Rethinking Tau

By Jason Socrates Bardi

Scientists would do well not to listen to the wisdom of famous jazz legend Louis Armstrong, who when asked to define jazz music, supposedly replied, if you have to ask, you'll never know. For scientists, better advice would be if you don't ask you'll never know.

One of the unanswered questions in the field of neuroscience has been how neurons in the brain develop and form connections—sometimes as many as 10,000 apiece. This question is of great interest to scientists because neurons are irreversibly damaged or lost in spinal cord injuries and neurodegenerative diseases like Alzheimer's.

Associate Professor Shelley Halpain, Research Associate Benoit Roger, and several of their colleagues at The Scripps Research Institute report progress in this area in a recent article in the journal Current Biology.

Similar Structures, Different Results

What the researchers were asking in particular was how two different neuronal proteins help maturing neurons send out neurites—the long finger-like processes characteristic of mature neurons that connect them with other neurons.

As these neurites are forming, they must be supported by the cell's cytoskeleton—its actin filaments and microtubules—which means that for the proper formation of the neurites, the microtubules and actin filaments must assemble at the same time. In recent years, scientists have also begun to appreciate that microtubules and actin filaments must interact with each other during this process. Scientists have identified a number of proteins that mediate this interaction, including the microtubule-associated proteins MAP2 and tau.

MAP2 and tau are abundant in neurons where they stabilize and promote the growth of the microtubules—something needed for neurite outgrowth.

Roger and Halpain's experiments showed that MAP2 also binds to actin. The results showed that the domain of MAP2 that binds to actin is the same domain that binds to the microtubules.

In contrast, the similar domain on the tau protein, which also binds microtubules, does not bind to actin the same way. In fact, tau has no actin binding at all.

This was a surprise because MAP2 and tau are so similar structurally—67 percent of the amino acid of the implicated cytoskeleton binding domain sequences are identical, and they both bind to microtubules with almost the same activity. However, Roger and Halpain found that MAP2 is sufficient to trigger neuritic growth but tau is not. And by making a "chimeric" protein of tau with one piece of MAP2 exchanged (the piece that binds to actin), Roger and Halpain showed that this altered tau could now induce neurites.

These differences between MAP2 and tau may cause scientists to rethink the role of tau in neurons and in various neurological disorders. For a long time, scientists have known that tau protein form abnormal aggregates inside cells in Alzheimer's disease, even though the amyloid proteins that form plaques outside of cells were thought to be the actual cause of the disease. Nevertheless, several other diseases are now known to result directly from defective tau—these are called the tauopathies. These rare hereditary dementias, which were just discovered in the last decade, are caused by single amino acid mutations in tau that cause the protein to form fibrous "neurofibrillary" tangles inside neurons.

Interestingly, no such mutations have been found to cause the MAP2 protein to form tangles. Perhaps the ability of MAP2 to interact with actin as well as microtubules may prevent it from forming neurofibrillary tangles. Such information may be used in the future to determine how altering tau's structure could prevent neurodegenerative diseases.

To read the article, "MAP2c, but Not Tau, Binds and Bundles F-Actin via Its Microtubule Binding Domain" by Benoit Roger, Jawdat Al-Bassam, Leif Dehmelt, Ronald A. Milligan, and Shelley Halpain, see the March 9, 2004 issue of Current Biology.


Send comments to:




Neuronal cell in the process of initiating a neurite. Microtubules are shown in green, actin filaments in red. Image courtesy of Leif Dehmelt, Halpain lab.