4 December 2009. One hypothesis for the death of motor neurons in amyotrophic lateral sclerosis (ALS) is that the neurons become hyperexcited, firing overtime and ultimately being poisoned by their own activity (Rao and Weiss, 2004). But the buck does not stop with the motor neurons, suggests a paper in the December 2 Journal of Neuroscience. Researchers at the Northwestern University Feinberg School of Medicine in Chicago, Illinois, report that the gung-ho motor neurons may be responding to excess input from spinal interneurons, upstream middle managers that induce motor neurons to fire. Mingchen Jiang is first author on the report of research carried out with his co-principal investigator C. J. Heckman.
Many nerves transmit signals by releasing glutamate into synapses, activating receiving neurons’ NMDA receptors. These receptors open up channels that allow calcium ions to enter the cell. Since motor neurons have particularly low levels of calcium-buffering proteins, they are poorly equipped to maintain calcium homeostasis (von Lewinski et al., 2008). Excess calcium may damage these cells in numerous ways, by altering gene expression, signaling, and cell structure, ultimately leading to cell toxicity and even cell death (reviewed in Trump and Berezesky, 1995).
Heckman and others have shown that motor neurons from a mouse model of ALS are hyperexcitable (Kuo et al., 2003). In the current work, Heckman and colleagues sought to determine what upstream signals lead to this overexcitation. The paper is the first to examine electrophysiology in preparations from adult mice, Jiang said; previous work relied on cell culture and material from neonates. He excised the sacral spinal cord to examine induction of signaling in the ventral root motor neurons (Jiang and Heckman, 2006). The researchers worked with mice carrying the gene for human superoxide dismutase 1 (SOD1) with a G93A mutation, which has been linked to human ALS. They examined the animals at presymptomatic (50 days old), early (90 days old), and late (120 days old) stages of the disease.
The scientists reasoned that motor neurons could be receiving above-normal inputs from two possible sources: peripheral sensory neurons in the short-latency reflex pathway, or spinal interneurons that transmit signals from the neuronal network.
The short-latency reflex system is responsible for quick actions in response to sensory input, such as pulling one’s finger away from a hot stove. One synapse separates the sensory neurons in the dorsal root from the spinal motor neurons sending signals into the ventral root, Jiang said. To examine the role of the short-latency reflex in motor neuron hyperexcitability, he stimulated the dorsal root and recorded activity in the ventral root of mSOD1 and non-transgenic mice in all three age groups. If the mSOD1 motor neurons were hyper-responding to the sensory input, Jiang reasoned, they should show higher rates of firing than the control preparations.
In fact, the mSOD1 ventral roots created lower signals, in terms of voltage, than the controls. However, the researchers suspected this decreased signal might simply be due to reduced overall activity of the sick motor neurons, not a specific response to the sensory input. When they analyzed the voltage output after stimulating the ventral neurons directly, they found that the total motor activity was also reduced. Statistical analysis showed that the decrease in sensory-response firing, compared to the overall reduction in activity, was not significant.
Having eliminated one source of motor neuron excitotoxicity, the researchers turned to interneuron input. They used the drugs strychnine and picrotoxin to induce interneuron signaling in the dissected preparations from 90- and 120-day-old animals. The interneuron and motor neuron activity bursts occurred in unison, confirming to the authors that the interneurons were stimulating the motor neurons.
Between mSOD1 and non-transgenic preparations, there were several differences in the patterns of motor neuron firing. The earliest firing rates were faster in preparations from mSOD1 animals, although over time these rates slowed, likely because the cells’ receptors became desensitized to drug stimulation. The mSOD1 spinal cord preparations were more likely to show this interneuron/motor neuron synchronized firing than non-transgenic tissue, and they had higher proportions of sub-bursts—multiple firings with one action potential starting before the previous has completed. The origin of action potentials was also altered in mSOD1 preparations. While control signals tended to initiate in the S3 segment of the sacral cord, mSOD1 output frequently began in S2, one vertebra higher. From these experiments, the authors concluded that the interneurons are the source of excess input leading to hyperexcitability in motor neurons.
In some animals in the later stages of disease, drugs were not even necessary to induce firing of interneurons and motor neurons in the dissected preparation. These animals generally had muscle tremors reminiscent of those experienced by people with ALS. The researchers used these cases to further investigate the circuit between interneurons and motor neurons.
If the rapidly firing interneurons were the cause of motor neuron hyperactivity, then NMDA receptors should be involved in the process. The researchers used high magnesium levels to block these receptors in their in vitro experiments. This treatment blocked the spontaneous bursting of motor neurons in the older mSOD preparations, suggesting that without interneuron input, the motor neurons do not fire of their own accord. “This is more evidence that this excitatory input is from interneurons,” Jiang said. “Both types of neurons are overactive.”
The work suggests a pathway by which hyperexcited interneurons, by activating motor neuron NMDA receptors, cause hyperexcitability and excess calcium influx, which could ultimately damage or kill the motor neurons. Diminishing the input to those receptors, then, might be a valid therapeutic target, Jiang suggested. There is some experimental evidence in mice to support this idea. Memantine, which blocks NMDA receptors in the brain (Johnson and Kotermanski, 2006) extends survival in ALS mice (Wang and Zhang, 2005). Any such treatment would require precise dosage, Jiang noted, because blocking the receptor completely might prevent the motor neurons from receiving any input, leading to paralysis or even death. (Memantine is currently approved for the treatment of moderate to severe Alzheimer disease in the U.S. and Europe.)
The work moves up one rung on the ladder—from motor neurons to interneurons—but leaves open the question of what causes the interneurons to over-fire in the first place. Could interneurons also be a valid target for ALS therapies? Jiang speculated that riluzole, the only approved drug for the disease and which reduces glutamate levels and calcium influx into motor neurons, might also affect interneurons.—Amber Dance.
Jiang M, Schuster JE, Fu R, Siddique T, Heckman CJ. Progressive changes in synaptic inputs in adult sacral spinal cord of a mouse model of amyotrophic lateral sclerosis. J Neurosci. 2009 Dec 2;29(48):15031-8. Abstract