For decades, the sticky amyloid-β peptide that accumulates into plaques around brain cells has garnered considerable attention from researchers who study Alzheimer's disease. Despite years of labor and millions of dollars invested in amyloid research, however, the neurodegeneration caused by Alzheimer's remains essentially untreatable, although several candidate treatments aimed at amyloid-β are now in clinical trials (CE&N, April 5, page 12).
Meanwhile, scientists are increasingly turning their attention to another culprit in this grim disease: the neuronal protein known as tau. Researchers in both academia and industry are studying the impact of malfunctioning tau and also searching for treatments to slow or even reverse its harmful effects on the brain.
"The ultimate result of tau dysfunction is that neurons lose their connections to other neurons, and when neurons are no longer communicating, that has profound effects on cognition—the ability to think and reason," says Travis Dunckley, who studies neurodegenerative diseases at the Translational Genomics Research Institute, in Phoenix.
In addition to Alzheimer's, tau pathology is associated with other "tauopathies" including frontotemporal dementia and progressive supranuclear palsy. It can also appear in Parkinson's patients. In Alzheimer's and most other tauopathies, malfunctioning tau accumulates into so-called paired helical filaments, which ultimately aggregate into neurofibrillary tangles.
Biochemist Marc W. Kirschner and colleagues first isolated tau in 1975 at Princeton University. The protein helps assemble and stabilize microtubules in neurons within the central nervous system. Neurons "require microtubule assembly for the growth and integrity of axons and dendrites and for the transport of molecular cargo between the cell body"—which contains the nucleus—"and distant synapses," Harvard University biochemist and neuroscientist Michael S. Wolfe says.
Tau's interaction with a microtubule is controlled by cycles of phosphorylation and dephosphorylation at specific sites on the tau molecule. When dephosphorylated by a phosphatase, tau binds to the exterior of the microtubule, helping to stabilize it, in part by offsetting the microtubule's negative charge with its own positive charge, notes Michael K. Lee, a neuroscientist who codirects the Center for Neurodegenerative & Neuromuscular Diseases at the University of Minnesota, Twin Cities.
Tau phosphorylation by some kinases can reduce binding of tau to the microtubules, enabling the cell to regulate and remodel the microtubules when needed, Lee adds. In a healthy cell, the microtubule-free tau is rapidly degraded. This natural ebb and flow contributes to neurons' ability to establish new connections and modify old ones, both during development and during learning in mature animals.
Under disease conditions, however, the process becomes unbalanced and tau becomes excessively phosphorylated. Hyperphosphorylated tau permanently dissociates from microtubules and begins to aggregate, says Donna M. Wilcock, a neuroscientist at Duke University Medical Center. Without tau's support, the microtubules' delivery of intracellular cargo can become disrupted.
"We don't really know whether this is an effect of too much kinase activity or too little phosphatase activity," she says. "Something appears to throw the balance between the kinases and the phosphatases off.
"There are also certain phosphorylation sites along tau that seem to be more pathological, so when you get phosphorylation at those sites, that seems to be more closely associated with disease," Wilcock says. She notes that some mutations associated with inherited forms of frontotemporal dementia affect those same sites and make them much easier to phosphorylate.
Among the factors that appear to disrupt proper tau phosphorylation are heavy metals. Heavy metals promote Alzheimer's, according to Ashley I. Bush, a neuroscientist at the University of Melbourne, in Australia, and Rudolph E. Tanzi, a neuroscientist at Harvard and Massachusetts General Hospital, in Charlestown. University of California, Irvine, neurobiologist Frank M. LaFerla has been investigating this connection. He recently reported that chronic copper exposure accelerates tau phosphorylation—possibly through kinase activation—and amyloid-β production in mice (J. Neurochem. 2009, 108, 1550).
Glycosylation, which attaches sugars to tau, also appears to promote hyperphosphorylation and has been detected in Alzheimer's cases, Leonard Petrucelli, a molecular neuroscientist at Mayo Clinic's Jacksonville, Fla., facility, notes in a review about tau (Mol. Neurodegener. 2009, 4, 13).
Another factor that can increase phosphorylation is aging-associated oxidative damage to kinases, Lee notes. Oxidation can also damage tau directly, making it less able to bind to microtubules and more prone to aggregate, he adds.
Furthermore, cells become less able to dispose of damaged tau as they age. As tau aggregates accumulate, they may block the intracellular transport of other biomolecules needed for the proper functioning of neurons, Petrucelli says.
In sum, there are numerous ways for tau to go bad and to disrupt its neuronal host. Complicating the situation further, tau trouble doesn't develop in isolation. In many Alzheimer's patients, for example, tau pathology is likely preceded and indeed exacerbated by amyloid-β pathology, suggests Einar M. Sigurdsson, a neuroscientist at New York University School of Medicine. In turn, tau pathology may make neurons more vulnerable to damage by amyloid-β.
Sigurdsson thinks accumulation of amyloid-β inside or outside neurons damages synapses, which start to retract. Hyperphosphorylated tau then begins to build up inside the afflicted neurons, and the synapses deteriorate further. The cells eventually die, leaving behind insoluble tau tangles that serve as a kind of tombstone for the vanished neurons, Wilcock says.
Amyloid-β and tau appear to share some characteristics. Amyloid-β plaques were for many years thought to be the most harmful form of amyloid-β. But an increasingly popular notion posits that amyloid plaques may represent the end result of the brain's effort to collect and sequester smaller but more toxic aggregates of amyloid-β, such as oligomers. Likewise, some researchers believe that soluble aggregates of tau oligomers are more likely to lead to neuronal dysfunction and death than insoluble helical fragments and tangles (Nature, DOI: 10.1038/nature08890). Harvard neurologist Bradley T. Hyman; University of Minnesota, Twin Cities, neurologist Karen Hsiao Ashe; and others are presently trying to identify which tau aggregates are especially neurotoxic.
Regardless of which form of amyloid-β and tau prove to be most harmful, "tau pathology appears to correlate better with the degree of dementia than amyloid-β pathology," Sigurdsson says. Furthermore, reducing tau production in a mouse model of Alzheimer's disease protected the animals from developing cognitive impairments (Science 2007, 316, 750).
That makes tau an attractive target for treatment. In fact, many drug companies are already pursuing it, as reported in a C&EN blog post ("CENtral Science," The Haystack, April 8).
Meanwhile, researchers in academia are also hunting for tau treatments.
The protein offers multiple targets for these investigators. Neurons produce six "isoforms" of the protein by varying the splicing of the precursor messenger RNA (pre-mRNA) from which tau is translated. Some mutations in the tau gene increase the production of isoforms containing four rather than three microtubule-binding regions, an alteration that is sufficient to cause dementia. With the goal of influencing splicing to adjust the isoform ratio back to normal, Wolfe is investigating compounds that bind to tau's pre-mRNA, including analogs of the anticancer drug mitoxantrone (J. Med. Chem. 2009, 52, 6523).
Another option for treating tauopathies is to increase the activity of phosphatases, which strip phosphate groups from tau, Wilcock says. Phosphatases are normally held in check by protein inhibitors that are bound to these enzymes, she notes, so the best approach might be to somehow remove those inhibitors.
Researchers are also trying to damp down the activity of kinases that phosphorylate tau, Wilcock notes. For example, University of Melbourne pathologist Kevin Barnham inhibited glycogen synthase kinase-3 (GSK-3) with a copper-bis(thiosemicar–bazone) complex that reduced the abundance of hyperphosphorylated tau and amyloid-β oligomers and also reversed cognitive deficits in mice (Proc. Natl. Acad. Sci. USA 2009, 106, 381).
Columbia University pathologist Karen E. Duff and others have experimented with lithium as an inhibitor for GSK-3 in mice. Duff is also studying compounds that disrupt tau aggregation, such as cyanine dyes (J. Med. Chem. 2009, 52, 3539).
Other dyes have been used as well. Claude M. Wischik, a molecular neuropathologist at the University of Aberdeen, in Scotland, has developed a form of methylene blue that prevents tau aggregation and breaks up existing tau aggregates. Wischik, who in 1988 helped determine that tangles are made of tau, cofounded Singapore-based TauRx Therapeutics to further develop the methylene blue compound, dubbed Rember.
Other researchers studying methylene blue include Chad A. Dickey, an assistant professor of molecular medicine at the University of South Florida, in Tampa, who has been investigating its mechanism of action. In work with University of Michigan biological chemist Jason E. Gestwicki, Dickey showed that the dye inhibits a tau-associated chaperone known as heat shock protein 70 (Hsp70) (J. Neurosci. 2009, 29, 12079). The researchers are testing additional chaperone inhibitors, including one based on a dihydropyrimidine scaffold.
Chaperone proteins help fold tau, control its association with microtubules, and regulate its degradation. Tau itself is an unstructured, or "intrinsically disordered," protein, Dickey explains. But interactions with chaperones and other biomolecules can confer specific conformations on tau, he believes. Hyperphosphorylation or truncation of tau under disease conditions might alter these interactions, thereby changing tau's folding behavior and degradation. Furthermore, diseased tau might bind to normal tau and force it to adopt the diseased conformation, spreading tau pathology in a self-sustaining manner reminiscent of prion behavior, Dickey suggests.
Dickey and other researchers—including Petrucelli—are studying how to disrupt this destructive sequence by manipulating chaperones including Hsp70 and Hsp90.
Compounds that help stabilize microtubules when healthy tau isn't readily available for this purpose have shown some promise in animal trials. One example is paclitaxel; another is nicotinamide. Besides enhancing microtubule stability, this form of vitamin B-3 reduces levels of hyperphosphorylated tau and prevents memory loss in mice genetically engineered to develop Alzheimer's, according to UC Irvine's LaFerla (J. Neurosci. 2008, 28, 11500). His Irvine colleague Steven S. Schreiber has begun a clinical trial of the compound.
Another approach enlists the immune system to do battle with pathological tau. Sigurdsson's team is testing two types of immunotherapy in mice genetically engineered to develop tau pathology as they age. In the "active immunization" method, the researchers inject the animals with small, hyperphosphorylated fragments of the tau molecule. The mouse immune system responds to these antigens by producing antibodies against the hyperphosphorylated fragments. These regions are either absent or less prevalent in normal tau, so the antibodies "target the diseased molecules, not the normal ones," Sigurdsson says.
Some of the antibodies move from the mouse's bloodstream into its brain, because inflammation caused by neurodegeneration makes the blood-brain barrier leaky. Diseased neurons take up some of the antibodies, which bind to pathological aggregates of hyperphosphorylated tau inside the cells.
Once inside the neurons, Sigurdsson thinks the antibodies promote the disassembly of tau aggregates by lysosomes. Lysosomes are cellular organelles loaded with enzymes that break down cellular components, such as proteins, that are defective or no longer needed. In an animal suffering from a tauopathy, the lysosomes are unable to keep up with the overwhelming quantity of defective tau aggregates that need to be digested. Sigurdsson believes the tau antibodies counteract this problem by "facilitating the enzymatic degradation that takes place in the lysosomes." The treatment prevents cognitive deterioration in mice (Curr. Alzheimer Res. 2009, 6, 446).
Recent studies have suggested that tau may be secreted and taken up into neighboring neurons, which could serve as a mechanism to spread tau pathology throughout the brain. In addition to degrading tau inside neurons, the antibodies might also promote clearance of extracellular tau, Sigurdsson suggests. This action might prevent uptake of the diseased tau into neighboring cells and limit the spread of pathology.
Sigurdsson and colleagues are also pursuing a "passive immunization" method in which they inject mice with monoclonal antibodies rather than antigens. They are assessing various regions of the diseased tau protein as targets for those antibodies.
Sigurdsson and Wilcock are both studying amyloid therapies in addition to tau treatments, believing that neither the amyloid nor the tau camp has all the answers. "If you can target both of these connected pathologies, you may get a better effect therapeutically than if you just target them individually," Sigurdsson explains.
Last year, Wilcock reported results from a study of mice that develop amyloid deposits, hyperphosphorylated and aggregated tau, and cognitive difficulties. On the basis of that study, Wilcock and her Duke colleague Carol A. Colton were the first to show that immunizing the mice against amyloid-β reduced both amyloid-β and tau pathology, halted neuron loss, and improved the animals' learning and memory capabilities (J. Neurosci. 2009, 29, 7957). Wilcock cautions that the therapy causes limited brain hemorrhaging, the consequences of which are unknown but must be addressed before the treatment can move forward.
The ultimate question, Wilcock says, is "What pathology do you target? If you target tau, is that going to be enough to treat Alzheimer's disease?"
Of course, tau-targeting therapies have relevance for "several tauopathies where you don't have any amyloid-β pathology," Sigurdsson notes. "And even though these are rarer, there is certainly treatment needed for these conditions as well."
The goal of these researchers is to reverse neurodegeneration, but they also believe that slowing the disease process is a worthy aim. "The progression into full-blown dementia takes several years," Sigurdsson notes. "If you can maintain someone as mildly cognitively impaired, that is a great battle won already."