19 March 2012. Stressful experiences impair memory and turn down glutamate signaling in the prefrontal cortex (PFC) of rats, according to a study published March 8 in Neuron. Led by Zhen Yan at State University of New York in Buffalo, the study finds that repeatedly stressing juvenile rats accelerates glutamate receptor turnover at the synapse, and stalling turnover undoes the effects of stress.
The relevance of the findings to mental illness is not immediately apparent, but there is hope that understanding the neural consequences of stress may provide clues to the formation of vulnerable brain states (de Kloet et al., 2005). Stressful events trigger release of stress hormones like cortisol, which then activates glucocorticoid receptors. This sets off a host of effects in the brain, resulting in altered behavior and structural changes in neurons (see SRF related news story). Taking a physiological perspective, the new study turns up dampened glutamate signaling specific to the PFC that is reminiscent of findings in schizophrenia (see SRF hypothesis). This suggests that the pattern of stress matters, as Yan’s group has previously linked an acute episode of stress with improved memory and increased glutamate signaling (Yuen et al., 2009).
In search of lost time
Because the brain is particularly sensitive to stress during adolescence, first authors Eunice Yen and Jing Wei studied the juvenile rat equivalent in a repeated stress paradigm that immobilized rats in a restraining tube two hours a day. After seven days of this, the rats were tested in a “temporal order recognition memory” task, which measures the ability to discriminate between objects encountered at different times in the past, and which requires medial PFC (Barker et al., 2007). The rats explored one object, followed by a different object one hour later. When presented with both objects three hours later, control rats spent more time exploring the first, less recent object, but the stressed rats lost this preference; however, they spent the same total time exploring the objects as controls did.
This deficit was attributed to activation of glucocorticoid receptors in the PFC because delivering RU486, a glucocorticoid receptor blocker, into the PFC 40 minutes before each stress session resulted in rats that could discriminate just as well as unstressed controls.
Dampened glutamate signaling
The state of glutamate signaling came under suspicion in these stressed rats, because injecting the PFC with α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-D-aspartate (NMDA) receptor blockers in control, unstressed rats produced a similar temporal order memory deficit. To explore this further, the researchers recorded from pyramidal neurons in PFC slices. In stressed rats, they found diminished glutamate-induced currents: compared to controls, AMPA-receptor currents were 40-60 percent smaller (depending on stimulus intensity), and NMDA-receptor currents were 38-57 percent smaller. Mini-analysis pointed to the post-synaptic side as mediating these decreases. These changes might be specific to PFC, because they were not found in hippocampus or striatum. Like the memory impairments, this dampened glutamate signaling did not transpire in stressed rats also receiving RU486 injections to PFC before each stress session.
To see if these changes stemmed from a decrease in the number of AMPA and NMDA receptors available at the surface for receiving glutamate signals, the researchers compared surface and total levels of receptor subunits using a method that labels the surface pool. In stressed rats, they found about half the level of surface subunits as that found in controls, and for GluR1 (a subunit in AMPA receptors) and NR1 (a subunit in NMDA receptors) specifically, this was also true for the total levels. Total levels of other synaptic and dendritic markers such as MAP2 were not affected by stress, which suggests that these subunit decreases did not merely reflect structural losses of dendrites or synaptic boutons, but rather a specific loss to the receptor composition of the post-synaptic membrane.
Stop the degradation
The researchers next explored whether the stress-induced loss in total levels of GluR1 and NR1 reflected a decrease in their synthesis or an increase in their breakdown. Repeated stress did not change mRNA levels of these subunits, arguing that synthesis was running normally. To measure protein degradation, the researchers measured the levels of GluR1 and NR1 proteins tagged for destruction by ubiquitin, and found nearly twice the amount in stressed rats compared to controls.
To interfere with the destruction of ubiquitinated proteins, the researchers infused MG132, which blocks the action of the cell’s protein-destruction chamber, into the PFC one hour before each stress session. This resulted in fairly normal AMPA- and NMDA-receptor currents in stressed rats, which did not differ significantly from unstressed animals injected with saline or MG132 alone. Surface and total protein levels of the GluR1 and NR1 subunits were also normal. MG132 infusions also blocked the stress-induced deficits in temporal order memory.
Looking one step upstream of degradation, the researchers found that RNA interference-mediated knockdown of the ligases responsible for tagging these subunits with ubiquitin mitigated the effects of repeated stress. Knockdown of Nedd4-1, a ligase that tags GluR1, and Fbx2, a ligase that tags NR1, did not produce the stress-induced increase in ubiquitination of GluR1 and NR1, respectively, nor the decrease in AMPA and NMDA receptor currents, respectively. In fact, the measurements were similar to those found in unstressed controls. Bringing the story full circle, knocking down both ligases prevented the stress-induced deficit in temporal order memory. Both Nedd4-1 and Fbx2 were expressed at higher levels in PFC compared to hippocampus and striatum, which might account for the region-specificity of these findings.
The findings highlight the notion that the ubiquitin protein degradation pathway does more than housekeeping for a cell. A role for the ubiquitination pathway has recently emerged in the remodeling of neural networks (Mabb et al., 2010), and gene expression profiling has fingered abnormalities in this pathway in schizophrenia (Middleton et al., 2002; Bousman et al., 2010). Though future work will have to connect the dots between glucocorticoid receptor activation and Nedd4-1 and Fbx2 activity, the findings identify some molecular consequences of stress that offer new ways to think about its role in mental illness.—Michele Solis.
Yuen EY, Wei J, Liu W, Zhong P, Li X, Yan Z. Repeated stress causes cognitive impairment by suppressing glutamate receptor expression and function in prefrontal cortex. Neuron. 2012 Mar 8;73(5):962-77. Abstract