4 October 2013. In healthy neurons, tau proteins stay in the axons, where they stabilize microtubules. In cells that have been poisoned by Aβ oligomers, however, tau accumulates in dendrites, and the microtubules fall apart. In the September 24 issue of The EMBO Journal, researchers led by Eva-Maria Mandelkow at the German Center for Neurodegenerative Diseases, Bonn, propose a molecular pathway for microtubule decay, a critical feature of Alzheimer’s and other neurodegenerative diseases. When tau invades dendrites, the authors report, enzymes tack polyglutamate side chains onto microtubules. This post-translational modification recruits and activates spastin, an ATPase that breaks microtubules by a mechanism that is not understood. Mandelkow and colleagues based their study on cell culture experiments. If their findings can be replicated and further developed in vivo, that would suggest therapeutic value to targeting microtubule-severing proteins, and lend support to ongoing development of microtubule stabilizers.
Mandelkow and others had previously reported that neurons cultured with Aβ oligomers accumulate dendritic tau and lose microtubules (see ARF related news story; Zempel et al., 2010; ARF related news story; Hoover et al., 2010). In the present study, first author Hans Zempel and colleagues explored why that might be. It is commonly assumed that Aβ induces tau hyperphosphorylation, which detaches tau from microtubules and renders them unstable. However, studies on rat neurons exposed to synthetic Aβ oligomers (in a 70-30 mix of Aβ40 and Aβ42) showed tau phosphorylation occurring about an hour after the researchers spotted tau missorting into dendrites and microtubules disassembling. That suggested that microtubule breakup is caused by something other than tau phosphorylation.
To find the cause, the researchers considered enzymes known to destabilize microtubules, focusing on the ATPases katanin and spastin. Immunocytochemistry with antibodies to these enzymes and to the microtubule building block tubulin revealed no change in the level of katanin or its localization after culturing rat hippocampal neurons with 1 microM synthetic Aβ oligomers. Instead, the authors found that spastin accumulated at the microtubules. Spastin recognizes substrates coated with glutamate amino acids and, sure enough, levels of polyglutamylated microtubules were doubled in Aβ-treated cells, a finding consistent with spastin recruitment to these cytoskeletal struts. Spastin overexpression occurs in human brain tumors (Draberova et al., 2011), and loss-of-function mutations in the enzyme can cause hereditary spastic paraplegia, a neurodegenerative condition marked by progressive stiffening and weakening of the legs (Charvin et al., 2003; Salinas et al., 2007).
How do microtubules rack up polyglutamates in the first place, the authors wondered? Enzymes called glutamylases regulate microtubule dynamics by adding polyglutamate chains to C-terminal glutamate residues on tubulins. Previous work identified tubulin-tyrosine-ligase-like-6 (TTLL6) as the main enzyme responsible for polyglutamylation leading to microtubule breakdown (Lacroix et al., 2010). In immunostaining experiments, the authors saw that TTLL6 lit up strongly in microtubules of Aβ-exposed neurons compared with control cells, and the enzyme co-localized with polyglutamylation sites in dendrites with missorted tau. Transfecting neurons with TTLL6 triggered microtubule polyglutamylation, spastin recruitment, microtubule breakdown, and tau mislocalization—even when no Aβ was present.
“The loss of microtubules after Aβ oligomer exposure is caused by polyglutamylation of microtubules by TTLL6, followed by recruitment of spastin and then severing,” the authors wrote. Tau mislocalization would be downstream of that cascade. Mandelkow does not know what drives TTLL6 to the microtubules.
The role of tau in tauopathies remains controversial. Some labs found more neurodegeneration (Dawson et al., 2010) and Parkinson’s disease-like symptoms (see ARF related news story) in tau-deficient mice, while others reported that tau reduction helped mice resist Aβ toxicity (see ARF related news story; ARF related news story). The current study adds to the evidence for tau being detrimental. In cultured neurons from tau-knockout mice, the pathological chain of events this study attributes to Aβ oligomers—TTLL6 and spastin recruitment, microtubule polyglutamylation, microtubule loss—were absent or sharply reduced. However, the changes reoccurred if the researchers transfected tau-deficient cells with human tau. These data support ongoing efforts to develop Alzheimer’s therapeutics that curb tau accumulation in neurons (see Hosokawa et al., 2012; ARF related news story), the authors believe.
In addition, the findings strengthen the rationale for testing brain-penetrant microtubule-stabilizing agents, such as epothilone D, in people with Alzheimer’s, said Kurt Brunden of the University of Pennsylvania School of Medicine, Philadelphia (see full comment below). This compound, and other cancer drugs that stabilize microtubules, improve outcomes in mouse models of tauopathy (see Brunden et al., 2010; Zhang et al., 2012; Barten et al., 2012; see also ARF related news story; and ARF conference story). Epothilone D is under investigation in a Phase 1b AD trial.
However, the current study suggests that microtubule loss results not only from instability, but also from excess cutting by spastin. That would argue for the development of new drugs that inhibit microtubule-severing proteins, Daphney Jean and Peter Baas of Drexel University College of Medicine, Philadelphia, Pennsylvania, noted in an accompanying commentary (Jean and Baas, 2013).
The study also addresses the longstanding mystery of how errant tau finds its way from axon to dendrite in AD. “The answer is, it doesn’t,” Eckhard Mandelkow told Alzforum. “Tau does not directly relocalize to dendrites, it is newly synthesized there.” The authors determined this by showing that tau carrying a fluorescent tag did not move into dendrites after treatment with Aβ oligomers, and that silencing tau translation with short-hairpin RNAs blocked the appearance of dendritic tau.
Aβ oligomers trigger calcium influx into the neuron, and this correlates with microtubule loss in dendrites. However, it is not clear how the calcium increase downstream of Aβ might unleash TTLL6 and spastin on microtubules, nor how the destruction of microtubules relates to newly synthesized dendritic tau. Neither TTLL6 nor spastin have been previously implicated in Alzheimer’s disease.—Esther Landhuis
Zempel H, Luedtke J, Kumar Y, Biernat J, Dawson H, Mandelkow E, Mandelkow EM. Aβ oligomers induce synaptic damage via Tau-dependent microtubule severing by TTLL6 and spastin. EMBO J. 24 Sep 2013. Abstract