When an animal is standing still or moving slowly, it's relatively easy to figure out what a specific muscle might do. Once the animal starts moving more quickly, the question gets a lot hairier – the body and the legs are swinging around; sometimes the legs hit the ground and sometimes they're in the air; and bones and joints have their own springiness. If we could answer the question, though, we'd be one step closer to understanding how animals run, jump, fly or swim as well as they do – and perhaps a step closer to building robots that could do the same thing.
Some groups have turned to computer models to examine how muscles and body interact during fast movements. But Simon Sponberg, Andrew Spence, Thomas Libby, Chris Mullens and Robert Full at the University of California, Berkeley, developed an alternative method to examine this interaction in running cockroaches.
They implanted recording and stimulating electrodes in a cockroach's middle leg, in the ventral femoral extensor muscle – a muscle that produces force to extend the leg and raise the body up or, conversely, to slow the body from falling down. Indeed, previous work, in which the muscle was excised from the body, had suggested that its primary function is as a brake, absorbing energy by slowing the body down.
Sponberg and his colleagues wanted to look at how the muscle functions in the animal during running. So, rather than pulling the muscle out, they developed a system that recorded the normal muscle action potentials from the nervous system, then used the stimulating electrode to rewrite that neural command – all in real time as the cockroach ran in an open arena. They couldn't erase the command entirely, but they could extend it for longer than normal or start it earlier. To see the effects of changing the neural command, they had the cockroaches carry acceleration-sensing backpacks and used high-speed video to accurately and rapidly track leg and body motions.
What they found was almost completely different from what they'd predicted based on older work. If the muscle normally operates like a brake, they expected that extending the normal command would make it brake even harder. Instead, it did the opposite, and started producing energy, lifting the body higher. Interestingly, the changes persisted into the following stride – for which they hadn't changed the neural command – and tilted the cockroaches up.
When the researchers started the neural command earlier, rather than extending it later, they found yet another effect – the animal started to turn. The earlier the command started, the stronger the turn.
In a companion paper, Sponberg and colleagues considered how the muscle changed its function during running. They found that the effect was due to a purely mechanical positive feedback between the muscle force and the leg position during the stance phase – no sensory inputs were needed. With more spikes, the muscle produced more force, extending further, which allowed it to produce even more force and extend even further.
Studying these neuromechanical feedback loops is challenging, but they hold the key to understanding the amazing flexibility of animal locomotion.