25 March 2011. Before forming neurofibrillary tangles, tau may pick up a few acetyl groups along with the slew of phosphates it brings to the pathological inclusions in tauopathies such as Alzheimer’s disease and frontotemporal lobar dementia (FTLD). Researchers report in the March 22 issue of Nature Communications that tau is acetylated in vitro, in cell culture, in a mouse tauopathy model, and in several human tauopathies. They observed no acetylated tau in healthy tissue, suggesting acetylation could be part of the pathological process. The acetylated tau also made insoluble fibrils more rapidly than normal tau. Coupled with a report last fall that indicated acetylation slows degradation of tau, the findings suggest that this form of protein modification might play a major role in regulating tau toxicity.
First author Todd Cohen and senior author Virginia Lee led the study at the University of Pennsylvania School of Medicine in Philadelphia. The work confirms an earlier report of tau acetylation by Li Gan and colleagues at the Gladstone Institutes in San Francisco, California (see ARF related news story on Min et al., 2010). The current data are “very clear, black and white, very solid,” said Li-Huei Tsai of the Picower Institute in Cambridge, Massachusetts. “Now there are two papers indicating acetylation of tau in aggregation pathology…that is definitely reassuring,” said Tsai, who was not involved in either study.
Cohen joined Lee’s group two years ago with an interest in acetylation of proteins involved in neurodegeneration. He started by testing whether tau or α-synuclein could be acetylated in vitro. α-synuclein was not, but tau was. When he incubated recombinant tau protein with the acetyltransferase Creb-binding protein (CBP), the tau acquired acetyl groups. Tau in the presence of CBP also reacted with an antibody to acetylated lysines. In HEK293 human embryonic kidney cells, tau acetylation rose when Cohen added an inhibitor of deacetylation, trichostatin A. The researchers used mass spectrometry to determine that lysines 163, 280, 281, and 369 were modified under these conditions.
Lysines 280, 281, and 369 occur in the microtubule-binding regions of tau, and the authors hypothesized that the acetylation would interfere with the tau-microtubule interactions. In vitro, acetylated tau did not promote microtubule assembly as well as the un-acetylated form did. In cell culture, a mutant tau in which each of the four lysines was swapped for a glutamine—to mimic the acetylated state—bundled microtubules poorly in comparison to wild-type tau.
“At this point, we were fairly certain that tau, at least in these artificial situations—in vitro, in cells—was a substrate for acetyltransferases,” Lee said. But what about in a whole animal? To test this, the scientists engineered an antibody specific for the acetylated lysine 280 (K280) form of tau and used it to examine two mouse models of tauopathy: a PS19 line expressing the P301S tau mutant, and one expressing that tau mutation plus the V717F human amyloid precursor protein mutant (Hurtado et al., 2010). Both evinced immunoreactivity with the acetyl-tau antibody in the cortex and hippocampus, compared to wild-type mice. The double-mutant mice showed greater immunoreactivity as they aged. The acetyl-tau signal frequently colocalized with that for the AT8 phospho-tau antibody and with thioflavin S staining for amyloid, indicating that tau tangles contain acetylated, phosphorylated tau.
Finally, the researchers examined tau acetylation in human samples, tapping coauthor John Trojanowski’s extensive set of cortical sections from people who died from various tauopathies. While they observed no signal from the acetyl-tau antibody in control samples, they did see antibody staining in Alzheimer’s disease, frontotemporal lobar dementia with tau pathology (both corticobasal degeneration and progressive supranuclear palsy), argyrophilic grain disease, tangle predominant senile dementia, and even some signal from Guam parkinsonism-dementia complex. The only negative result was from Pick’s disease, in which the predominant tau isoform lacks K280. Though sample numbers were small, it appears tau acetylation may be a common feature of many tauopathies.
Lee suggested that as such, tau acetylation might be useful as a biomarker to distinguish different forms of frontotemporal lobar dementia (FTLD). While some FTLDs have tau pathology, others exhibit inclusions of the protein TDP-43 instead, but levels of phosphorylated tau are the same between the two groups. Lee speculated that acetyl-tau levels in the cerebrospinal fluid might allow clinicians to differentiate patients that need tau-based therapies from those that would be best served by TDP-43-related drugs. No such drugs are currently approved, and, at this point, the idea is only “a hope and a prayer,” Lee said, but she thinks it is a possibility worth exploring.
“What is the significance of this post-translational modification of tau in Alzheimer’s disease pathogenesis?” asked Khalid Iqbal of the New York State Institute for Basic Research in Staten Island, who was not involved in the study. Acetylation might promote phosphorylation, or block tau degradation, as suggested in Gan’s study. Acetylated tau is also more fibrillogenic, which could be a key pathological change. Using sedimentation analysis, electron microscopy, and thioflavin T staining to examine tau aggregation, Lee and colleagues found that acetylated tau formed fibrils much faster than non-acetylated. However, the current evidence is not yet enough to convince Iqbal that acetylation stands among the most important factors in disease. For example, he noted, it is not yet known how much of tau in fibrils is actually acetylated.
A key step toward understanding the importance of acetylation will be to identify the enzymes that acetylate and deacetylate tau. These enzymes would also be potential therapeutic targets, Lee noted. Cohen’s experiments with histone deacetylase 6 suggest it affects tau acetylation, while Gan’s study pointed to roles for histone acetyltransferase p300 and the deacetylase SIRT1; none has yet been proven to be a primary modifier of tau in vivo. Perhaps both deacetylases are correct, Tsai suggested: “I would not be surprised if there were redundant functions.”
Also unknown at this point is whether tau acetylation is a purely pathological phenomenon, or whether is serves some function in a healthy cell. One of the acetylation motifs in tau also appears in the related microtubule associated proteins MAP2 and MAP4. The study authors suggest that perhaps acetylation is a mechanism to regulate MAP-microtubule binding. Cohen might not have been able to observe acetylation in normal samples, Lee suggested, because deacetylation happens quickly. “I think it is likely that there is a role” in normal development, said Mark Mattson of the National Institute on Aging in Baltimore, Maryland. He speculated, for example, that acetylation might act to knock tau off microtubules when axons reach the appropriate length and stop growing.—Amber Dance.
Cohen TJ, Guo JL, Hurtado DE, Kwong LK, Mills IP, Trojanowski JQ, Lee VM. The acetylation of tau inhibits its function and promotes pathological tau aggregation. Nat Commun. 2011 Mar;2:252. Abstract