A New Model for Anxiety…and More


by Rob Mitchum

An advantage and disadvantage of hypothesis-free studies looking for genes associated with various traits or diseases is that they often point to genetic candidates that don’t make immediate sense. One example of this occurrence was the 2005 discovery of an association between the gene Glo1 and anxiety-like behaviors in mice. Previously, scientists knew Glo1’s protein product, glyoxylase 1, primarily as a enzyme involved in in glycolysis — the cellular digestion of glucose. Nobody had considered that glyoxylase 1 played a role in brain function, much less behavior, leading some to question the validity of the genetic association.

“When people discover a gene, they’re always most comfortable when they discover something they already knew,” said Abraham Palmer, assistant professor of human genetics at the University of Chicago Medicine. “The alarming thing here was there was a discovery of something that nobody knew, and therefore it seemed less likely to actually be correct.”

But Palmer’s laboratory continued chasing down the Glo1/anxiety connection, and their experiments paid off in the discovery of an entirely new mechanism for anxiety disorders. Their study, published today in the Journal of Clinical Investigation, also describes a previously unrecognized inhibitory neural factor, offers a promising new target for the treatment of anxiety disorders and other psychiatric symptoms, and suggests an intriguing connection between metabolism and neurobiology.

“What’s neat is that we started with exploratory, open-ended genetic studies in mice, and we’ve now gotten into some fundamental new physiology that nobody had appreciated or put together before,” Palmer said. “Now we’re starting to reap some of the fruit from those types of genetic studies to enrich our understanding of more classical aspects of biology.”

Lead author Margaret Distler, an MD/PhD student in the Pritzker School of Medicine, started her examination of the Glo1/anxiety link by testing it in a new way. In a 2009 paper, Palmer’s group hypothesized that mice with more copies of the Glo1 gene — a genetic phenomenon known as copy number variants — were more likely to show anxious behaviors. Distler tested this theory by inserting two, eight, or ten copies of the gene into mouse lines, and measuring their anxiety with various laboratory tests, including the open field test and the light-dark box test. As predicted, more Glo1 copies equated to more anxiety-like behavior, lending more evidence to the link.

“Animals transgenic for Glo1 had different levels of anxiety-like behavior, and more copies made them more anxious,” Palmer said. “We showed that Glo1 was causally related to anxiety-like behavior, rather than merely correlated.”

The next step was to figure out how Glo1 accomplished its unexpected influence. In glycolysis, glyoxylase 1’s job is to metabolize and reduce levels of a byproduct called methylglyoxal, or MG for short. So Distler tried an experiment so simple it almost obvious: if increasing Glo1 increases anxiety, would increasing MG levels alleviate it? After injecting mice with MG, the mice were less anxious in the behavioral experiments, suggesting that the metabolic byproduct does in fact play a role in anxiety — and a relatively fast role, at that.

“Methylglyoxal changed behavior within 10 minutes of administration, which means it’s a rapid onset. It’s not changing gene expression, and it’s not having long-term downstream effects,” Distler said. “That was our first breakthrough.”

A major clue to what MG was doing in the brain was provided by trying higher doses of the factor, which caused mice to become hypothermic and sedated — so much so that Distler couldn’t test their anxiety. Those symptoms are often seen with drugs, such as barbiturates and benzodiazepines, that activate receptors for the neurotransmitter GABA and are sometimes prescribed to treat anxiety. A collaboration with electrophysiologist Leigh Plant found that MG does in fact activate the GABA-A receptor, making it only the second endogenous factor (after GABA itself) discovered to have that effect in the brain.

Those findings led to a model for the role MG plays in the brain. In the synapse, higher levels of the neurotransmitter GABA would be expected to activate the GABA receptors on the receiving neuron. But away from the synapse,  MG might out-weight the influence of GABA, making it the most important inhibitory factor at these locations. Here too, the connection to glycolysis fits into the model, as more active neurons would ramp up glycolysis, producing more MG, which can then act as a negative feedback “brake” on neuronal activity.

“It’s well known that if neurons become too excited, intracellular calcium levels rise and the cell will die spontaneously,” Palmer said. “There are known mechanisms by which excessive metabolic activity can be toxic to neurons, so you would want to have a brake that would quiet the system down.”

That neurobiological brake might also provide a new target for the treatment of anxiety and other psychiatric conditions, such as epilepsy and sleep disorders, that are commonly treated with GABA agonist drugs. To test this premise, Palmer and Distler contacted John Termini of the Beckman Research Institute of the City of Hope, who had previously created a Glo1 inhibitor for an entirely different reason: the treatment of cancers where tumor cells over-express the Glo1 gene. Termini provided samples of the Glo1 inhibitor to the researchers, who tested its effects on mouse anxiety. As with MG treatment, inhibiting Glo1 successfully lowered anxiety-like behaviors, suggesting that a similar approach might be useful clinically for treating psychiatric conditions while minimizing the side effects of existing drugs.

“The GABA-A receptor agents already out there have a lot of side effects, such as sedation and hypothermia, as well as a high abuse liability,” Distler said. “It’s possible that taking a Glo1 inhibitor will increase only MG levels to a certain maximum. You could have the potential for more specificity, given that you’re activating a system that’s already in place, not just dumping methylglyoxal or some other GABA-A receptor agent throughout the brain.”

“It’s a different way of hitting these GABA-A receptors,” Palmer said. “We have yet to determine if that’s a better way of doing it, but it’s certainly different, and it gives us a unique angle of attack on this system and potential advantages that we have yet to evaluate.”

But the potential connection between metabolic activity and neuronal inhibition also presents a non-pharmacological possibility for affecting behavior. Activities that increase cellular metabolism, such as exercise or eating, could alter a person’s brain levels of MG, and thereby influence their behavior. Conversely, metabolic diseases such as diabetes may disrupt the factor, with previously unrecognized psychiatric consequences.

“This could also be an interesting system where a person or animal’s metabolic load reflects their neuronal inhibitory tone and behavior,” Distler said. “Maybe it’s not a drug, but it’s an interesting therapeutic approach or concept for how could we affect behavior by changing our metabolic state.”


Distler, M., Plant, L., Sokoloff, G., Hawk, A., Aneas, I., Wuenschell, G., Termini, J., Meredith, S., Nobrega, M., & Palmer, A. (2012). Glyoxalase 1 increases anxiety by reducing GABAA receptor agonist methylglyoxal Journal of Clinical Investigation DOI: 10.1172/JCI61319

About Rob Mitchum (525 Articles)
Rob Mitchum is communications manager at the Computation Institute, a joint initiative between The University of Chicago and Argonne National Laboratory.
%d bloggers like this: