In vivo synaptic recovery following optogenetic hyperstimulation - PubMed (original) (raw)
In vivo synaptic recovery following optogenetic hyperstimulation
Maike Kittelmann et al. Proc Natl Acad Sci U S A. 2013.
Abstract
Local recycling of synaptic vesicles (SVs) allows neurons to sustain transmitter release. Extreme activity (e.g., during seizure) may exhaust synaptic transmission and, in vitro, induces bulk endocytosis to recover SV membrane and proteins; how this occurs in animals is unknown. Following optogenetic hyperstimulation of Caenorhabditis elegans motoneurons, we analyzed synaptic recovery by time-resolved behavioral, electrophysiological, and ultrastructural assays. Recovery of docked SVs and of evoked-release amplitudes (indicating readily-releasable pool refilling) occurred within ∼8-20 s (τ = 9.2 s and τ = 11.9 s), whereas locomotion recovered only after ∼60 s (τ = 20 s). During ∼11-s stimulation, 50- to 200-nm noncoated vesicles ("100nm vesicles") formed, which disappeared ∼8 s poststimulation, likely representing endocytic intermediates from which SVs may regenerate. In endophilin, synaptojanin, and dynamin mutants, affecting endocytosis and vesicle scission, resolving 100nm vesicles was delayed (>20 s). In dynamin mutants, 100nm vesicles were abundant and persistent, sometimes continuous with the plasma membrane; incomplete budding of smaller vesicles from 100nm vesicles further implicates dynamin in regenerating SVs from bulk-endocytosed vesicles. Synaptic recovery after exhaustive activity is slow, and different time scales of recovery at ultrastructural, physiological, and behavioral levels indicate multiple contributing processes. Similar processes may jointly account for slow recovery from acute seizures also in higher animals.
Keywords: channelrhodopsin; chemical synapse; electron microscopy; synaptic vesicle recycling.
Conflict of interest statement
The authors declare no conflict of interest.
Figures
Fig. 1.
Recovery of synaptic transmission after optical hyperstimulation, using ChR2(H134R). (A) Velocity of C. elegans expressing ChR2 in cholinergic neurons (transgene zxIs6) is decreased during a 30-s photostimulation and recovers 30–60 s later. Displayed is the mean normalized velocity during periods (time bins) before, throughout, and after 30-s illumination (473 nm; 22.6 mW/mm2) for animals raised in the absence (red squares; n = 14) and presence (blue circles; n = 15) of ATR; statistically significant differences refer to the velocity before illumination (*P < 0.05; **P < 0.01; ***P < 0.001). Also shown is a one-phasic fit of the data, revealing a time constant of τ = 20.0 s for speed recovery (dashed red line). (B) Exemplary 100-s m(e)PSC trace, measured in patch-clamped muscle cells of zxIs6 animals before, during, and after 30-s photostimulation (the trace is in the same time scale as the behavioral experiment in A). Peak currents are typically 1,200–1,600 pA (see Inset in D for group data). Inset in the gray box shows expanded regions of the trace, before (red), during (blue and cyan), and after (green) the light stimulus, color coded by bars/boxes. (C) Rates of mPSC events, reflecting single SVs fusing, were determined for consecutive 1-s intervals before, during, and after the photostimulus (means obtained from n = 5–6 traces, analogous to the one in B). Red dashed line shows a monophasic decay of the ePSCs during the stimulus (τ = 10.11 s). (D) Recovery of photo-ePSCs (Upper) or photocurrents (Lower) evoked by ChR2 expressed in cholinergic neurons (Upper) or in muscles (Lower), recorded from postsynaptic, voltage-clamped body wall muscle cells. After an initial 30-s photostimulus (blue bar), cells were left in the dark for various recovery periods (6, 10, 20, 30, 60 s) before a second, 10-ms stimulus was applied (n = 5–7; blue tick mark). (Inset) Group data and statistical analysis. (E and F) Synapses regained full release capability after about 20 s, whereas ChR2 showed maximal currents already after 6 s. Currents were first normalized to the peak (E), and then muscle currents postsynaptically evoked by (ChR2 in) cholinergic neurons were corrected for the observed recovery of photocurrents of ChR2 directly expressed in muscle (F). A one-phasic fit of the data shows a time constant for ePSC recovery of 11.94 s (dashed black line in F). In A_–_C, illumination is indicated by blue shading. Displayed are means ± SEM throughout the figure, except in B and D.
Fig. 2.
HPF-EM coupled with photostimulation of cholinergic neurons in vivo. (A) Experimental setup. A laser was directed at the sample holder with the HPF platelet that contains C. elegans in a paste of bacteria. The laser was shutter-controlled. (B) Sequence of the experiment, interval of photostimulation, and time points of HPF are indicated, relative to the end of the photostimulus (6, 8, 10, 15, 20, 60 s). (C, Left) Typical ultrastructure of a wild-type cholinergic NMJ, obtained after HPF of intact C. elegans. Indicated are the proteinaceous DP material at the AZ, the postsynaptic muscle arm process, SVs, and a DCV. Also visible are endosomal structures (endo) and a mitochondrion (mito). Endosomes, like smooth ER, have an irregular (nonspherical) shape in thin sections and a core that is less electron-dense than the cores of DCVs and SVs. Furthermore, regions in which SVs, in direct contact with the PM and at different distances to the DP, were counted and classified into categories (also see D) and are indicated by different color shading: categories I (yellow), II (red), III (blue), IV (green), and V (purple). (C, Right) Enlarged view of the boxed region. (Scale bars: 100 nm.) (D) Statistical analysis of the occurrence of SVs in single thin sections through the DP, in the different groups classified according to the pictogram, at different time points following a 30-s photostimulus. Displayed are means ± SEM; n = 15–46 synapses were analyzed (also see related
Fig. S2_A_
). (E) All docked SVs and SVs tethered to the DP were counted and are shown as mean (± SEM), number of SV per section for untreated and photostimulated synapses following indicated recovery periods. Docked SVs recover with a time constant of 9.18 s (see related
Fig. S2_B_
). In D and E, significance was determined by one-way ANOVA with Dunnett’s post hoc test (comparison with untreated). (F) Recovery of ePSCs mediated by postsynaptic nAChRs was probed by applying cholinergic agonists [ACh, levamisole (Lev)] from a pipette directly to the NMJ, after the indicated ISIs following the presynaptic photostimulus. Displayed are means ± SEM and the percentage of agonist-evoked currents compared with currents in nonphotostimulated animals (n = 7–9); nonsignificance was determined by Student's t test.
Fig. 3.
Three-dimensional reconstruction of serial sections of photostimulated synapses shows formation of 100nm vesicles. (A) Wild-type synapse expressing ChR2, photostimulated, and frozen after 6-s recovery. (Scale bar: 100 nm.) (B) Wild-type synapse, untreated. (A_–_C, Left) Single sections through the DP (indicated by an arrowhead); 100nm vesicles are labeled by an asterisk. (A and B, Center and Right) Three-dimensional reconstructions, in two perpendicular orientations. (A and B, Center) View along the axon. (A and B, Right) Lateral view on the axon, with the DP in front. Light gray, transparent: axonal membrane; dark gray: large vesicular structures; blue: SVs; black: DCVs; red: DP. (C) Morphology of a photostimulated cholinergic NMJ synapse (arrowhead) of a wild-type zxIs6 animal raised without ATR. (D) Statistical analysis of diameters of SVs and endosomes in untreated wild-type synapses and of 100nm vesicles observed following 30-s photostimulation and 6-s recovery. The indicated number of organelles was analyzed. Displayed are means ± SEM. ***P < 0.001; *P < 0.05 (Kruskal–Wallis test with Dunn’s post hoc test). (E) Comparison of morphologies of hyperstimulation-induced 100nm vesicle (asterisk) (Right) and synaptic endosome-like structures present in wild-type synapses (black arrow) (Left) (this is a part of the image shown in Fig. 2_C_). SV marked by an open arrowhead and DCV by a black arrowhead. See also
Fig. S3
for analyses of SV and DCV diameters, reconstruction of a neuron–neuron synapse, and more examples of endosomes vs. 100nm vesicles.
Fig. 5.
Aberrant endocytic structures found in synaptic terminals of endophilin unc-57, synaptojanin unc-26, and dynamin dyn-1 mutants. Synaptic sections and reconstructions of unc-57(e406) (A), unc-26(s1710) (B), and dyn-1(ky51) (C) mutant synapses were prepared from animals treated as indicated. Filled arrowhead indicates the DP, and asterisks indicate 100nm vesicles. In dyn-1 mutants, white arrow points to membrane continuity of 100nm vesicle with the PM. Vesicles trees indicate that the disassembly of these large structures into smaller vesicles cannot be completed (open arrowheads). Synaptic vesicles (blue) and DCVs (black) are absent from the vicinity of the DP of dyn-1 mutants. (D) Statistical analysis of the diameters and occurrence of the indicated number of endosomes and light-induced, bulk-endocytosed 100nm vesicles, in different genetic backgrounds and conditions, as indicated; 100nm vesicles in unc-26 mutants were significantly larger than in the wild-type or in the other mutants tested. ***P < 0.001 (Kruskal–Wallis test with Dunn’s post hoc test). (Scale bar: 100 nm.)
Fig. 4.
Spatiotemporal development of 100nm vesicles during and after a photostimulus analyzed by HPF-EM. (A) Experimental sequence and different HPF time points using two ChR2 variants. ChR2(C128S) (transgene zxIs22) remains open for minutes after a 1-s stimulus; thus, synapses remain stimulated until freezing and events during stimulation are observed. For late time points, we used ChR2(H134R). (B) Representative single sections obtained at indicated times of continuous stimulation via ChR2(C128S); 100nm vesicles denoted by asterisks. (C) Decomposition of 100nm vesicles requires 6–8 s. Representative single sections frozen at indicated times after the end of a 30-s photostimulus, in zxIs6 animals. Large 100nm vesicles (asterisks) close to the DP (black arrowheads) are indicated. Percentage of AZs observed where 100nm vesicles were found is indicated, as well as number of synapses analyzed. (Scale bars: B and C, 100 nm.) (D) Sizes and occurrence of 100nm vesicles observed in synapses following photostimulation and indicated recovery times. Displayed are means ± SEM. *P < 0.05 (Kruskal–Wallis test with Dunn’s post hoc test). (E and F) The 100nm vesicles are not large transmitter content “compound” vesicles. (E) mPSC amplitudes, measured before (gray shade), during (blue shade), and following (green shade) 30-s photostimulation and averaged for 1-s time bins. Displayed are means ± SEM. (F) Analysis of the amplitudes of measured SV fusion events, as a correlate for SV size and/or transmitter content. For each 5-s window before (gray), during (blue), or after (green) the photostimulation, the number of single mPSC events, falling into bins of amplitudes as indicated on the x axis, was summed up for n = 7 preparations. For better visibility, only 5 of the 12 time periods analyzed are shown here. See also
Fig. S4
for a representation of the data in F, for all time periods, and normalized to peak event number.
Fig. 6.
Decomposition of 100nm vesicles is delayed in animals expressing mutant endophilin, synaptojanin, and dynamin. (A_–_C) Representative single sections frozen at indicated times after a 30-s photostimulus, in endophilin unc-57; zxIs6 (A), synaptojanin unc-26; zxIs6 (B), and dynamin dyn-1(ky51); zxIs6 (C) mutants, incubated at permissive (15 °C) and nonpermissive (25 °C) temperatures. Large vesicles (asterisks) close to the DP (arrowheads) are indicated. Red bars indicate time after which 100nm vesicles are observed in less than 50% of synapses analyzed. Percentage observation of 100nm vesicles is indicated for each condition, as well as the number of synapses analyzed. Similar to wild-type, dyn-1 mutants show formation of 100nm vesicles (asterisk) in synaptic terminals of motoneurons at the permissive temperature after stimulation. When dynamin is destabilized at 25 °C, several 100nm vesicles are found even after 15 s following stimulation. (D) Temporal sequence of decomposition of the 100nm vesicles in photostimulated wild-type, unc-57, unc-26, and dyn-1(ky51) synapses (the latter at nonpermissive temperature), based on data in A_–_C. Also see
Fig. S5
. (E) The number of large endocytic vesicles was counted for each terminal. Compared with wild-type, dyn-1 mutants at nonpermissive temperature show significantly more 100nm vesicles. ***P < 0.001 (ANOVA with Dunnett’s multiple comparison test). (F) The distance between the DP and presumable endocytic sites was analyzed, wherever membrane contact or continuity (the latter only in dyn-1 mutants) between 100nm vesicle and the PM was observed; genotypes and conditions are indicated.
References
- Gundelfinger ED, Fejtova A. Molecular organization and plasticity of the cytomatrix at the active zone. Curr Opin Neurobiol. 2012;22(3):423–430. -PubMed
- Dittman J, Ryan TA. Molecular circuitry of endocytosis at nerve terminals. Annu Rev Cell Dev Biol. 2009;25:133–160. -PubMed
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Other Literature Sources
Molecular Biology Databases