Presynapses contain distinct actin nanostructures

Abstract

The architecture of the actin cytoskeleton that concentrates at presynapses remains poorly known, hindering our understanding of its roles in synaptic physiology. In this work, we measure and visualize presynaptic actin by diffraction-limited and super-resolution microscopy, thanks to a validated model of bead-induced presynapses in cultured neurons. We identify a major population of actin-enriched presynapses that concentrates more presynaptic components and shows higher synaptic vesicle cycling than their non-enriched counterparts. Pharmacological perturbations point to an optimal actin amount and the presence of distinct actin structures within presynapses. We directly visualize these nanostructures using Single Molecule Localization Microscopy (SMLM), defining three distinct types: an actin mesh at the active zone, actin rails between the active zone and deeper reserve pools, and actin corrals around the whole presynaptic compartment. Finally, CRISPR-tagging of endogenous actin allows us to validate our results in natural synapses between cultured neurons, confirming the role of actin enrichment and the presence of three types of presynaptic actin nanostructures.

Figure S1.

Postsynaptic actin impedes the visualization of presynaptic content and organization in hippocampal neurons, even using state-of-the-art super-resolution microscopy. (A–C) Images of neurons after 9 d in vitro (div), stained for synaptophysin (purple), map2 (blue), ß2-spectrin (purple), and actin (gray). (A) Widefield image. (B) Zooms from the highlighted area in A, overlaying the widefield image of synaptophysin, map2, and ß2-spectrin with the STORM image of actin. (C) Zooms from the highlighted area in B, with the top row showing overlayed widefield image of synaptophysin and map2 with the STORM image of actin (gray) and bottom showing the isolated actin STORM image with axon shaft highlighted in orange dashed lines. (D–F) Images of neurons after 9 div, stained for VGAT (green), VGLUT (orange), ß2-spectrin (purple), and actin (gray). (C) Widefield image. (D) Widefield image. (E) Zoom from the highlighted area in D, overlaying widefield images of VGAT, VGLUT, and ß2-spectrin with the STORM image of actin. F, zooms from the highlighted areas in E, with the top row showing overlayed widefield image of VGAT and VGLUT with the STORM image of actin (gray) and the bottom showing the isolated actin STORM image with axon shaft highlighted in orange dashed lines. Scale bars for A and D, 10 µm; for B and E, 2 µm; for C and F, 0.5 µm.

Figure 1.

Polylysine-coated beads induce functional presynaptic specializations at axon-bead contacts. (A) Widefield fluorescence image of cultured neurons seeded with polylysine-coated beads (gray) after 8 div and fixed 48 h later at 10 div, labeled for actin (green), bassoon (purple), synapsin (orange), and map2 (gray). (B) Zooms corresponding to the areas highlighted in A: top row, natural synapse at axon-dendrite contact (NS); middle row, induced presynapse at an axon-bead contact (S+); bottom row, axon-bead contact with no induced presynapse (S−). The position of the bead is indicated by the dashed gray circle. (C) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin (green), synaptophysin (purple), vamp2 (orange), and map2 (gray). (D) Zooms corresponding to the natural synapse, S+ and S− axon-bead contacts highlighted in C. (E) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin (green), synaptophysin (purple), map2 (gray), and feeding with anti-synaptotagmin antibody (syt) during constitutive vesicular cycling (orange). (F) Zooms corresponding to the natural synapse (NS), S+, and S− axon-bead contacts highlighted in E. Scale bars for A, C, and E: 20 µm; for B, D, and F: 2 µm. (G) Quantification of the proportion of S+ (blue) and S− (gray) axon-bead contacts. (H) Quantification of the labeling intensity for bassoon (dark purple), synapsin (orange), synaptophysin (purple), and vamp2 (yellow) at natural synapses (NS), induced presynapses at axon-bead contacts (S+) and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at S+ contacts. (I) Quantification of the labeling intensity for actin at natural synapses (NS), induced presynapses at axon-bead contacts (S+) and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at S+ contacts. (J) Quantification of the labeling intensity after syt feeding for constitutive (orange) and stimulated (KCl-induced, yellow) vesicular cycling at natural synapses (NS), induced presynapses at axon-bead contacts (S+), and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at S+ contacts. Significance signs on graphs compare to the value for the same labeling in bead-induced presynapses (S+, normalized to 1.0). See Data S1 file for detailed statistics.

Figure S2.

Optical methods to assess vesicular cycling at bead-induced presynapses. (A–C) Anti-synaptotagmin (syt) antibody feeding experiments to assess the cycling activity of presynapses. (A) Cartoon of the anti-syt feeding experiments. Living neurons are incubated with the syt antibody (orange) directed against an extracellular epitope of synaptotagmin (yellow), either for 60 min to measure constitutive cycling, or for 3 min in the presence of 50 mM KCl to measure stimulated cycling. Neurons are then fixed and the syt antibody is revealed with a secondary antibody (green). (B) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin (green), synaptophysin (purple), map2 (gray), and feeding with anti-synaptotagmin antibody (syt) during a 3-min incubation with KCl (stimulated cycling). (C) Zooms corresponding to the natural synapse (NS), S+, and S− axon-bead contacts highlighted in B. Scale bars for B, 20 µm; for C, 2 µm. (D–G) FM1-43 dye loading/release experiment to assess the cycling activity of presynapses. (D) Cartoon of the FM1-43 experiment. Living neurons are first loaded with FM1-43 using a 3-min incubation in 50 mM KCl followed by 15 min of recovery, then images of the “loaded” time point are taken. FM1-43 is then released using a second 3-min incubation with 50 mM KCl and 15-min recovery, before images are taken of the “released” time point. (E) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin using SiR (green) after loading with FM1-43 (orange). (F) Zooms corresponding to area highlighted in E, showing the images obtained after loading (left column) and release (right column) of FM1-43. Beads are indicated by dashed-line circles labeling the induced presynapses (S+) and axon-bead contact with no presynapse (S−). Scale bars for B, 20 µm; for C, 5 µm. (G) Quantification of the FM staining intensity after loading (orange) and release (yellow) in natural synapses (NS) and bead-induced presynapses (S+) as well as axon-bead contacts devoid of a presynapse (S−). Significance signs on the graph compare to the value for the same labeling condition in bead-induced presynapses (S+, normalized to 1.0 for the loading condition). See Data S1 file for detailed statistics.

Figure 2.

A major proportion of bead-induced presynapses is enriched in actin and concentrates more presynaptic components than non-enriched induced presynapses. (A) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin (green), bassoon (purple), synapsin (orange), and map2 (gray). (B) Zooms corresponding to the areas highlighted in A: top row, actin-enriched induced presynapse at an axon-bead contact (A+); second row, induced presynapse at an axon-bead contact with no actin enrichment (A−); third row, axon-bead contact with no induced presynapse (S−); bottom row, natural synapse at axon-dendrite contact (NS). (C) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin (green), synaptophysin (purple), vamp2 (orange), and map2 (gray). (D) Zooms corresponding to the A+, A−, S− axon-bead contacts and natural synapses (NS) highlighted in C. Scale bars in A and C: 20 µm; B and D: 2 µm. (E) Quantification of the proportion of A+ (dark blue), A− (blue), and S− (gray) axon-bead contacts. (F) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (orange), and vamp2 (yellow) at actin-enriched presynapses (A+), induced presynapses with no actin enrichment (A−), and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at A+ presynapses. Significance signs on graphs compare to the value for the same labeling in actin-enriched presynapses (A+, normalized to 1.0). See Data S1 file for detailed statistics.

Figure S3.

Presynapses induced 48 h after seeding more mature, 14-div neurons show a similar proportion of actin enrichment and with similar presynaptic content. (A) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 14 div, labeled for actin (green), bassoon (purple), synapsin (orange), and map2 (gray). (B) Zooms corresponding to the areas highlighted in A: top row, actin-enriched induced presynapse at an axon-bead contact (A+); second row, induced presynapse at an axon-bead contact with no actin enrichment (A−); third row, axon-bead contact with no induced presynapse (S−); bottom row, natural synapse at axon-dendrite contact (NS). (C) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 14 div, labeled for actin (green), synaptophysin (purple), vamp2 (orange), and map2 (gray). (D) Zooms corresponding to the A+, A−, S− axon-bead contacts and natural synapses (NS) highlighted in C. Scale bars in A and C: 20 µm; B and D: 2 µm. (E) Quantification of the proportion of A+ (dark blue) and A− (blue) at axon-bead contacts that resulted in an induced presynapse. (F) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (orange), and vamp2 (yellow) at actin-enriched presynapses (A+), induced presynapses with no actin enrichment (A−), and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at A+ presynapses. Significance signs on graphs compare to the value for the same labeling in bead-induced presynapses (S+, normalized to 1.0). See Data S1 file for detailed statistics.

Figure S4.

Presynapses induced for 48 and 96 h show a similar proportion of actin enrichment with similar presynaptic content. (A) Widefield fluorescence image of cultured neurons 96 h after bead seeding at 6 div, labeled for actin (green), bassoon (purple), synapsin (orange), and map2 (gray). (B) Zooms corresponding to the areas highlighted in A: top row, actin-enriched induced presynapse at an axon-bead contact (A+); second row, induced presynapse at an axon-bead contact with no actin enrichment (A−); third row, axon-bead contact with no induced presynapse (S−); bottom row, natural synapse at axon-dendrite contact (NS). (C) Widefield fluorescence image of cultured neurons 96 h after bead seeding at 6 div, labeled for actin (green), synaptophysin (purple), vamp2 (orange), and map2 (gray). (D) Zooms corresponding to the A+, A−, S− axon-bead contacts and natural synapses (NS) highlighted in C. Scale bars in A and C: 20 µm; B and D: 2 µm. (E) Quantification of the proportion of A+ (dark blue) and A− (blue) at axon-bead contacts that resulted in an induced presynapse for neurons after 48 and 96 h incubation with beads. (F) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (orange), and vamp2 (yellow) at actin-enriched presynapses (A+), induced presynapses with no actin enrichment (A−), and axon-bead contacts devoid of presynapse (S−), 48 h (left), and 96 h (right) after bead seeding, normalized to the intensity of A+ presynapses after 48 h. Significance signs on graphs (bottom: *, ns) compare to the value for the labeling on bead-induced presynapses (S+, normalized to 1.0 at 48 h); significance signs (top: °, ns) compare to the value for the same labeling and type of induced presynapse (A+ or A−) between 48 and 96 h. See Data S1 file for detailed statistics.

Figure S5.

Bead-induced inhibitory and excitatory presynapses show similar actin and presynaptic component content distribution. (A) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin (green), synapsin (purple), VGLUT (orange), and map2 (gray). (B) Zooms corresponding to the areas highlighted in A: top row, actin-enriched induced presynapse at an axon-bead contact (A+); second row, induced presynapse at an axon-bead contact with no actin enrichment (A−); third row, axon-bead contact with no induced presynapse (S−); bottom row, natural synapse at axon-dendrite contact (NS). (C) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin (green), synaptophysin (purple), VGAT (orange), and map2 (gray). (D) Zooms corresponding to the A+, A−, S− axon-bead contacts and natural synapses (NS) highlighted in C. Scale bars in A and C: 20 µm; B and D: 2 µm. (E) Quantification of the proportion of A+ (dark blue) and A− (blue) at axon-bead contacts that resulted in an induced VGLUT-positive (left bar) and VGAT-positive (right bar) presynapse. (F) Left graph, quantification of the labeling intensity for actin (green), synapsin (dark purple), and VGLUT (orange) at actin-enriched presynapses (A+), induced presynapses with no actin enrichment (A−), and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at A+ presynapses. Right graph, quantification of the labeling intensity for actin (green), synaptophysin (purple), and VGAT (yellow) at A+ and A− induced presynapses as well as S− axon bead contacts. Significance signs on graphs compare to the value for the same labeling in actin-enriched presynapses (A+, normalized to 1.0). See Data S1 file for detailed statistics.

Figure S6.

Induced presynapses are similar to natural presynapses between 48 and 96 h after bead seeding and between inhibitory and excitatory presynapses; acute actin perturbation effect on the accumulation of presynaptic components is similar at natural synapses and induced presynapses. (A) Quantification of the labeling intensity for bassoon (dark purple), synaptophysin (purple), synapsin (orange), and vamp2 (yellow) at natural synapses (NS), induced presynapses at axon-bead contacts (S+) and axon-bead contacts devoid of presynapse (S−) 48 h (left) or 96 h (right) after bead seeding, normalized to the intensity at S+ contacts after 48 h. (B) Quantification of the labeling intensity for actin at natural synapses (NS), induced presynapses at axon-bead contacts (S+), and axon-bead contacts devoid of presynapse (S−) 48 h (left) or 96 h (right) after bead seeding, normalized to the intensity at S+ contacts after 48 h. (C) Quantification of the labeling intensity for synapsin (dark purple) and VGLUT (orange) at excitatory natural synapses (NS), induced presynapses at axon-bead contacts (S+), and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at S+ contacts. (D) Quantification of the labeling intensity for synaptophysin (purple) and VGAT (yellow) at inhibitory natural synapses (NS), induced presynapses at axon-bead contacts (S+) and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at S+ contacts. (E) Quantification of the labeling intensity for actin at natural synapses (NS), induced presynapses at axon-bead contacts (S+), and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at S+ contacts for excitatory (left) or inhibitory (right) presynapses. Significance signs on graphs A–E compare to the value for the same labeling in bead-induced presynapses (S+, normalized to 1.0). (F–H) Representative images of natural presynapses in control condition (F) or after treatment with swinholide A (G) or cucurbitacine E (H) labeled for actin, bassoon, and synapsin (green, purple, and orange respectively, top-left image); actin, synaptophysin, and vamp2 (green, purple and orange respectively, top-right image); actin, synaptophysin and constitutive feeding with anti-synaptotagmin antibody (green, purple and orange respectively, bottom-left image); actin, synaptophysin, and stimulated feeding with anti-synaptotagmin antibody (green, purple, and orange, respectively, bottom-right image). Zooms are taken from the images shown in Fig. 4, A–C. Scale bars, 2 µm. (I) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (dark blue), vamp2 (blue), and after syt feeding for constitutive (orange) and stimulated (yellow) vesicular cycling at natural synapses in the control condition and after swinholide A treatment, normalized to control natural synapses. (J) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (dark blue), vamp2 (blue), and after syt feeding for constitutive (orange) and stimulated (yellow) vesicular cycling at natural synapses in the control condition and after cucurbitacin E treatment, normalized to control natural synapses. Significance signs on graphs I and J compare to the value for the same labeling in the control condition for natural synapses (normalized to 1.0). See Data S1 file for detailed statistics.

Figure 3.

Actin-enriched induced presynapses have a higher vesicular cycling than non-enriched induced presynapses. (A) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin (green), synaptophysin (purple), map2 (gray), and feeding with anti-synaptotagmin antibody (syt) during constitutive cycling (orange). (B) Zooms corresponding to the A+, A−, S− axon-bead contacts and natural synapses (NS) highlighted in B. (C) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for actin (green), synaptophysin (purple), and map2 (gray) and feeding with anti-synaptotagmin antibody (syt) during KCl-stimulated vesicular cycling (orange). (D) Zooms corresponding to the A+, A−, S− axon-bead contacts and natural synapses (NS) highlighted in C. Scale bars in A and C, 20 µm; B and D, 2 µm. (E) Quantification of the labeling intensity for actin (green), synaptophysin (purple), and after syt feeding for constitutive (orange) and stimulated (yellow) vesicular cycling at actin-enriched presynapses (A+), induced presynapses with no actin enrichment (A−), and axon-bead contacts devoid of presynapse (S−), normalized to the intensity at A+ presynapses. (F) Zooms on axon-bead contacts of living neurons stained with SiR-actin (green) and loaded with FM1-43 (orange, left image) before release (right image). Left two images show an A+ presynapse, right images show an A− presynapse. Scale bar, 2 µm. (G) Quantification of the FM staining intensity after loading (orange) and release (yellow) in A+ and A− induced presynapses as well as axon-bead contacts devoid of a presynapse (S−). Significance signs on graphs compare to the value for the same labeling condition in actin-enriched presynapses (normalized to 1.0 for A+ in E, normalized to 1.0 for the A+ in loading condition in G). See Data S1 file for detailed statistics.

Figure 4.

Acute actin disassembly and over-stabilization lower the accumulation of presynaptic components at induced presynapses. (A) Top, widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, control treated with vehicle (DMSO) for 1 h and labeled for actin (green), bassoon (purple), synapsin (orange), and map2 (gray). Bottom, panels showing A+ presynapses labeled for actin, presynaptic components, and after syt feeding. The first panel is a zoom of the area highlighted in the top image. (B) Top, widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, treated with 1 µM swinholide A (swin) for 1 h, and labeled for actin (green), bassoon (purple), synapsin (orange), and map2 (gray). Bottom, panels showing A− presynapses labeled for actin, presynaptic components, and after syt feeding. The first panel is a zoom of the area highlighted in the top image. (C) Top, widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, treated with 5 µM cucurbitacin E (cuc) for 3 h, and labeled for actin (green), bassoon (purple), synapsin (orange), and map2 (gray). Bottom, panels showing A+ presynapses labeled for actin, presynaptic components, and after syt feeding. The first panel is a zoom of the area highlighted in the top image. Scale bars on large images in A–C, 20 µm; on zoomed images, 2 µm. (D) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (dark blue), vamp2 (blue), and after syt feeding for constitutive (orange), and stimulated (yellow) vesicular cycling at actin-enriched presynapses (A+) and induced presynapses with no actin enrichment (A−) in the control condition, and at A− presynapses after swin treatment. (E) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (dark blue), vamp2 (blue), and after syt feeding for constitutive (orange), and stimulated (yellow) vesicular cycling at actin-enriched presynapses (A+) and induced presynapses with no actin enrichment (A−) in the control condition, and at A+ presynapses after cuc treatment. Significance signs (bottom: *, ns) on graphs compare to the value for the same labeling in the control condition for A+ presynapses (normalized to 1.0); significance signs (top: °, ns) compare to the value for the same labeling in the control condition for A− presynapses. See Data S1 file for detailed statistics.

Figure 5.

Arp2/3 and formins inhibitors have distinct effects on actin and presynaptic components at induced presynapses. (A) Top, widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, control-treated with vehicle (DMSO) for 1 h and labeled for actin (green), bassoon (purple), and map2 (gray). Bottom, panels showing A+ and A− presynapses labeled for actin, presynaptic components, and after syt feeding. The top panels are zooms of the areas highlighted in the top image. (B) Top, widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, treated with 50 μM CK666 (Arp2/3 inhibitor) for 1 h, and labeled for actin (green), bassoon (purple), and map2 (gray). Bottom, panels showing A+ and A− presynapses labeled for actin, presynaptic components, and after syt feeding. The top panels are zooms of the areas highlighted in the top image. (C) Top, widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, treated with 30 µM SMIFH2 (formins inhibitor) for 3 h, and labeled for actin (green), bassoon (purple), and map2 (gray). Bottom, panels showing A+ and A− presynapses labeled for actin, presynaptic components and after syt feeding. Top panels are zooms of the areas highlighted in the top image. Scale bars on large images in A–C, 20 µm; on zoomed images, 2 µm. (D) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (dark blue), vamp2 (blue), and after syt feeding for constitutive (orange), and stimulated (yellow) vesicular cycling at actin-enriched presynapses (A+) in the control condition, and after CK666 or SMIFH2 treatment. Significance signs on graphs compare to the value for the same labeling in the control condition for A+ presynapses (normalized to 1.0). See Data S1 file for detailed statistics.

Figure S7.

Acute actin nucleators inhibition effect on the accumulation of presynaptic components is similar at natural synapses and induced presynapses; effect of CK666 on additional presynaptic components. (A–C) Representative images of natural presynapses in control condition (A) or after treatment with CK666 (B) or SMIFH2 (C) labeled for actin, bassoon, and synapsin (green, purple, and orange, respectively, top-left image); actin, synaptophysin, and vamp2 (green, purple and orange respectively, top-right image); actin, synaptophysin and constitutive feeding with anti-synaptotagmin antibody (green, purple, and orange, respectively, bottom-left image); actin, synaptophysin and stimulated feeding with anti-synaptotagmin antibody (green, purple, and orange, respectively, bottom-right image). Zooms are taken from the images shown in Fig. 5, A–C. Scale bars, 2 µm. (D) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (dark blue), vamp2 (blue), and after syt feeding for constitutive (orange), and stimulated (yellow) vesicular cycling at induced presynapses devoid of actin enrichment (A−) in the control condition and after CK666 or SMIFH2 treatment, normalized to control A+ presynapses. Significance signs on graph compare to the value for the same labeling in the control condition for A− presynapses (normalized to 1.0). (E) Quantification of the labeling intensity for actin (green), bassoon (dark purple), synaptophysin (purple), synapsin (dark blue), vamp2 (blue), and after syt feeding for constitutive (orange), and stimulated (yellow) vesicular cycling at natural synapses in the control condition and after CK666 or SMIFH2 treatment, normalized to control natural synapses. Significance signs on the graph compare to the value for the same labeling in the control condition for natural synapses (normalized to 1.0). (F and G) Representative images of induced and natural presynapses in the control condition (F) or after treatment with CK666 (G) labeled for actin, syntaxin, and synapsin (green, purple, and orange respectively, top row); actin, SV2 and munc13 (green, purple, and orange, respectively, bottom row). Scale bars, 2 µm. (H) Quantification of the labeling intensity for actin (green), syntaxin (dark purple), SV2 (purple), synaptophysin (dark blue), and munc13 (orange), at induced presynapses enriched in actin (A+) or devoid of actin enrichment (A−), in the control condition and after CK666 treatment, normalized to control A+ presynapses. Significance signs (bottom: *, ns) on graphs compare to the value for the same measurement in the control condition for the same type of presynapse (A+ or A−, normalized to 1.0); significance signs (top: °, ns) compare to the value for the same measurement in the control condition for A+ presynapses. (I) Quantification of the labeling intensity for actin (green), syntaxin (dark purple), SV2 (purple), synaptophysin (dark blue), and munc13 (orange), at natural presynapses (NS), in the control condition and after CK666, normalized to control natural synapses. Significance signs on graphs I and J compare to the value for the same labeling in the control condition for natural synapses (normalized to 1.0). See Data S1 file for detailed statistics.

Figure 6.

STORM visualizes distinct actin nanostructures at induced presynapses. (A) Widefield fluorescence image of cultured neurons 2 d after bead seeding at 8 div, labeled for ß2-spectrin (purple), synapsin (orange), and actin (gray). (B) Zooms on a natural synapse at an axon–dendrite contact (top row) and on an induced presynapse at an axon–bead contact (bottom row), corresponding to the highlighted areas in A. Left columns are widefield images of actin (gray), middle channels are overlays of widefield images of ß2-spectrin (purple), synapsin (orange), and actin (gray); right columns are overlays of widefield images of ß2-spectrin (purple) and synapsin (orange) with the STORM image of actin (gray). (C) Zoomed STORM images of the natural synapse (left) and bead-induced presynapse (right). (D) Same image as C, left, with color highlighting of presynaptic actin structures: active zone mesh (red), actin rails (green), perisynaptic corral (blue). (E) Additional representative images of bead-induced presynapses. The first column shows the widefield image of ß2-spectrin (purple), synapsin (orange), and actin (gray). The second column is a zoom of the highlighted area in the first image. The third column shows an overlay of widefield images of ß2-spectrin (purple) and synapsin (orange) with the STORM image of actin (gray). The fourth column is the isolated STORM image of actin (gray). The fifth column is the STORM image of actin with highlighted presynaptic actin structures: active zone mesh (red), actin rails (green), and perisynaptic corral (blue). Scale bars in A, 10 µm; in B–D, 1 µm; in E, 5 µm (first column) and 1 µm (other columns).

Figure 7.

Actin mesh, rails, and corrals are present in various types of induced presynapses. (A) Three-color SMLM image of a bead-induced presynapse labeled for ß2-spectrin (DNA-PAINT image), synapsin (DNA-PAINT image), and actin (STORM image). The fourth column is an overlay of the three channels, while the fifth column shows the actin STORM image with highlighted presynaptic actin structures. (B) Actin nanostructures in A+ and A− induced presynapses along the same axon. The top-left panel shows the widefield image of ß2-spectrin (purple), synaptophysin (orange), and actin (gray). The top-center panel is a zoom of the highlighted area in the first image. The top-right panel shows an overlay of widefield images of ß2-spectrin (purple) and synaptophysin (orange) with the STORM image of actin (gray). The bottom-left panel is the isolated STORM image of actin (gray). The bottom-center panel is the isolated 3D-STORM image of actin, color-coded for depth (with Z scale in the top right corner). The bottom-right panel is the STORM image of actin with highlighted presynaptic actin structures in the A+ and A− induced presynapses: active zone mesh (red), actin rails (green), perisynaptic corral (blue). (C) Semi-quantitative assessment of the frequency of identification for presynaptic actin structures: actin mesh (red), rails (green), and corral (blue) from A+ and A− induced presynapses. (D) Quantification of the size of the mesh (red) and corral (blue) on 2D-projected STORM images of A+ and A− induced presynapses. Significance signs on the graph compare to the value of a given nanostructure size between A+ and A− synapses. See Data S1 file for detailed statistics. (E) Actin nanostructures in an identified bead-induced glutamatergic presynapse. First column shows the widefield image of ß2-spectrin (purple), VGLUT (orange), and actin (gray). The second column is a zoom of the highlighted area in the first image. The third column shows an overlay of widefield images of ß2-spectrin (purple) and VGLUT (orange) with the STORM image of actin (gray). The fourth column is 3D-STORM image of actin, color-coded for depth (with Z scale in the top right corner). The fifth column is the STORM image of actin with highlighted presynaptic actin structures: active zone mesh (red), actin rails (green), perisynaptic corral (blue). (F) Actin nanostructures in an identified bead-induced GABAergic presynapse. The first column shows the widefield image of ß2-spectrin (purple), VGAT (green), and actin (gray). The second column is a zoom of the highlighted area in the first image. The third column shows an overlay of widefield images of ß2-spectrin (purple) and VGAT (green) with the STORM image of actin (gray). The fourth column is 3D-STORM image of actin, color-coded for depth (with Z scale in the top right corner). The fifth column is the STORM image of actin with highlighted presynaptic actin structures: active zone mesh (red), actin rails (green), and perisynaptic corral (blue). Scale bars in A, 1 µm; in B, 5 µm (top left panel) and 1 µm (other panels); in E and F, 5 µm (left column panels) and 1 µm (other panels).

Figure 8.

Presynaptic actin nanostructures are distinctly dependent on Arp2/3 and formins. (A–C) Images of induced presynapses from cultured neurons 2 d after bead seeding at 8 div. The first column shows a widefield image of ß2-spectrin (purple), synapsin (orange), and actin (gray). The second column is a zoom of the highlighted area in the first column image. The third column shows an overlay of widefield images of ß2-spectrin (purple) and synapsin (orange) with the STORM image of actin (gray). The fourth column is the isolated 3D-STORM image of actin, color-coded for depth (with Z scale in the top right corner). The fifth column is the STORM image of actin with highlighted presynaptic actin structures: active zone mesh (red), actin rails (green), perisynaptic corral (blue). (A) Representative examples from neuronal cultures treated with DMSO for 1 h (control condition). (B) Two examples after treatment with 50 μM CK666 for 1 h. (C) Two examples after treatment with 30 µM SMIFH2 for 1 h. (D) Semi-quantitative assessment of the frequency of identification for presynaptic actin structures: actin mesh, rails, and corrals, from all the presynapses imaged in the control condition (ctrl) or after treatment with CK666 (CK) or SMIFH2 (SMIF). (E) Two-color STORM images of actin and Arp2/3 (anti-p34 antibody) at two representative bead-induced presynapses. The first column shows an overlay of the widefield image of synaptophysin (purple) with the STORM images of actin (gray) and Arp2/3 (yellow). The second column shows the overlayed STORM images of actin (gray) and Arp2/3 (yellow). The third column adds highlighted presynaptic actin structures: active zone mesh (red), actin rails (green), and perisynaptic corral (blue) to the image in the second column. Scale bars in A–C, 5 µm (first column) and 1 µm (other columns); in E, 1 µm.

Figure 9.

Revealing presynaptic actin enrichment in natural synapses and its role using CRISPR-tagging of endogenous actin. (A) Widefield fluorescence image of cultured neurons infected after plating with CRISPR AAVs for the knock-in of an HA tag on endogenous actin, fed with a synaptotagmin antibody for 1 h (orange) and fixed at 9 div, then labeled for HA (green), synaptophysin (purple), and map2 (gray). (B) Zooms corresponding to the area highlighted in A showing isolated channels and overlay of the axon of a HA-actin knocked-in neuron forming actin-enriched A+ (1 and 3) and non-enriched A− (2) presynapses. (C) Zooms on individual presynapses highlighted in B. (D) Quantification of the proportion of A+ (dark blue) and A− (blue) presynapses along the axons of HA-actin neurons. (E) Quantification of the labeling intensity for HA-actin (green), synaptophysin (purple), and constitutive synaptotagmin feeding (orange) at HA-actin-enriched presynapses (A+), induced presynapses with no actin enrichment (A−), and axon shafts contacts devoid of presynapse (S−), normalized to the intensity at A+ presynapses. Significance signs on graphs compare to the value for the same labeling in A+ presynapses (normalized to 1.0).

Figure 10.

Actin mesh, rails and corrals are present in natural presynapses. (A) Widefield fluorescence image of cultured neurons infected after plating with CRISPR AAVs for the knock-in of a FLAG-ALFA tag on endogenous actin, fixed at 9 div and labeled for FLAG (gray), ß2-spectrin (purple), map2 (blue), and synapsin (orange). (B) Zooms corresponding to the area highlighted in A showing the FLAG-ALFA-actin isolated channel and overlay for the axon arborization of a FLAG-ALFA-actin knocked-in neuron forming presynapses on nearby neurons. (C and D) STORM images of actin nanostructures in the presynapses of ALFA-actin knock-in neurons labeled for ALFA using primary and secondary antibodies. The first column shows the widefield image of map2 (blue), synaptophysin (orange), and ALFA-actin (gray). The second and third columns are zooms of the highlighted areas, showing an overlay of widefield images of map2 (blue) and synaptophysin (orange) with the STORM image of ALFA-actin (gray). Fourth column is 3D-STORM image of ALFA-actin, color-coded for depth (with Z scale in the top right corner). Fifth column is the STORM image of ALFA-actin with highlighted presynaptic actin structures: active zone mesh (red), actin rails (green), perisynaptic corral (blue). (E and F) DNA-PAINT images of actin nanostructures in the presynapses of 3ALFA-actin knock-in neurons labeled for ALFA using an anti-ALFA nanobody. The first column shows the overlay of the widefield image of map2 (blue) and synaptophysin (orange) with the PAINT image of 3ALFA-actin (gray). The second column is 3D-STORM image of 3ALFA-actin, color-coded for depth (with Z scale in the top right corner). The third column is the PAINT image of 3ALFA-actin with highlighted presynaptic actin structures: active zone mesh (red), actin rails (green), perisynaptic corral (blue). Scale bars in A, 50 µm; in B, 10 µm; in C, 5 µm (first panel), 2 µm (second panel), 1 µm (other panels); in D, 5 µm (first panel), 2 µm (second panel), 1 µm (third panel), 0.5 µm (fourth and fifth panels); in E and F, 1 µm (first column), 0.5 µm (second and third columns). (G) Cartoon summarizing the identified presynaptic actin nanostructures and their putative role: the branched actin mesh at the active zone where vesicles are exocytosed; linear actin rails (green) help vesicles move between the exocytosis, endocytosis zones, and intracellular pools; the perisynaptic actin corral (blue) scaffolds the vesicular pool.