Tangles and Neurodegenerative Disease 

<DIV align=center>Tangles and Neurodegenerative Disease — A Surprising Twist
</DIV>Rudolph E. Tanzi, Ph.D.Nearly 100 years ago, a group of Bavarian psychiatrists gathered in a small conference hall in Tübingen to hear a presentation by Dr. Alois Alzheimer. To their surprise, he proposed that presenile dementia in his 51-year-old patient might be related to two brain lesions he referred to as "miliary bodies" lying outside of neurons and "dense bundles of fibrils" choking the interiors of neurons. Today, we consider these lesions to be the neuropathological hallmarks of the disease that is named after him and refer to them as senile plaques and neurofibrillary tangles, respectively. Although abundant levels of these brain lesions are required to confirm the diagnosis of Alzheimer's disease, the issue of whether they are neurotoxic, protective, or simply incidental markers of disease has remained controversial. The numbers of neurofibrillary tangles are more closely correlated with neuronal loss and dementia than are the numbers of senile plaques.¹ Until now, however, the potential neurotoxicity of neurofibrillary tangles has not been evaluated experimentally. A recent report by Santacruz and colleagues² addresses this issue and provides surprising and encouraging data relevant to the goal of clinically reversing the ravages of Alzheimer's disease and related dementias.

The authors engineered mice to express a mutant form of the tau gene, known to cause frontotemporal dementia, so they could turn the gene on and off by exposing the mice to doxycycline (Figure 1). Tau is a constituent of microtubules and the main component of neurofibrillary tangles, where it appears in a hyperphosphorylated form.³ Expression of the mutant tau gene resulted in age-related loss of neurons, atrophy of the forebrain, and tau-positive lesions, along with a dramatic age-dependent decline in spatial memory in the transgenic mice (Figure 1). Santacruz and colleagues then assessed the potential for the recovery of brain function during various stages of neurofibrillary abnormalities by deactivating the mutant tau gene. The progression of neurofibrillary abnormalities in mice was independent of the expression of mutant tau after four months of age, at which time the number of neurofibrillary tangles continued to increase in the absence of expression of the transgene. Surprisingly, even though the levels of neurofibrillary tangles steadily increased after the mutant tau gene was turned off, there was a clear plateau in the rate of neuronal loss and brain atrophy. Moreover, spatial-memory retention simultaneously improved, even in the presence of abundant neurofibrillary tangles and significant losses in brain weight and hippocampal neurons. Even mice that had already lost half their hippocampal CA1 neurons were able to learn and retain new information after the mutant tau gene had been suppressed. Coauthor Dr. Hsiao Ashe captured the striking and surprising nature of these results: "When I saw the memory getting better I actually thought I had done something wrong in the experiment."⁴

The study has profound implications for understanding both the pathogenic mechanism of tau-related lesions and the potential for restoring brain function in those with either Alzheimer's disease or frontotemporal dementia. It points the finger at the mutant tau protein and not the neurofibrillary tangle as the neurotoxic entity, supporting the possibility that the neurofibrillary tangle is an incidental marker for the neurotoxic cascade or may even represent a protective neuronal response aimed at sequestering mutant tau. Interestingly, a similar situation has emerged in recent studies of senile plaques and their principal component, the -amyloid peptide A₄₂. All but a handful of mutations in three genes for early-onset familial Alzheimer's disease increase the generation of A₄₂, which in turn enhances the formation of -amyloid fibrils and the deposition of these fibrils in senile plaques. However, the numbers of these lesions do not correlate well with the severity of dementia. Instead, small, pre-plaque oligomers of A₄₂ appear to cause neuronal damage, as evidenced by recent experiments showing that they interfere with synaptic function and impair memory in mice.⁵ And other studies indicate that both tau and A₄₂ wreak havoc on neurons and synapses long before neuropathological lesions form. Thus, the same hypothesis may pertain to tangles and plaques: each may represent neutral or possibly even protective final destinations for their protein components — tau and A₄₂, respectively.

The observation that brain function was restored in the mice despite dramatic neuronal loss, brain atrophy, and the presence of abundant neurofibrillary tangles suggests that recovery of cognitive function may be possible in the early-to-middle stages of dementia. However, it is important to be cautious when one is extrapolating disease-related results from mice to humans. Furthermore, the mutant form of tau used in this study does not cause Alzheimer's disease; thus, these results may be more relevant for understanding and eventually treating frontotemporal dementia and other tauopathies. These and other findings continue to prompt a paradigm shift in which pathogenicity is more strongly correlated with pre-lesion molecules of tau and A₄₂ than with neurofibrillary tangles and plaques. Therapeutically, these findings strongly suggest that removing the neurotoxic species of tau (and A₄₂) from the brain, by curbing their generation, interfering with the formation of their toxic conformers, or enhancing their clearance and degradation, may be our best bet for someday treating and preventing age-related dementias. In the meantime, let us hope for more surprises.

Figure 1. Recovery from Dementia in a Mouse Model.

A recent study by Santacruz and colleagues² suggests that the neurofibrillary tangles observed in the brains of patients with frontotemporal dementia may not cause the dementia. The authors engineered mice to express a tau gene containing a mutation that causes frontotemporal dementia. The mice produced abnormal tau protein and had a progressive loss of neuronal cells, atrophy of the forebrain, and memory impairment. Deactivation of the mutant gene did not prevent further accumulation of neurofibrillary tangles, but it did halt neuronal loss and brain atrophy. Surprisingly, memory also improved. The pathological effects of mutant tau may be caused by an insult to the integrity of microtubules — that is, incorporation of the mutant protein may destabilize microtubules and kill neuronal cells. The mutant tau proteins are subsequently sequestered into neurofibrillary tangles. P denotes phosphate.