Role of Calretinin at High-Frequency Endbulb of Held Synapses
Chuangeng Zhang, Meijian Wang, Shengyin Lin, and Ruili Xie
(see pages 2729–2742)
Auditory nerve fibers are categorized by their spontaneous spike rate and auditory threshold, and they differ in their expression of the calcium binding protein calretinin (CR). Whereas low-threshold/high-spontaneous-rate fibers express calretinin in their presynaptic structures (calyces of Held) in the cochlear nucleus, fibers with medium-high threshold and low-medium spontaneous rates do not. Zhang, Wang, et al. report that calretinin helps fibers maintain high-frequency transmission.
To understand how calretinin affects synaptic transmission, the authors analyzed spontaneous and evoked EPSCs in postsynaptic bushy cells. The frequency of spontaneous EPSCs was similar in cells receiving input predominantly from calretinin-positive fibers (CR synapses) and those receiving input predominantly from calretinin-negative fibers (non-CR synapses). The average amplitude of EPSCs evoked by single presynaptic spikes was also similar across cells. In contrast, the initial release probability was greater and the size of the readily releasable pool of vesicles was smaller at CR synapses than non-CR synapses. Although one would expect this to result in more rapid vesicle depletion at CR synapses, evoked EPSC amplitude decreased more slowly during high-frequency stimulation at CR synapses than at non-CR synapses. This was attributable to faster vesicle replenishment in CR synapses, which, in turn, was partly attributable to differences in the expression of vesicular glutamate transporters: whereas CR synapses only expressed VGluT1, non-CR synapses expressed both VGluT1 and VGluT2; VGluT2 refills vesicles with glutamate more slowly than VGluT1. In contrast, total release—which includes both synchronous and asynchronous release—decreased more rapidly during high-frequency stimulation at CR synapses than at non-CR synapses. In fact, asynchronous release was lower at CR synapses throughout stimulation, likely because calretinin sequestered calcium required for asynchronous release. Indeed, an exogenous calcium buffer reduced synaptic depression and asynchronous release at non-CR synapses, but not at CR synapses.
These results suggest that calretinin supports high-frequency transmission at calyx-of-Held synapses between low-threshold/high-spontaneous rate fibers and bushy cells. Notably, calretinin binds calcium more efficiently when calcium levels are high than when they are low. Consequently, calretinin affects synaptic transmission predominantly during high-frequency spiking, when calcium influx is high. Enhanced calcium buffering reduces asynchronous vesicle release and promotes synchronous release, which, along with rapid refilling of synaptic vesicles, allows CR synapses to maintain firing at high rates.
Caffeine greatly decreased the activity of iMSNs in PNKD mice (red, bottom), but produced little change, on average, in iMSNs of wild-type mice (black, bottom), dMSNs of PNKD mice (green, top), and dMSNs of wild-type mice (black, top). See Nelson et al. for details.
Reduced Firing in Indirect Pathway Neurons in Rare Dyskinesia
Alexandra B. Nelson, Allison E. Girasole, Hsien-Yang Lee, Louis J. Ptáček, and Anatol C. Kreitzer
(see pages 2835–2848)
The performance of smooth voluntary movements depends on balanced activity in two populations of neurons in the striatum: direct-pathway medium spiny neurons (dMSNs) and indirect-pathway MSNs (iMSNs). Disruption of MSN activity underlies many movement disorders. For example, Huntington's disease involves degeneration of MSNs, and Parkinson's disease results from degeneration of dopaminergic inputs to these cells. Even in the absence of degeneration, however, imbalances in MSN activity can cause movement disorders, including involuntary, erratic movements called dyskinesia. For example, l-DOPA-induced dyskinesia is associated with increased activity in a subset of dMSNs and reduced activity in iMSNs. Nelson et al. report that iMSN activity is selectively affected in a mouse model of paroxysmal nonkinesigenic dyskinesia (PNKD), a rare condition in which dyskinesia occurs after consumption of caffeine or alcohol.
Under baseline conditions, the firing rates of both dMSNs and iMSNs were similar in wild-type and PNKD mice. Intraperitoneal injection of caffeine in wild-type mice increased the average spike rate of dMSNs, but did not affect iMSNs. In PNKD mice, in contrast, caffeine had no effect on the average spike rate of dMSNs, but greatly decreased the average spike rate of iMSNs. Notably, decreases in iMSN spiking slightly preceded the onset of dyskinesia, and rates returned to baseline when dyskinetic movements stopped. The effects of caffeine were mimicked by chemogenetic inhibition of iMSNs in PNKD, but not wild-type mice, and the effects were blunted by chemogenetic activation of iMSNs. Finally, the caffeine-induced reduction of iMSN activity was attributable to a decrease in the probability of glutamate release onto iMSNs, which was likely mediated by endocannabinoid-dependent long-term depression. Consistent with this, pretreating PNKD mice with a cannabinoid receptor antagonist reduced the severity of caffeine-induced dyskinesia, whereas potentiating cannabinoid signaling lowered the amount of caffeine needed to induce dyskinesia in PNKD mice.
These results demonstrate that dyskinesia can result from decreased iMSN activity without changes in dMSN activity. PNKD is caused by mutations in a synaptic protein (named pnkd) with unknown function. Determining how pnkd mutations enable caffeine, an antagonist of Gs-coupled adenosine A2a receptors, to promote endocannabinoid-dependent depression of input onto iMSNs is a rich area for future research.
Footnotes
This Week in The Journal was written by Teresa Esch, Ph.D.
