A classical model of how neurons power their chemical messages may need revision. Neurons from the rat hippocampus use three times less energy to propagate an action potential down an axon than was previously believed, according to a new study published in this week's issue of Science -- providing important clues for interpreting brain imaging techniques.

Golgi staining of pyramidal cells in human hippocampus Image: Wikimedia commons, MethoxyRoxy

"Many people will be surprised by this," said neuroscientist Johan Storm of the University of Oslo, who did not participate in the research. "It was thought that action potentials were quite expensive from an energetic point of view, [but] in these axons, they have found that you have only a minimal amount of this shunting or waste of energy."

More than 50 years ago, researchers Alan Lloyd Hodgkin and Andrew Huxley proposed a model for the ionic currents underlying action potentials. The model was based on their observations in squid giant axons, for which they later won the 1963 Nobel Prize in Physiology or Medicine. The Hodgkin-Huxley model predicts a large degree of overlap between when sodium ions rush into the cell -- the "gas pedal" of the action potential -- and when potassium ions rush out of the cell -- the "brake" -- explained neuroscientist Michael Häusser of University College London, who was not involved in the research. This overlap essentially equates to "hitting the gas and the brake at the same time," said Häusser, "and that wastes a lot of energy" -- four times the energy that would be required if there was no such overlap.

Suspicious of this widely accepted "factor of 4," neuroscientist Henrik Alle, then at the Max Planck Institute for Brain Research in Frankfurt, Germany, and his colleagues tested the theory by measuring the currents underlying action potentials in rat hippocampal neurons. After recording the wave form of the action potential (AP) as it traveled down the axon, the researchers removed a small patch of membrane to directly measure the movement of sodium and potassium ions in response to a simulated AP. They found that the flow of the two ions was nearly completely separated in time. The APs appeared to use just 1.3 times the amount of energy that would be needed if there were no overlap at all -- a "quite high efficiency," Alle said.

"Based on this study, it is likely that energy consumption connected with action potential propagation in the mammalian brain is much less than the energy required by other steps in neural signaling, and much less than previously believed," said Storm.

The results of this study have important implications for the interpretation of brain imaging techniques such as functional MRI (fMRI), said Häusser. "There has been a lot of controversy in the field about what fMRI signals represent," he said. Although it was generally understood that the fMRI readouts were related to energy consumption, researchers didn't know which neural processes used that energy. These findings suggest that when specific brain regions appear activated in an fMRI scan, that activity comes not from the action potential, but from other steps in neural signaling, such as synaptic transmission. "Understanding that is crucial for interpreting fMRI signaling correctly," Häusser added.

The question now is whether this energy efficiency is typical of other neurons in the brain, said Alle, now at the Institute of Neurophysiology at the Charité-Universitätsmedizin Berlin. "We have shown this at one axon, [but] to make sure that this is not an exception, our community has to look at other axons," said Alle

In addition, Häusser said, "it's going to be really important to directly measure the energy budget of neurons." This study simply predicted the amount of ATP consumption based on measurements of the ionic currents in the axon, but once ATP consumption is measured directly, "then we can really put hard numbers on the energy budget for different [neural] processes," he said.


Related stories:

Single neuron power
[15th October 2008]

Mental music
[28th March 2008]

Bit by Bit, the Structure of the Potassium Ion Channel Emerges
[22nd July 2002]