Counting Open Calcium Channels at an Active Central Synapse
Shankar Ramachandran, Shelagh Rodgriguez, Mariana Potcoava, and Simon Alford
(see pages 2385–2403)
In presynaptic terminals, vesicles are docked at the active zone by the protein complex that mediates fusion. Voltage-gated calcium channels, which provide calcium for triggering release, are tethered to this complex, increasing the efficiency of coupling between depolarization and vesicle release, and thus minimizing the amount of calcium that must enter the terminal. How many channels must open to trigger release at most synapses is unclear. Studies in large calyx synapses, which have numerous active zones, suggest relatively few channels open. Indirect evidence suggests the situation is similar in simpler synapses that have only a single active zone, but the tight apposition of presynaptic and postsynaptic membranes has precluded direct measure of calcium channel opening at these synapses. Ramachandran et al. have overcome this obstacle by acutely dissociating reticulospinal axons from their postsynaptic partners in lamprey spinal cord. This enabled patch-clamp recording from the release face of intact active zones.
Consistent with previous work, currents mediated by N-, P/Q-, R-, and L-type calcium channels were evoked at each active zone. Excluding L-type channels, which do not contribute to evoked synaptic transmission, each active zone contained 11–51 channels. When axons were depolarized to levels reached during action potentials, 1–7 (mean of 4) calcium channels opened. Similarly, quantitative imaging of fluorescent calcium indicators in isolated spinal cords suggested that ∼4 channels open per action potential. These results were further corroborated using lattice light-sheet microscopy to analyze hot spots of calcium elevation evoked by action potentials in intact spinal cord. These experiments revealed that each action potential evokes opening of a different subset of calcium channels distributed across the active zone. Finally, paired recordings from presynaptic terminals and postsynaptic neurons indicated that the number of calcium channels in an active zone is approximately equal to the number of primed vesicles.
These results suggest that each primed vesicle at the active zone of lamprey reticulospinal axons is associated with a single N-, P/Q-, or R-type calcium channel. Each action potential causes up to 7 channels to open, but opening of a single channel can trigger vesicle release. Future work should determine whether the single-channel requirement applies to all channel subtypes and whether single action potentials open multiple subtypes.
Expression of IL-6 (red) is greater in astrocytes (labeled with GFAP, green) from rats expressing mutant GFAP (bottom) than in wild-type rats (top). See Wang et al. for details.
Benefit of Glial Apoptosis in Alexander Disease Model
Liqun Wang, Hassan Bukhari, Linghai Kong, Tracy L. Hagemann, Su-Chun Zhang, et al.
(see pages 2584–2597)
Apoptosis is an important mechanism for removing excess cells during development and damaged cells throughout life. Insufficient trophic support from growth factors, accumulation of misfolded proteins or DNA damage, or external toxins initiate signaling pathways that ultimately activate executioner caspases with numerous targets throughout the cell. Caspase-mediated protein cleavage produces the key hallmarks of apoptosis: cell shrinkage, chromatin condensation, and nuclear fragmentation. Although cell death was long thought to be inevitable after activation of executioner caspases, recent work revealed that cells near death can regain health if apoptosis-inducing toxins are removed. This phenomenon, called anastasis, presents an obstacle for cancer treatment, because it means some cancer cells can survive chemotherapy and even become more aggressive. Nevertheless, the ability to restore health to cells undergoing apoptosis may be beneficial for some neurodegenerative diseases. Wang et al. tested this idea in a Drosophila model of Alexander disease, a disease in which mutations in glial fibrillary acidic protein (GFAP) cause this protein to accumulate, triggering apoptosis in astrocytes and eventually causing neurodegeneration.
Blocking glial apoptosis in flies expressing mutant GFAP led to greater, rather than reduced neuron death. This likely occurred because blocking apoptosis caused glia to assume a senescent phenotype characteristic of aging. Many senescent cells had previously expressed caspases, as shown by a fluorescent reporter, but lacked current caspase expression and instead showed increased expression of the cytokine upd3. Notably, astrocytes in a rat model of Alexander disease also showed hallmarks of senescence, as well as elevated expression of IL-6, a cytokine related to upd3. Similar indicators of senescence and IL-6 upregulation were found in frontal cortical tissue from people with Alexander disease. Importantly, reducing neuronal expression of dome, the upd3 receptor, reduced neuron death in Drosophila expressing mutant GFAP.
These results suggest that increased senescence in astrocytes in Alexander disease leads to increased release of cytokines from these cells, producing an inflammatory state that kills neurons. In contrast, apoptosis of astrocytes may protect neurons by removing astrocytes that would otherwise become senescent. Future work should determine whether blocking senescence is beneficial in models of Alexander disease.
Footnotes
This Week in The Journal was written by Teresa Esch, Ph.D.
