In Alzheimer and some other neurodegenerative diseases, the microtubule-associated protein tau weaves a web of neurofibrillary tangles (NFTs) that might be toxic to neurons. But alternative interpretations suggest that NFTs, by sequestering tau, are in fact neuroprotective, or, at the very worst, they are the product of an incidental side reaction. To untangle this sticky problem, Karen Hsiao Ashe of the University of Minnesota Medical School in Minneapolis and collaborators from Massachusetts General Hospital and the Mayo Clinic in Jacksonville, Florida, have generated a new tau transgenic mouse in which the overexpression of mutant human tau can be regulated by tetracycline. In this model, turning off tau expression after deposits have formed halts neuron loss and reverses memory defects. But surprisingly, neurofibrillary tangles continue to accumulate, suggesting that they are not responsible for neurodegeneration.

The data, reported in today’s Science, mesh with a recent report from Peter Davies and Karen Duff (Andorfer et al., 2005), which suggested that neuronal loss induced by human tau expression in mice does not track with neurofibrillary tangle formation. Together, these results should turn attention away from tangles to a search for other mechanisms of tau-induced neuronal death. And the new observation that some tau-related cognitive deficits are reversible in mice raises the hope that early intervention to blunt tau’s damage in humans could yield benefits in AD and other tauopathies.

The new tau mice, the product of a three-way collaboration between co-first authors Karen SantaCruz, Jada Lewis, and Tara Spires, manifested tau toxicity early in life as a result of overexpression of the P301L human tau mutant in the forebrain. By the time they were 4 months old, the mice had visible tau deposits. At 5.5 months, the animals showed a decrease in brain weight and in the number of hippocampal neurons. Pathology progressed by 10 months to include gross atrophy of the forebrain, accompanied by accumulation of NFTs containing hyperphosphorylated tau. Behaviorally, by 2.5 months the mice had impaired performance in the Morris water maze, while by 4 months old they were swimming randomly, a sign of severe spatial memory deficits.

After characterizing the “tau-on” mice, the investigators switched tau off by administering doxycycline, which represses the tetracycline promoter, and watched what happened to the tangles, neurons, and behavioral and cognitive performance. In mice fed doxycycline, tau mRNA decreased by 85 percent and soluble tau protein levels diminished by three-quarters. When tau was switched off at 2.5 months, the progression of tangle formation and neuronal loss was halted. But if repression of human tau was delayed until 4 months, NFTs continued to increase. Around this age, a 64-kd insoluble tau species first appeared, which the authors speculate could act as a sink for the residual mutant tau in these older mice.

But despite the continued accumulation of tangles, turning off tau in these older mice did protect them against neuron loss and cognitive deficits. When tau was suppressed continuously from 5.5 months to 9.5 months, the mice did not lose brain weight, for example. Turning off tau for short periods between 2.5 and 8.5 months also stabilized the number of hippocampal neurons, suggesting that NFTs “do not invariably cause neuron death,” write the authors.

In the memory task, early cognitive deficits observed in young (2.5-month-old) mice in the Morris water maze actually improved when tau was turned off, suggesting that the processes leading to early memory defects might be reversible. Tau suppression at later times (5.5 months) elicited an improvement of memory function, but not to levels seen in young mice. “That memory function improved in > 4-month-old mice fed doxycycline despite ongoing accumulation of NFTs clearly implies dissociation between the processes that lead to memory loss and those that cause NFTs, and that the NFTs remaining after tau suppression are not sufficient to disrupt cognitive function,” the authors conclude.—Pat McCaffrey


  1. In this study, SantaCruz and colleagues have created an inducible mutant tau transgenic model, and observed the neuropathological consequences of strategically turning on and off tau expression for set periods. Profound forebrain neurodegeneration and memory loss by tau overexpression was observed. Remarkably, following termination of tau expression, tau hyperphosphorylation and tangle formation progressed while memory function recovered. These findings provide compelling evidence that neurofibrillary tangles do not directly serve a role in neuronal loss or cognitive impairment. This is consistent with a recent report by Peter Davies and colleagues that neurofibrillary pathology was not correlated with neuronal death in a human tau transgenic model displaying neurodegeneration (Andorfer et al., 2005). The study by SantaCruz et al. also provides the exciting prospective that recovery of cognitive function is possible even after significant progression of neurodegeneration. These findings have profound implications in the understanding of AD and other neurodegenerative disorders featuring tauopathy.


    . Cell-cycle reentry and cell death in transgenic mice expressing nonmutant human tau isoforms. J Neurosci. 2005 Jun 1;25(22):5446-54. PubMed.

  2. SantaCruz and colleagues manipulated transgenic tau levels using the tet-off system. Their results indicate that NFT accumulation is not sufficient to cause neuronal death and cognitive decline. This result was a bit of a surprise, because NFT formation increased even after transgenic tau expression was suppressed. Memory function also improved as suppressing tau expression halted brain atrophy. These results suggest that NFT formation, neuronal death, and cognitive decline may occur through different mechanisms in tauopathies. Moreover, it suggests that developing a tau aggregation inhibitor might not produce therapeutic benefits in such diseases. However, these results might very well be due to the effects of tau overexpression. Mandelkow’s group clearly showed that tau overexpression impaired anterograde axonal trafficking, which may cause synaptic dysfunction, eventually inducing memory failure and neuron death. When suppressing tau expression, axonal transport recovered. Memory function, therefore, might improve with the return of synaptic function.

    Once the tau concentration increases, conformational changes, which account for the earliest aberrations in tau in AD (Weaver et al., 2000), occur that produce tau by-products that are resistant to degradation. These residual products form tau fibrils, NFTs, thereby explaining why NFTs increase even after tau expression is suppressed, and they may generate tau oligomers that cannot be removed. Tau fibrils would then remain even after suppressing tau overexpression. In the authors’ mouse model, the effect of tau overexpression is commonly used to express phenotype; yet, patients with the P301L mutation do not have overexpressed mutant tau. Rather, mRNA levels of mutant tau are the same as wild tau levels, and soluble mutant tau levels are smaller than those for wild tau (Miyasaka et al., 2001). This mutant tau expression, albeit small, induces tauopathy. Thus, it is not overexpression, but tau mutations that provoke tauopathy (that is, NFT formation, neuronal loss, and dementia). Therefore, investigating the effects of tau mutation might provide insights into what is happening in patients with FTDP-17.


    . Some problems concerning the determination of different corrinoids by the plate method with Escherichia coli 113-3. Acta Microbiol Pol B. 1975;7(2):103-10. PubMed.

    . Selective deposition of mutant tau in the FTDP-17 brain affected by the P301L mutation. J Neuropathol Exp Neurol. 2001 Sep;60(9):872-84. PubMed.

  3. Neurofibrillary Tangles: Villain or Merely Vilified? What’s Next? TADDLs/TAuDDLs!
    For most investigators in the field of Alzheimer disease (AD), pathology has become equated with pathogenesis. As such, the two leading theories concerning the disease revolve around the amyloid-β of senile plaques and the phospho-tau of neurofibrillary tangles (NFTs). Contrasting this viewpoint, there a growing contingency suggesting that pathology is not central to pathogenesis; rather, pathology may even be the anti-pathogenesis and, rather than causing the disease, the pathology may be protecting from the disease (Rottkamp et al., 2002; Smith et al., 2002; Lee et al., 2005). A recent transgenic mouse study from Karen Ashe and colleagues (Santacruz et al., 2005) goes a long way toward bringing light to this specific issue. These studies quite convincingly demonstrate that NFT-like accumulations of phospho-tau are not associated with neurodegeneration, echoing a recent similar conclusion from the Peter Davies group (Andorfer et al., 2005). While these studies implicate NFTs as beneficial, since they are disconnected with cell death, better evidence for the exoneration of phospho-tau/NFTs in the disease process is provided by human studies (reviewed in Lee et al., 2005). For example, in patients with AD, phospho-tau and NFTs are unlikely to be factors in the neurodegenerative process since, while both the total number and length of microtubules are significantly and selectively reduced in pyramidal neurons, such decrements are unrelated to NFTs (Cash et al., 2003). This poses the question as to what causes the neurodegeneration. One suggestion is that NFTs sequester “abnormal” phospho-tau microaggregates, thereby rendering them harmless (Binder et al., 2005). Therefore, much like micro-aggregates of amyloid-β (ADDLs; Klein, 2002), the tau story has shifted to micro-aggregates of phospho-tau…let’s call them TADDLs or TAuDDLs (Tau-Derived Diffusible Ligands). The evidence for such TADDLs, much like their counterpart ADDLs (Lee et al., 2004), is scant but at least it keeps the prime suspects front and center. However, as we recently reviewed (Lee et al., 2005), there is far, far more evidence that phospho-tau, as TADDLs or NFTs, are a protective compensatory response mounted by neurons in an effort to stave off another driving pathogenic force. Indeed, oxidative damage not only chronologically precedes phospho-tau (Nunomura et al., 2000; Liu et al., 2005), but, supporting a protective response, such oxidative stress is significantly reduced in AD in neurons containing phospho-tau (Nunomura et al., 2001). Perhaps the time has come to reassess our old prejudices and exonerate the “villain.”


    . Cell-cycle reentry and cell death in transgenic mice expressing nonmutant human tau isoforms. J Neurosci. 2005 Jun 1;25(22):5446-54. PubMed.

    . Tau, tangles, and Alzheimer's disease. Biochim Biophys Acta. 2005 Jan 3;1739(2-3):216-23. PubMed.

    . Microtubule reduction in Alzheimer's disease and aging is independent of tau filament formation. Am J Pathol. 2003 May;162(5):1623-7. PubMed.

    . Abeta toxicity in Alzheimer's disease: globular oligomers (ADDLs) as new vaccine and drug targets. Neurochem Int. 2002 Nov;41(5):345-52. PubMed.

    . Perspectives on the amyloid-beta cascade hypothesis. J Alzheimers Dis. 2004 Apr;6(2):137-45. PubMed.

    . Tau phosphorylation in Alzheimer's disease: pathogen or protector?. Trends Mol Med. 2005 Apr;11(4):164-9. PubMed.

    . Alzheimer-specific epitopes of tau represent lipid peroxidation-induced conformations. Free Radic Biol Med. 2005 Mar 15;38(6):746-54. PubMed.

    . Neuronal oxidative stress precedes amyloid-beta deposition in Down syndrome. J Neuropathol Exp Neurol. 2000 Nov;59(11):1011-7. PubMed.

    . Oxidative damage is the earliest event in Alzheimer disease. J Neuropathol Exp Neurol. 2001 Aug;60(8):759-67. PubMed.

    . The state versus amyloid-beta: the trial of the most wanted criminal in Alzheimer disease. Peptides. 2002 Jul;23(7):1333-41. PubMed.

    . Tau suppression in a neurodegenerative mouse model improves memory function. Science. 2005 Jul 15;309(5733):476-81. PubMed.

    . Amyloid-beta and tau serve antioxidant functions in the aging and Alzheimer brain. Free Radic Biol Med. 2002 Nov 1;33(9):1194-9. PubMed.

  4. I found this paper fascinating and very intriguing, but could not determine whether the effects attributed to lowered tau levels (from the otherwise very high levels) are instead due to a neuroprotective effect of the drug used to control tau expression.

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Cell-cycle reentry and cell death in transgenic mice expressing nonmutant human tau isoforms. J Neurosci. 2005 Jun 1;25(22):5446-54. PubMed.

Further Reading

Primary Papers

  1. . Tau suppression in a neurodegenerative mouse model improves memory function. Science. 2005 Jul 15;309(5733):476-81. PubMed.