Reducing PERK Activity Improves Memory
Vijendra Sharma, Hadile Ounallah-Saad, Darpan Chakraborty, Mohammad Hleihil, Rapita Sood, et al.
(see pages 648–658)
Secreted and membrane-associated proteins are folded into their mature conformations in the endoplasmic reticulum (ER), where monitors ensure that improperly folded proteins, which may be inactive or toxic, are refolded or targeted for degradation. If new proteins are synthesized too quickly for the folding and monitoring machinery to keep pace, however, misfolded proteins accumulate in the ER. This activates several stress-related molecules, including the kinase PERK. PERK phosphorylates and thus inactivates the translation initiation factor eIF2α, leading to a general reduction in protein synthesis, which allows the folding machinery to catch up. If misfolded proteins continue to accumulate, however, the unfolded protein response becomes persistently active, ultimately causing cell death.
Although the unfolded protein response is essential for maintaining protein and ER integrity, it can also impair learning, which requires ongoing protein synthesis. Indeed, reducing eIF2α phosphorylation enhances memory formation (Buffington et al. 2014 Ann Rev Neurosci 37:17). In addition, Sharma et al. report that knocking down or inhibiting PERK in mouse hippocampal area CA1 improved memory in a fear-conditioning task. This effect was at least partially attributable to increased excitability of hippocampal neurons. After PERK was inhibited, the neuronal spike threshold was lowered (i.e., hyperpolarized), neurons fired more action potentials in response to depolarizing current steps, and the spike afterhyperpolarization was reduced.
The authors also found that PERK mRNA levels were higher in 12-month-old than in 5-month-old mice, and that older mice exhibited poorer memory in the fear conditioning task. At the same time, hippocampal neurons in older mice had a more pronounced spike afterhyperpolarization and spiked less than neurons in younger mice. Importantly, knocking down PERK in the hippocampus of older mice eliminated these differences.
These results suggest that excessive PERK activity contributes to cognitive decline in normal aging. Because a previous study in which PERK was knocked out selectively in CaMKII-expressing (presumably glutamatergic) forebrain neurons found no effects on memory, it is possible the effects in this study were due to loss of PERK in GABAergic neurons. Future work should examine this possibility. In any case, the results indicate that reducing PERK activity might enhance cognitive function in the elderly.
Mutant Huntingtin Impairs Cannabinoid Signaling
Marja D. Sepers, Amy Smith-Dijak, Jeff LeDue, Karolina Kolodziejczyk, Ken Mackie, et al.
(see pages 544–554)
Huntington's disease (HD) is an inherited neurodegenerative disease characterized by production of involuntary, jerky movements, along with other cognitive, psychiatric, and motor symptoms. HD is caused by elongation of a polyglutamine sequence in huntingtin protein. Although huntingtin is expressed ubiquitously, its mutation primarily affects striatal medium spiny neurons (MSNs), which degenerate as HD progresses. The reason for this selective effect is unknown, but altered synaptic transmission and impaired neurotrophin-mediated signaling at corticostriatal synapses are thought to be involved. Defining pathological mechanisms in HD is challenging, however, because expression of mutant huntingtin disrupts multiple cellular functions, including transcription, axonal transport, and synaptic transmission, and it alters the expression and/or function of numerous proteins and signaling molecules. Determining which changes constitute primary pathological events and which are secondary or compensatory is difficult.
Glutamatergic signaling at corticostriatal synapses is regulated partly by retrograde endocannabinoid signaling. Activation of cortical afferents causes MSNs to release the endocannabinoids 2-arachidonoylglycerol (2-AG) and/or anandamide, which activate presynaptic CB1 receptors. This leads to a reduction in glutamate release, producing long-term depression (LTD). But Sepers et al. report that stimulation of cortical afferents failed to reduce release probability or produce LTD in MSNs of YAC128 mice, which express mutant human huntingtin. Notably, LTD was impaired selectively in MSNs that expressed D2 dopamine receptors, which are the first to degenerate in HD, and it was apparent even in young mice. Surprisingly, CB1 expression levels appeared similar in wild-type and YAC128 mice, and CB1 agonists reduced field EPSP amplitudes by similar amounts in the two strains. Furthermore, two processes mediated by 2-AG—depolarization-induced suppression of excitation and LTD induced by a metabotropic glutamate receptor agonist—were similar in wild-type and YAC128 mice, indicating that production of 2-AG and signaling downstream of CB1 receptors were normal. Finally, preventing 2-AG breakdown rescued LTD in YAC128 mice. Preventing breakdown of anandamide was unable to rescue LTD, however.
Together, these data suggest that mutant huntingtin reduces anandamide production by D2-MSNs. The resulting loss of LTD might promote excitotoxicity, leading to synaptic loss and degeneration. Alternatively, the reduction in anandamide might serve to compensate for other impairments in synaptic transmission induced by mutant huntingtin. Future work should determine whether enhancing cannabinoid signaling improves or exacerbates motor impairments in YAC128 mice.
Footnotes
This Week in The Journal was written by Teresa Esch, Ph.D.