Single protein appears central to the formation of the long-term memories

Oct 16 2004

Researchers at the National Institute of Child Health and Human Development have discovered that a single protein appears central to the formation of the long-term memories.

The protein is known as mBDNF, which stands for mature brain-derived neurotrophic factor. It appears to chemically alter neurons, boosting their ability to communicate with one another.

"Most of what we accomplish as human beings depends on what we learn," said Duane Alexander, M.D., Director of the National Institute of Child Health and Human Development. "This discovery brings the possibility of studying this protein system in people with disorders of learning and memory and perhaps designing new medications that might help to compensate for these problems."

The study was undertaken by NICHD's Petti Pang, Ph.D, and Bai Lu, Ph.D, along with their colleagues at NICHD, Weill Medical College of Cornell University in New York City, and The Chinese University of Hong Kong.

Researchers recognize two broad categories of memory--short term memory, and long term memory. Short term memory refers to the transient memories that last from minutes to hours. Long term memory refers to the ability to remember things for more than a day--sometimes for many years.

Scientists have suspected that BDNF played a role in memory, but had not known whether it exerted its effect directly, or in combination with other substances. The first clue came in 1996. Then, Dr. Lu and his colleagues reported in Nature that, in a laboratory simulation using rodent brain cells, BDNF fostered changes in the cells indicative of memory. In 1998 Nobel laureate Eric Kandel reported that tissue plasminogen activator was also involved in the formation of long-term memory. Tissue plasminogen activator, or tPA, is best known for its use in dissolving clots in stroke and heart attack patients. Dr. Lu and his colleagues then sought to determine how tPA and BDNF might interact with each other to form long term memory.

A breakthrough came in 2001, when another author of the current paper, Barbara Hempstead, M.D., Ph.D, of Cornell, and her colleagues deciphered the chemical reaction leading to the formation of mBDNF. In that study, Dr. Hempstead and her coworkers reported that the enzyme plasmin chemically converts the early, or precursor, form of BDNF--proBDNF--into mBDNF. Previously, other researchers had determined that tPA converts another substance, plasminogen, into plasmin. (An illustration of the entire chemical sequence by which tPA brings about the formation of mBDNF appears at http://www.nichd.nih.gov/new/releases/conversion_model_image.cfm.)

However, deciphering a chemical reaction in test tubes does not prove that the same reaction occurs naturally in the brain or that the reaction underlies the formation of long-term memory.

"In the Science article, we describe a series of experiments showing that the chemical reaction that generates mBDNF actually takes place in the brain, and that mBDNF is essential to the long-term memory process," said Dr. Pang.

To conduct their experiments, the researchers relied on observations of a laboratory phenomenon thought to mirror the changes that occur in the brain when a long-term memory is formed. Briefly, neurons communicate via a relay system of electrical impulses and specialized molecules called neurotransmitters. A neuron generates an electrical impulse, causing the cell to release its neurotransmitters. The neurotransmitters, in turn, bind to special sites, or receptors, on nearby neurons. The recipient neurons then generate their own electrical impulses and release their own neurotransmitters, triggering the process in still more neurons, and so on.

When a long term memory is made, researchers believe that neurons gain the capacity to transmit a much stronger electrical impulse than they otherwise would, and require much less neurotransmitter.

To simulate memory, the researchers relied on a laboratory test involving slices taken from the brains of mice. The test involves attaching micro-electrodes to brain cells. The micro-electrodes are tiny probes that detect the cells' electrical impulses. The brain slices come from a region of the brain known as the hippocampus, believed to be involved in forming long-term memories. When the hippocampal cells are stimulated with a specific pattern of electric signals, they begin to transmit the stronger electric signals characteristic of neurons involved in memory. A comparatively small burst of electric current simulates a phenomenon thought to parallel short term memory, and is referred to as early long-term potentiation, or E-LTP. A larger burst of current is thought to simulate long-term memory. This phenomenon, which the researchers reported on at length in the Science article, is referred to as late long term potentiation, or L-LTP.

In the first of the experiments, the NICHD researchers treated the mouse hippocampal slices with a compound that prevents new proteins from being made. Protein synthesis is essential for the formation of long-term memory and, before the current study, researchers had searched for years to identify exactly which proteins were needed in the process. As expected, applying current to the slices failed to bring about L-LTP, because mBDNF could not be made.

The researchers then applied mBDNF directly to the hippocampal slices before again applying current. The researchers found that mBDNF completely restored L-LTP even when protein synthesis is inhibited. This demonstrated that mBDNF was essential for memory formation, and mBDNF is the newly-synthesized protein that scientists have been looking for that underlies L-LTP and long-term memory.

In the next series of experiments, the NICHD researchers tested hippocampal slices from the brain tissue of mice that were genetically incapable of producing the chemicals needed to manufacture mBDNF.

In one experiment, the researchers could not induce L-LTP in hippocampal slices from mice incapable of producing tPA. However, the researchers could induce L-LTP in the brain slices of these mice if they first supplied the slices with mBDNF. Similarly, the scientists also found that mBDNF could restore L-LTP in brain slices from mice incapable of producing plasminogen.

Next, the researchers treated plasminogen-deficient brain slices with a form of proBDNF that could not be converted to mBDNF. Again, L-LTP could not be induced in these hippocampal sections. Likewise, adding proBDNF to hippocampal sections deficient in tPA failed to bring about L-LTP. These experiments showed that both plasminogen and tPA are needed to bring about L-LTP.

"Our study has provided a link between these two seemingly unrelated molecule systems in L-LTP," Dr. Lu said.

Next, the researchers analyzed brain tissue from the mice that were deficient in tPA and plasminogen. Brains from tPA-deficient mice contained increased amounts of proBDNF, demonstrating that without tPA, the animals could not make the plasmin needed to produce mBDNF. Similarly, brains of mice deficient in plasminogen also contained increased amounts of proBDNF.

Dr. Lu and his colleagues are now trying to find exactly where and how in the neuron the proBDNF is converted to mBDNF and whether defects in this conversion process could lead to disorders of long-term memory. Dr. Lu added that mBDNF may also play a role in Alzheimer's disease, as some studies have shown that the brains of Alzheimer's patients have reduced levels of mBDNF or plasmin.