By Sean Henahan, Access Excellence

St. Louis, MO (Jan. 22, 1997) Muscle tissue is very particular in its choice of which nerve cells it will form a long-term relationship with, a new study shows.The finding could lead to a better understanding of how memories are preserved in the brain.

"The nerve cells try desperately to talk to the muscle, but the muscle has its fingers in its ears," says Jeff Lichtman, M.D., Ph.D., professor of anatomy and neurobiology at Washington University School of Medicine in St. Louis.

Neurons communicate through axons -- long "wires" that reach out to other neurons or muscle fibers during development and stay connected for life unless they're pruned away. The axons branch at their tips, where they join to target cells at junctions called synapses. Because synapses in the brain are so inaccessible, though there are trillions of them, the more easily reached neuromuscular junction has become the standard synapse for study.

Studies in both mice and humans indicate that signficant pruning of surplus axons from muscle fibers occurs during the first four weeks of life. In this way, each fiber ends up with only one axon. The manner in which elimination occurs has been one of the intractable problems in neuroscience, notes Howard Colman, an M.D./Ph.D., also of the Washington University School of Medicine

In the 1980s, Lichtman and colleagues invented a way to look repeatedly at the same synapses in living mice. "You can literally see neurons that are co-innervating the same target cell," he says. "And you can watch the connections disappear over time."

Nerves activate muscles by releasing packets containing acetylcholine at synapses. Muscle fibers detect this message with membrane proteins called acetylcholine receptors, which cluster directly under nerve terminals. Receptor activation sparks a chain of events that makes the muscle contract.

Studies in Lichtman's lab with a snake toxin that inactivates acetylcholine receptors led to a surprising finding, described in Nature in 1994. Before this time, scientists thought that two axons on the same fiber might act like battering rams, competing with each other for growth factors and space. "But we found that the receptors on the postsynaptic membrane disappeared before the nerve terminal withdrew," Lichtman says. "So the muscle cell is not just a passive witness -- it seems to know which synapses are going to be removed before elimination actually occurs."

These structural studies made Lichtman wonder whether muscle fibers rebuff already dying axons, leaving only healthy ones in place, or whether a muscle fiber decides among healthy axons which one will remain. "We had nothing to prove that the change in the muscle fiber occurs before any change in the nerve terminal," he says. "So we wanted to find out if healthy axons are still trying to communicate with muscles at the time acetylcholine receptors disappear."

Colman and Lichtman addressed this question with electrophysiologic experiments on living trapezius muscle, which they removed from the neck of newborn mice. The muscle remained attached to the two nerve branches that control it in the body.

As axon pruning proceeded, Colman periodically stimulated each nerve branch and recorded the resulting electrical responses of individual muscle fibers. He studied only those fibers that were innervated by axons from each nerve. To better interpret the results, Lichtman invented a way to display the electrical recordings as color-coded 3-D images.

One part of the study measured the relative strengths of the chemical messages from pairs of competing axons -- the number of packets of acetylcholine released. In the first few days of life, the two axons in a pair tended to have similar strengths. But the strengths diverged as the days went by. When the mice were nine days old, for example, the majority of the remaining pairs had one axon that responded to stimulation by sending more than four times as many acetylcholine packets to a fiber as the other axon.

"So the loser does not disappear suddenly- the strength of one input gets stronger while the other gets weaker," Lichtman says. "And we found that the stronger one wins."

Another part of the study focused on the effectiveness of the acetylcholine packets. By ranking responses to each axon, the researchers were able to see that, in many doubly innervated fibers, packets from the losing axon evoked weaker responses in the muscle than those from the stronger one, even though they contained the same amount of acetylcholine. It was as if the muscle was slighting a boring suitor in favor of a more interesting one.

"Our data, combined with other studies from our lab, support the idea that acetylcholine receptors are disappearing under the losing but otherwise healthy axon," Lichtman says.

The work suggests that synapse elimination may be a step-wise process. As a newborn animal learns to use its muscles to move around, an individual fiber will receive different signals from each of the axons that reach it. The signals may differ slightly in strength, or one axon may fire more often than the other.

"So the muscle fiber may decide it likes one axon better than another and may start pulling receptors out from under the axon it doesn't like," Colman says. "That magnifies the small differences between the two axons, which rapidly come to differ more and more. Eventually, the losing axon takes the hint and disappears. So experience leads to axon elimination."

Although this process continues during the first four weeks of a mouse's life, the contest on any one fiber is over in a couple of days. "So that's an interesting analogy to learning and memory, where you have synaptic changes over a short period of time that lead to a permanent alteration in brain circuitry," Lichtman says.

Additional research revealed that synaptic elimination also can occur in adult animals by this same process. Nabekura made electrophysiologic measurements from fibers of injured muscles that were re-establishing connections with nerves. During the transition to single-axon innervation, signals from the losing axon had less and less effect on the muscle.

"We concluded that the process of synapse elimination at the neuromuscular junction of adult animals is similar to that seen during development," Colman says. "So it's possible it also could occur in the adult brain."

No one knows how the brain preserves memories for long periods of time, Lichtman stresses. "But the idea that learning involves loss of connections requires the circuitry for associating any objects you might ever see to be in your brain from birth," he says. "Making a logical connection between a face and a name, for example, would involve preserving and strengthening synaptic connections between certain neurons while eliminating those between others. Our study links the large body of work that shows that experience modifies synaptic strength to a structural mechanism that permanently disconnects some axons and preserves others. If target cells in the brain could modify their connections in the same way, that would be an effective way for experience to permanently change the circuitry of the brain."

The research appeared in Science, 275, 356-361, 1997.