Data from: Alternative splicing of synaptotagmin 7 regulates oligomerization and short-term synaptic plasticity
Data files
Sep 24, 2026 version files 29.31 GB
-
Fig._1.zip
802.34 MB
-
Fig._2.zip
323.38 MB
-
Fig._3_and_4_Syt7_Alpha.zip
5.49 GB
-
Fig._3_and_4_Syt7_Beta.zip
3.99 GB
-
Fig._3_and_4_Syt7_Gamma.zip
3.44 GB
-
Fig._3_and_4_Syt7_KO.zip
5.58 GB
-
Fig._3_and_4_Syt7_WT.zip
5.22 GB
-
Fig._5.zip
845.02 MB
-
Fig._6.zip
342.89 MB
-
Fig._S13.zip
1.21 GB
-
Fig._S16.zip
29.04 MB
-
Fig._S5.zip
205.07 MB
-
Fig._S6.zip
595.01 MB
-
Fig._S8.zip
1.24 GB
-
README.md
10.72 KB
Abstract
Synaptic plasticity is crucial for learning and memory. The presynaptic calcium sensor synaptotagmin 7 (syt7) regulates aspects of short-term plasticity (STP), but the underlying mechanisms remain unclear. Here, we show that alternative splicing of the syt7 juxtamembrane linker acts as a molecular switch at both biochemical and functional levels. The α and β variants undergo liquid-liquid phase separation to form condensates, while the γ variant forms aggregates. Using iGluSnFR imaging, we found that when expressed at equal levels, these three isoforms also diverge regarding their abilities to regulate two key aspects of STP: paired-pulse facilitation and synaptic depression. STED microscopy showed that all three isoforms form active zone-associated clusters that colocalize with syt1, while MINFLUX super-resolution microscopy resolved syt7 clusters within the active zone, well-positioned to directly control synaptic vesicle dynamics. Thus, alternative splicing might fine-tune STP by differentially impacting syt7 oligomerization.
Dataset DOI: 10.5061/dryad.4qrfj6qpw
Code repository: https://github.com/nikunjrmehta/MINFLUX-Data-Analysis
General Information
Authors: Mehta, Nikunj; Larson, Devin T.; Wozney, Mitch; Mishra, Shweta; Subramani, Smrithika; Kaur, Simi; Jain, Avani; Chapman, Edwin R.
Author affiliations:
Department of Neuroscience, University of Wisconsin-Madison
Howard Hughes Medical Institute
Funding: Howard Hughes Medical Institute; National Institute of Mental Health (R01MH061876); National Institute of Neurological Disorders and Stroke (R35NS136306)
License: This dataset is released under a CC0 (public domain) waiver.
Description of the data and file structure
This dataset contains the raw data and code required to replicate analyses in Mehta et al. The data include fluorescence recovery after photobleaching (FRAP) experiments in vitro and in cells, iGluSnFR imaging, MINFLUX imaging as the main dataset, and an additional dataset for supplementary information in the paper.
Files and variables
The dataset is organized by the main and supplementary figures of the associated manuscript (main Figures 1-6; Supplementary Figures 1-20). To analyze the dataset, all the relevant codes and scripts are provided in the GitHub repository linked in the Related Works section. Files with common descriptions, structure, or methods are mentioned below:
File: Fig._1.zip
Description: Fig. 1 contains grouped data for in vitro FRAP experiments for the three splice variants of synaptotagmin 7 (syt7): alpha, beta, and gamma, C-terminally tagged with mRuby, in a cell-free reconstitution (25 mM Tris pH 7.4, 100 mM NaCl, plus PEG 8000 as a molecular crowding agent). FRAP was performed on a Zeiss 880 Airyscan LSM (63X/1.4 NA oil, Fast Airyscan mode). Circular ROIs of 0.8-2 µm diameter were bleached within droplets/aggregates/clusters (2-2.5 µm diameter) using 405 nm and 488 nm laser lines (25% and 100% power, respectively). Imaging was performed at 15 frames/min, with 20 s pre-bleach, a 500 ms bleach event, and 400 s of recovery monitored per trace. All images were Airyscan-deconvolved; recovery curves were fitted with a hyperbolic function to extract % recovery and t1/2. HaloTag-fusion constructs (used for full-length syt7 in cells) were labeled with JF549 dye.
Data type: FRAP time series (pre-bleach, bleach, recovery) per the common FRAP parameters. Format: .czi (Zeiss Airyscan confocal). Opens in Zeiss ZEN or the open-access FIJI/ImageJ (Bio-Formats plugin).
File: Fig._2.zip
Description: Cellular FRAP of full-length (fl), HaloTag-tagged syt7 variants at the plasma membrane, in two cell types: HEK293T cells (Supplementary Fig. 6) and rat hippocampal neurons, imaged separately at interbouton segments and within boutons (main Fig. 2; beta and gamma neuronal data specifically in Supplementary Fig. 7). Each combination includes a control and a 10% 1,6-HD-treated condition. 1,6-hexanediol (Sigma, 88571) is an aliphatic alcohol that acutely disrupts the weak hydrophobic interactions holding together LLPS condensates, without dissolving solid aggregates. Per the Methods, HEK293T cells and cultured hippocampal neurons were gently treated with 10% (v/v) 1,6-HD for 10 minutes; FRAP was then repeated in the same manner as the paired control condition (untreated). Condensates formed by the alpha and beta variants dissolved upon 1,6-HD treatment (consistent with an LLPS origin), whereas gamma-variant aggregates were unaffected (consistent with a non-liquid, aggregated state).
Data type: FRAP time series per cell/compartment/condition (common FRAP parameters above). Format: .czi (Zeiss Airyscan confocal); ZEN or FIJI.
File: Fig._3_and_4_Syt7_WT.zip
File: Fig._3_and_4_Syt7_KO.zip
File: Fig._3_and_4_Syt7_Alpha.zip
File: Fig._3_and_4_Syt7_Beta.zip
File: Fig._3_and_4_Syt7_Gamma.zip
Description: iGluSnFR (glutamate sensor) imaging in wild-type neurons, syt7KO neurons, and syt7KO neurons re-expressing one splice variant at a level matched to WT PPF (alpha) or to the alpha expression level (beta, gamma; expression levels validated by the Western blots in Supplementary Fig. 9). Each variant/genotype folder contains both the paired-pulse protocol (main Fig. 3; ISI dependence in Supplementary Fig. 11; per-AP amplitude in Supplementary Fig. 10; PPR restricted to dual-responding ROIs in Supplementary Fig. 12) and the depression/HFS protocol (main Fig. 4; WT-vs-KO comparison replotted in Supplementary Fig. 14). Imaging used an Olympus IX83 inverted microscope (X-Cite 120 LED, ORCA-Fusion CMOS camera), 33-34°C, neurons at 14-15 DIV, in standard extracellular imaging solution with D-AP5, CNQX, and picrotoxin to block recurrent network activity. Two stimulation protocols were used: 1) Paired-pulse facilitation: two action potentials (APs) at one of five inter-stimulus intervals (50, 100, 200, 500, or 1000 ms, corresponding to 20, 10, 5, 2, and 1 Hz), imaged at 100 Hz (10 ms exposure, 2x2 binning); 150 frames were collected for the 20-2 Hz conditions and 200 frames for the 1 Hz condition. 2) Synaptic depression: high-frequency stimulation (HFS) of 50 APs at 20 Hz (2.5 s total), imaged as 350 frames at 10 ms exposure (2x2 binning), covering 3.5 s total.
Data type: Wide-field time-lapse image series (not static frames) per the common iGluSnFR parameters above: 150-200 frames for paired-pulse recordings, 350 frames for depression/HFS recordings. Format: .tiff, wide-field fluorescence microscopy; a custom FIJI/ImageJ plugin and downstream analysis scripts (ROI detection, PPR and SVRR calculation) are provided in the GitHub repository linked in the Related Works section.
File: Fig._5.zip
Description: 3D STED super-resolution imaging of syt7KO hippocampal neurons overexpressing FLAG-tagged alpha, beta, or gamma syt7, immunostained for syt7 (anti-FLAG), syt1 (mAb48), and confocal-imaged for bassoon (active-zone marker). Used to test whether the three variants differ in active-zone clustering or colocalization with syt1; they did not (similar cluster counts, colocalization, and bassoon-positive occupancy across all three, ~64-71%).
Data type: 3D (z-stack), 2-channel STED plus 1-channel confocal acquisitions per field of view; figures shown are maximum Z-projections, with additional zoomed-in en face single-synapse views. Format: .obf (Abberior). Opens in Abberior Lightbox or FIJI (with the appropriate reader).
File: Fig._6.zip
Description: Two-color 3D MINFLUX single-molecule localization imaging of alpha-syt7 (anti-FLAG, FX680) and syt1 (mAb48, FX640) at active zones, paired with confocal imaging of bassoon to define the active-zone centroid. Localization precision achieved was 5.39 ± 0.26 nm (X), 5.14 ± 0.26 nm (Y), 3.01 ± 0.22 nm (Z) (~21 nm practical resolution once ~20 nm antibody linkage error is included). Used to localize syt7 relative to the active-zone boundary and to syt1 clusters.
Data type: Single-molecule localization ("point-cloud") data: individual fluorophore localizations grouped by Trace ID (TID), with X/Y/Z coordinates, not conventional pixel images. Raw .msr files were exported to .npy for downstream Python analysis (filtering by EFO/CFR/localizations-per-TID, DBSCAN clustering, nearest-neighbor-distance analysis); analysis code is in the repository. Format: .msr (Abberior). Opens in Abberior Inspector or FIJI (with the appropriate reader).
File: Fig._S5.zip
Description: Alpha-syt7cyto-mRuby assembled on the surface of a giant unilamellar vesicle (GUV; 93.5% DOPC, 5% 18:1 DGS-NTA, 1.5% Atto647N DOPE, labeled with Atto-647), used to visualize 2D LLPS directly on a membrane rather than in free solution.
Data type: A confocal z-stack through the GUV (used to generate the 3D rendering shown in the figure) plus a representative single Z-section plane; not a FRAP time series or a 2D single-plane-only acquisition. Format: .czi (Zeiss Airyscan confocal); ZEN or FIJI.
File: Fig._S6.zip
Description: See description of Fig._2.zip. Cellular FRAP of full-length (fl), HaloTag-tagged syt7 variants at the plasma membrane, in HEK293T cells.
Data type: FRAP time series per cell/compartment/condition (common FRAP parameters above). Format: .czi (Zeiss Airyscan confocal); ZEN or FIJI.
File: Fig._S8.zip
Description: FRAP of alpha-syt7-fl-HaloTag in three subcellular compartments of hippocampal neurons (soma, dendrite, axon initial segment/AIS), control vs. 10% 1,6-HD. Recovery was low (<15%) at all three locations regardless of treatment, indicating that mistargeted somatodendritic syt7 is relatively immobile. The AIS was identified using a live-cell surface stain against pan-neurofascin (extracellular labeling, AF647 secondary) prior to bleaching.
Data type: FRAP time series per compartment/condition (common FRAP parameters above), plus static two-channel (neurofascin/HaloTag) marker images used to locate the AIS before bleaching. Format: .czi (Zeiss Airyscan confocal); ZEN or FIJI.
File: Fig._S13.zip
Description: Airyscan immunocytochemistry of syt7KO neurons expressing alpha, beta, or gamma syt7-fl, fixed in methanol and stained with a vGlut antibody (presynaptic terminal marker). Puncta were counted from 30 µm axonal segments (10 FOVs/condition, 2 culture preparations) to confirm that presynaptic terminal density did not differ across variants, ruling this out as a confound for the functional (iGluSnFR) differences seen in Figs. 3-4.
Data type: static two-channel (variant marker + vGlut) images. Format: .czi (Zeiss Airyscan confocal); ZEN or FIJI. Analysis code for puncta counting is provided in the GitHub repository linked in the Related Works section.
File: Fig._S16.zip
Description: MINFLUX imaging control. Non-transduced syt7KO neurons (no FLAG-tagged syt7 present) were stained with the same anti-FLAG antibody used in Fig_6.zip, to measure non-specific/background labeling; little signal was observed, while endogenous syt1 and bassoon staining were normal.
Data type: Same type of data as Fig_6.zip: 2-color 3D MINFLUX point-cloud localizations (anti-FLAG, syt1) plus a confocal bassoon channel.Format: .msr (Abberior). Opens in Abberior Inspector or FIJI (with the appropriate reader).
Code/software
Analysis code and scripts used to generate the figures in this manuscript are maintained at: https://github.com/nikunjrmehta/MINFLUX-Data-Analysis