23 Jul 2007

The dogma in memory circles states that the conversion of short-term memories to long-term storage requires protein syntheses. Part of the support for this tenet comes from experiments where application of protein synthesis inhibitors produce amnesia in experimental animals. But the story may be a bit more complicated, says a paper out this week in PNAS online. Clinton Canal, Qing Chang, and Paul Gold of the University of Illinois at Urbana-Champaign, show that injection of anisomycin into the amygdala of rats results in a massive release of neurotransmitters, which could explain the memory effects of the widely used protein synthesis inhibitor. While not a dogma-buster (there are other lines of evidence for the requirement of protein synthesis in long-term memory), the study should serve as a cautionary tale about the use of “dirty” pharmacological agents in complex biological systems.

The accepted model of memory formation includes an early phase, independent of protein synthesis, and a later phase, where a permanent memory is written in the protein record of cells. In animals, injection of protein syntheses inhibitors before a training regimen allows the formation of short-term memory, but not long-term memory. Citing many experimental exceptions to this rule, the authors explored alternative actions of one protein synthesis inhibitor, anisomycin, in a test of memory in rats, requiring they learn to avoid a mild electric foot-shock.

When the researchers injected anisomycin into the amygdala, followed by the foot-shock training, inhibition of protein synthesis was apparent by a lack of newly made c-Fos protein, and the rats did not remember the training 48 hours later. The investigators then used microdialysis to measure release of the neurotransmitters norepinephrine (NE), dopamine (DA), and serotonin (5-HT) at the site of the anisomycin infusion and found it to be increased from 12-fold (for norepinephrine) to 50-fold (for DA and 5-HT) just after the time of injection. Between 2 and 3 hours later, the levels of NE and DA had dropped below normal, where they remained for at least 8 hours. By 48 hours, all three neurotransmitters recovered to baseline.

Norepinephrine is well known to modulate memory, so to test whether the spike and decline in NE accounted for the animals’ amnesia, the researchers injected the β-adrenergic receptor antagonist propranolol before anisomycin, or the agonist clenbuterol after, to coincide with the peak increases or decreases in NE, respectively. They found that either of these interventions ameliorated amnesia, even in the presence of a continued block in protein synthesis. Finally, they showed that dosing NE itself before training impaired memory to a similar extent as anisomycin. “These findings suggest that altered release of neurotransmitters may mediate amnesia produced by anisomycin,” the authors conclude. They do not know if the effect on neurotransmitters is special to anisomycin, or a general property of protein synthesis inhibitors, or if the long-term changes in neurotransmitters reflect changes in protein synthesis.

The findings do not rule out de novo protein synthesis playing a role in memory formation. Rather, they inject a note of caution in drawing conclusions based on the use of nonspecific inhibitors. Fortunately, the increasing availability of targeted genetic and biochemical approaches are superseding the use of such pharmacological agents, and offer a wealth of independent and more detailed evidence for the role of protein syntheses in memory consolidation (for a recent example, see ARF news story on Costa-Mattioli et al., 2007).—Pat McCaffrey