Proteins Run Amok

by Guy McKhann, M.D.

February 21, 2012

There are a group of progressive diseases of the brain in which specific populations of nerve cells (neurons) degenerate. These so-called neurodegenerative diseases can be very specific, depending on which populations of neurons are affected. If the retina is affected, you might go blind; if it's the cerebellum, you may develop ataxia, or clumsiness; if nerves in the spinal cord, ALS (also called Lou Gehrig's disease). And if neurons in the cortex are affected, you may develop Alzheimer's disease.

For a given disease, do all the nerve cells that will ultimately be affected get sick at once? Or does the disease start in a specific area and then spread to other areas? Here we have some observations, and at least partial answers.

In Alzheimer’s disease, the degeneration starts in a very specific part of the brain, the entorhinal cortex, a part of the temporal lobe. This structure is part of one of the memory systems of the brain, which also involves the hippocampus and is particularly involved in episodic memory—the memory for explicit events such as a conversation, an answer to a question, or a recent trip. Failure to retain this type of recently acquired information is one of the hallmarks of early Alzheimer’s.

Over time, the disease advances clinically to involve other aspects of memory and other cognitive functions. On pathological examination, it is noted that the disease “spreads,” and involves other parts of the brain.

The basic pathology of Alzheimer’s disease involves the two proteins:
1. Amyloid protein, involving a fragment of the protein called A-Beta 42. This protein is seen in the form of “plaques,” as originally described by Alois Alzheimer.
2. Tau, also originally identified by Alzheimer, appears within nerve cells as “tangles.”

There has been considerable controversy about which protein is the primary culprit in the disease process. For a variety of reasons, which I have outlined in previous columns, the emphasis has been on amyloid—millions of dollars have been spent trying to develop therapies aimed at modifying A-Beta or getting it out of the brain. These therapeutic attempts have not succeeded.

In recent years, the conversation about therapeutic considerations has shifted to include tau. Further, if one correlates the presence of plaques or tangles with the degree of cognitive impairment, the correlation is strongest with the presence of tau-containing tangles.

Occasionally insights into a disease are gained when investigators ask a different question. In this case, the question shifted from What is the protein doing? to How is the disease spreading? Two research groups (one at Harvard led by Brad Hyman; the other at Columbia by Karen Duff and Scott Small), almost simultaneously, came up with an answer outlined by Gina Kolata in The New York Times.

These investigators, using a mouse model of Alzheimer’s disease, engineered a mouse so that the gene for human tau protein was active only in the nerve cells of its entorhinal cortex. They waited for several weeks and looked to see where the human protein was located. What they found was that the protein was not only where they expected (in the neurons of the entorhinal cortex), but also in other neurons connected to the targeted neurons. Further, over time the neurons of the entorhinal cortex had less human tau, presumably because they began to die. The other neurons in other areas, which I will call the recipient neurons, did not have the gene to make human tau, so the human tau could only have come by some form of neuron-to-neuron transfer. At the times when the entorhinal neurons were dying, the recipient neurons continued to accumulate human tau.

The results suggest that the tau protein is carried from one neuron to another through established connections, perhaps going across the synapse, the usual signaling route for neurons. Thus the spread is not random, but through existing pathways, many of which are part of memory mechanisms.

The observations also provide an approach to therapy. This spread is a slow process, taking place over months to years in the mouse and, presumably, in a similar timeframe in humans. The shift in therapeutic emphasis could be to clearly understand this interneuronal transfer of tau, and to slow it down even further. The result would be slowing the progression of the disease. Just how that might be done, I leave to Drs. Hyman and Duff.