerged platform—a spatial memory task that stymied the vehicle-treated and high-dose (10 mg/kg) groups. This treatment effect held up in post-training probe trials. The cognitive benefits were associated with a significant reduction in the hippocampal CA1 neuronal loss typical of rTg4510 mice. Low-dose epothilone treatment also noticeably affected tau pathology, reducing hyperphosphorylated and conformationally abnormal tau, as well as Gallyas silver-positive neurofibrillary tangles. The latter results in particular came as a surprise, Andorfer said, because no significant pathological changes were seen in the earlier paclitaxel study (Zhang et al., 2005). Coupled with the absence of obvious adverse consequences such as weight loss or motor dysfunction, these findings support the hypothesis that tau dysfunction drives microtubule dynamics through a destabilization effect that epothilones and related drugs appear to counter. Bristol-Myers Squibb collaborator Charlie Albright said that further mouse studies are underway to better understand the drug’s mechanism, which will, in turn, provide new insight into how abnormal tau wreaks havoc in the brain in AD and related diseases. For more on that, see Part 3 of this series.—Esther Landhuis.

This is Part 2 of a three-part series on tau. See also Part 1 and Part 3.