Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Aug 21:8:e46356.
doi: 10.7554/eLife.46356.

Structural plasticity of dendritic secretory compartments during LTP-induced synaptogenesis

Yelena D Kulik  1 , Deborah J Watson  1 , Guan Cao  1 , Masaaki Kuwajima  1 , Kristen M Harris  1
Affiliations

Structural plasticity of dendritic secretory compartments during LTP-induced synaptogenesis

Yelena D Kulik et al. Elife. .

Abstract

Long-term potentiation (LTP), an increase in synaptic efficacy following high-frequency stimulation, is widely considered a mechanism of learning. LTP involves local remodeling of dendritic spines and synapses. Smooth endoplasmic reticulum (SER) and endosomal compartments could provide local stores of membrane and proteins, bypassing the distant Golgi apparatus. To test this hypothesis, effects of LTP were compared to control stimulation in rat hippocampal area CA1 at postnatal day 15 (P15). By two hours, small spines lacking SER increased after LTP, whereas large spines did not change in frequency, size, or SER content. Total SER volume decreased after LTP consistent with transfer of membrane to the added spines. Shaft SER remained more abundant in spiny than aspiny dendritic regions, apparently supporting the added spines. Recycling endosomes were elevated specifically in small spines after LTP. These findings suggest local secretory trafficking contributes to LTP-induced synaptogenesis and primes the new spines for future plasticity.

Keywords: 3D electron microscopy; development; long-term potentiation; neuroscience; plasticity; rat; synapse; ultrastructure.

PubMed Disclaimer

Conflict of interest statement

YK, DW, GC, MK, KH No competing interests declared

Figures

Figure 1.
Figure 1.. Within-slice experimental design and electrophysiological outcome.
(A) Illustration of an acute slice from a P15 rat hippocampus with a recording electrode (rec.) in the middle of CA1 stratum radiatum between two bipolar stimulating electrodes (S1 and S2). S1 and S2 are separated by 600-800 µm. The two experiments were counterbalanced for which of the two electrodes delivered TBS or control stimulation. Tissue samples collected for 3DEM were located ~120 µm beneath and to the side of the stimulating electrodes. D.G., dentate gyrus; Sub., subiculum. (B) Representative waveforms from control (CON, blue) and TBS (LTP, red) sites. Each waveform is the average of the final 10 responses to each stimulating electrode obtained for the last 20 min before delivery of TBS at time 0 (light color) and for 20 minutes before the end of the experiment at 120 min after TBS (dark color). The stimulus intensity was set at population spike threshold to activate a large fraction of the axons in the field of each stimulating electrode. The positive deflection in the post-TBS waveform at ~3-4 ms reflects synchronous firing of pyramidal cells with LTP. (C) Changes in the slope of the field excitatory postsynaptic potential (fEPSP), expressed as a percentage of the average baseline response to test-pulses, were recorded for 20 min before delivery of TBS at time 0 (red) or control stimulation (blue). Responses were recorded for n=2 slices for 120 min after the first TBS train, then fixed and processed for 3DEM as described in Methods. Error bars are SEM. Adapted from Watson et al. (2016) where it was originally published under a CC BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/).
Figure 2.
Figure 2.. The limited occupancy of spines by SER does not increase during spinogenesis in the LTP condition.
(A) Sample serial section EMs (left) and representative 3D reconstructions of dendrites (right) from control (top) and LTP (bottom) conditions, illustrating dendrites (yellow), SER (green), and synapses (red). Synaptic area was measured as the total surface area of the PSD. Arrows point to SER-containing spines. (B) Spine density (#/µm) binned for PSD area. Significant increase in spines following TBS was carried by spines in the category with the smallest PSD areas (<0.05 µm2; ANOVA F(1,12)=50.707, p=0.00001, η2 = 0.81). No statistically significant changes occurred in the frequency of spines with larger synapses (PSD area 0.05 to 0.1 µm2, ANOVA F(1,12)=1.079, p=0.31941; PSD area 0.1 to 0.15 µm2, ANOVA F(1,12)=0.09638, p=0.76154; PSD area 0.15 to 0.2 µm2, ANOVA F(1,12)=3.5065, p=0.08569; PSD area >0.2 µm2, ANOVA F(1,11)=3.0778, p=0.10484). Control n = 8, LTP n = 8 dendrites. (C) Decrease in percentage of spines containing SER following TBS (ANOVA F(1,12)=10.599, p=0.00688, η2 = 0.87). Control n = 8, LTP n = 8 dendrites. (D–F) SER content for spines with PSD areas less than 0.05 µm2. (D) No statistically significant difference between control and LTP conditions in density of spines with SER (ANOVA F(1,12)=2.59, p=0.13322). Control n = 8, LTP n = 8 dendrites. (E) No statistically significant difference in average SER volume per SER-containing spine between control and LTP conditions (hnANOVA F(1,14)=.73111, p=0.40692). Control n = 12, LTP n = 15 spines. (F) No statistically significant difference in SER surface area per SER-containing spine between control and LTP conditions (hnANOVA F(1,14)=3.3120, p=0.09022). Control n = 12, LTP n = 15 spines. (G–I) SER content for spines with total PSD area equal to or greater than 0.05 µm2. (G) No statistically significant difference in density of spines with SER between control and LTP conditions (ANOVA F(1,12)=2.1641, p=0.16700). Control n = 8, LTP n = 8 dendrites. (H) Reduction in average SER volume per SER-containing spine in the LTP relative to control condition (hnANOVA F(1,38)=5.7205, p=0.02182, η2 = 0.13). Control n = 29, LTP n = 25 spines. (I) Reduction in average SER surface area in SER-containing spines in the LTP relative to control condition (hnANOVA F(1, 38)=4.5873 p=0.03868, η2 = 0.12). Control n = 29, LTP n = 25 spines. Bar graphs show mean ± S.E.M. Control (CON, blue) and TBS (LTP, red).
Figure 2—figure supplement 1.
Figure 2—figure supplement 1.. All analyzed dendrites fully reconstructed with SER, aligned left to right from least to greatest spine density.
Scale cube is 0.5 µm on each side.
Figure 3.
Figure 3.. Reduction in shaft SER following LTP.
(A) Electron micrographs showing the dendrite (yellow), SER (green), and synapses (red). For both control and LTP, the SER in the aspiny segments forms small cross-sectioned tubules, whereas in the spiny segments the SER tubules are broadly expanded. (B) Sample 3D reconstructions from serial section electron micrographs of SER-containing dendrites, illustrating spiny segments (yellow) and aspiny segments (blue) while the other colors match Figure 2. Aspiny segments consist of two or more sections (>100 nm) of no spine origins. Spiny segments had at least one spine and were surrounded by aspiny segments. Scale cube is 0.5 µm on each side. (C) No statistically significant differences between control and LTP conditions were found in surface area of SER in the dendritic shaft (ANOVA F(1,12)=3.8833, p=0.07228). Control n = 8, LTP n = 8 dendrites. (D) Volume of dendritic SER network was reduced in the LTP relative to control conditions (ANOVA F(1,12)=6.4397, p=0.02605, η2 = 0.35). Control n = 8, LTP n = 8 dendrites. (E) Summed cross-sectional area of SER tubules and cisterns as a measure of changes in complexity. More SER on spiny than aspiny sections within both control (hnANOVA F(1,1432) = 51.672, p<0.00000, η2 = 0.034; spiny n = 493, aspiny n = 955 sections) and LTP conditions (hnANOVA F(1,324)=17.535, p=0.00003, η2 = 0.013; spiny n = 714, aspiny n = 626 sections). Reduced SER complexity with LTP for both spiny (hnANOVA F(1,1191) = 51.745, p<0.00000, η2 = 0.019; Control n = 493, LTP n = 714 sections) and aspiny sections (hnANOVA F(1,1565) = 29.991, p<0.00000, η2 = 0.042; Control n = 955, LTP n = 626 sections) relative to control. Bar graphs show mean ± S.E.M. Control (CON, blue) and TBS (LTP, red).
Figure 4.
Figure 4.. Identification of endosomal compartments.
(A) Model of the dendritic endosomal pathway. Clathrin-coated pits (CPs) invaginate, becoming clathrin-coated vesicles (CVs) and large vesicles (LVs) after coat shedding. Large vesicles fuse to form tubules, recycling complexes (RCs), and sorting complexes (SCs) with a multivesicular body (MVB). From here, the sorted material may be sent to the plasma membrane via small vesicles (SVs) that pinch off coated tips of tubules. MVBs may serve as exosomes (Exo) or primary lysosomes, that are more darkly stained than exosomes due to the acidic cytomatrix of lysosomes (adapted from Cooney et al., 2002). Sample electron micrographs illustrate (B) recycling complex (pink arrow) and small vesicles (purple arrow), (C) clathrin-coated pit (orange arrow), (D) sorting complex (light blue arrows point to multivesicular body (MVB) in the center and tubules around it), (E) amorphous vesicle (green arrow), (F) lysosome (black arrow), and (G) whorl (black arrow). Scale bar in (G) is 0.5 µm for all images.
Figure 4—figure supplement 1.
Figure 4—figure supplement 1.. Sample images from the LTP condition of dendritic (yellow) recycling complex with multiple tubules (pink) entering the spine neck, and two small vesicles (purple arrow) in a different dendritic spine.
D28 FZYJV sections 108–111. Scale bar 0.5 µm.
Figure 4—figure supplement 2.
Figure 4—figure supplement 2.. Sample images of coated pit (orange) inside the dendritic shaft (yellow) from a dendrite in the LTP condition, D25 DCPBM sections 121–124.
Scale bar 0.5 µm.
Figure 4—figure supplement 3.
Figure 4—figure supplement 3.. Sample images of sorting complex (turquoise) inside the dendritic shaft from the control condition, with one tubule entering a spine neck (right side row 3).
D26 PWCNZ sections 41–44. Scale bar 0.5 µm.
Figure 4—figure supplement 4.
Figure 4—figure supplement 4.. Sample image of an amorphous vesicle in the dendritic shaft of the LTP condition from D35 DCPBM sections 25–28.
Scale bar 0.5 µm.
Figure 4—figure supplement 5.
Figure 4—figure supplement 5.. Sample images of degradative lysosome (black) in the dendritic shaft (yellow) of the LTP condition from D17 FZYJV sections 146–149.
Scale bar 0.5 µm.
Figure 4—figure supplement 6.
Figure 4—figure supplement 6.. Sample images of degradative whorl (black) in a dendrite (yellow) of the control condition from D69 FXBVK sections 176–180.
Scale bar 0.5 µm.
Figure 5.
Figure 5.. Increased occurrence of endosomes in small spines after LTP.
(A) Sample serial EM sections and representative 3D reconstructed dendrites illustrate the distribution of endosomal compartments from control and LTP conditions. Dendrites are yellow, synapses are red, and color-coded arrows point to endosome-containing spines. The color-coded key in the lower left corner indicates amorphous vesicles (AV), recycling complexes (RC), coated pits (CP), coated vesicles (CV), large vesicles (LV), sorting complexes (SC), small vesicles (SV) and degradative structures (DEG); these abbreviations apply also to the graphs. Vesicles are represented as 100 nm spheres (AV, CP, CV, LV, and SV). The other structures (RC, SC, DEG) are reconstructed in 3D to scale. (B) Endosomal structures in dendritic shafts (#/µm) with relative distributions to aspiny and spiny segments in control (CON) and LTP conditions. Overall, shaft endosomes (hnANOVA F(1,293)=0.93104, p=0.33539), degradative structures (hnANOVA F(1,293)=0.47789, p=0.48993) or constructive endosomal compartments (Constr. = all minus degradative; hnANOVA F(1,293)=0.62167, p=0.43107) did not differ between LTP and control conditions or segment locations. Recycling complexes (RC) were greater in the LTP than control dendritic shafts (hnANOVA F(1,293)=6.4920, p=0.01135, η2 = 0.022), but no significant differences occurred in the other categories: amorphous vesicles (hnANOVA F(1,293)=1.5092, p=0.22025); small vesicles (hnANOVA F(1, 293)=1.1699, p=28031); coated pits, coated vesicles, and large vesicles (hnANOVA F(1,293)=0.89152, p=0.34584); and sorting complexes (hnANOVA F(1,293)=0.45286, p=0.50151). (For control (CON) n = 151 aspiny + spiny segments and for LTP n = 158 aspiny + spiny segments.) (C) More dendritic spines contained endosomes along the dendrites in the LTP than the control condition (ANOVA F(1,12)=18.047, p=0.00113, η2 = 0.60), an effect that was carried by spines with PSD areas less than 0.05 µm2 (ANOVA F(1,12)=23.642, p=0.00039, η2 = 0.66) but not in spines with PSD area >0.05 µm2 (ANOVA F(1,12)=0.84714, p=0.37550). (D) Stability in percentage of spines containing endosomes following TBS (ANOVA F(1,12)=.72158, p=0.41225). (E) Among spines with PSD area less than 0.05 µm2, the increase in occupancy of endosomes was due to more with coated pits, coated vesicles, and large vesicles (ANOVA F(1,12)=4.94433, p=0.046140, η2 = 0.29), recycling complexes (ANOVA F(1,12)=11.009, p=0.00613, η2 = 0.48), and more with small vesicles (ANOVA F(1,12)=5.2575, p=0.04072, η2 = 0.30). No significant changes in spine occupancy occurred for amorphous vesicles (ANOVA F(1,12)=1, p=0.33705), sorting complexes (ANOVA F(1,12)=1, p=0.33705), or degradative structures (ANOVA F(1,12)=0.46689, p=0.5074). Bar graphs show mean ± S.E.M. (For C–E), Control (CON, n = 8 full dendrite reconstructions) and LTP (n = 8 full dendrite reconstructions).
Figure 5—figure supplement 1.
Figure 5—figure supplement 1.. All analyzed dendrites fully reconstructed with constructive endosomes, aligned left to right from least to greatest spine density.
Scale cube is 0.5 µm on each side.
Figure 5—figure supplement 2.
Figure 5—figure supplement 2.. All analyzed dendrites fully reconstructed with intracellular degradative structures, aligned left to right from least to greatest spine density.
Scale cube is 0.5 µm on each side.
Figure 6.
Figure 6.. Model of the contribution of dendritic secretory compartments to LTP-induced synaptogenesis.
Smooth endoplasmic reticulum (SER, green), postsynaptic density (PSD, red), small vesicle or recycling endosome (RE, turquoise), new silent spines (orange), control activation (Con), theta-burst stimulation (TBS), long-term potentiation (LTP), AMPA receptors (AMPAR).
See this image and copyright information in PMC

References

    1. Ashley J, Cordy B, Lucia D, Fradkin LG, Budnik V, Thomson T. Retrovirus-like gag protein Arc1 binds RNA and traffics across synaptic boutons. Cell. 2018;172:262–274. doi: 10.1016/j.cell.2017.12.022. - DOI - PMC - PubMed
    1. Bailey CH, Kandel ER, Harris KM. Structural components of synaptic plasticity and memory consolidation. Cold Spring Harbor Perspectives in Biology. 2015;7:a021758. doi: 10.1101/cshperspect.a021758. - DOI - PMC - PubMed
    1. Bell ME, Bourne JN, Chirillo MA, Mendenhall JM, Kuwajima M, Harris KM. Dynamics of nascent and active zone ultrastructure as synapses enlarge during long-term potentiation in mature Hippocampus. Journal of Comparative Neurology. 2014;522:3861–3884. doi: 10.1002/cne.23646. - DOI - PMC - PubMed
    1. Borgdorff AJ, Choquet D. Regulation of AMPA receptor lateral movements. Nature. 2002;417:649–653. doi: 10.1038/nature00780. - DOI - PubMed
    1. Bourne JN, Kirov SA, Sorra KE, Harris KM. Warmer preparation of hippocampal slices prevents synapse proliferation that might obscure LTP-related structural plasticity. Neuropharmacology. 2007;52:55–59. doi: 10.1016/j.neuropharm.2006.06.020. - DOI - PubMed

Publication types