Research ArticleCELLULAR NEUROSCIENCE

Control of synaptic vesicle release probability via VAMP4 targeting to endolysosomes

See allHide authors and affiliations

Science Advances  30 Apr 2021:
Vol. 7, no. 18, eabf3873
DOI: 10.1126/sciadv.abf3873
  • Fig. 1 Synaptic accumulation of VAMP4 lowers SV fusion competence.

    Primary cultures of hippocampal neurons were cotransfected with syp-mOr2 and either VAMP4-pHluorin or syb2-pHluorin. Neurons were stimulated with 400 APs (40 Hz) and pulsed with NH4Cl imaging buffer after 200 s. (A and B) Representative images of the fluorescent response of syp-mOr2 (A and B), VAMP4-pHluorin (A), and syb2-pHluorin (B) are displayed at rest, during stimulation, or during exposure to NH4Cl [white arrows, nerve terminals that display a syp-mOr2 response (active synapses); yellow arrows, those that do not (silent synapses)]. Scale bar, 5 μm. (C and D) Hippocampal neurons transduced with PSD95-EGFP after a prior transfection with syp-mOr2 were subjected to an identical protocol as above. (C) Representative images of syp-mOr2 and PSD95-EGFP at rest, during stimulation, or during NH4Cl exposure (white arrows, active synapses; yellow arrows, silent synapses). Scale bar, 5 μm. (D) Number of active or silent syp-mOr2 puncta colocalizing with PSD95-EGFP (n = 13 coverslips from four independent preparations, P = 0.82, Mann-Whitney test). (E and F) Number of synapses with fusion events reported by pHluorin (E) or the expression level of pHluorin at synapses that show an activity-dependent syp-mOr2 response (F). n = 15 coverslips, each from four independent preparations, ****P < 0.001 and ***P = 0.001, Mann-Whitney test. (G to I) Expression level of syb2-pHluorin (G), sypHy (H), and VAMP4-pHluorin (I) at active and silent synapses. n = 10 (G) or n = 14 (H and I) coverslips from four independent preparations, ***P = 0.0005, **P = 0.002, and *P = 0.039, either Wilcoxon matched-pairs signed-rank test (G and I) or paired t test (H). (J to L) Correlation between the extent of the sypHy response with the expression of sypHy (J) or the syp-mOr2 response with the expression of syb2-pHluorin (K) and VAMP4-pHluorin (L). ns, not significant.

  • Fig. 2 VAMP4 does not form SNARE complexes efficiently during synaptic activity.

    Primary cultures of hippocampal neurons were cotransfected with YFP-syntaxin1a and either mCer-syb2 or mCer-VAMP4 and stimulated with 800 APs (40 Hz) while monitoring the fluorescence lifetime (τ) of the mCer donor. (A, B, E, and F) Representative images of the fluorescence intensity (A and E) or τ [using pseudo-color to report τ (B and F)] of either mCer-syb2 (A and B) or mCer-VAMP4 (E and F) before (control) or during (40-Hz, 20-s) stimulation. Scale bars, 10 μm. (C, D, G, and H) Normalized fluorescence decay (C and G) and τ (D and H) of either mCer-syb2 (C and D) or mCer-VAMP4 (G and H) from the experiments in (B) and (F), respectively. (I) ΔFRET indicates the average change in mCer-syb2 or mCer-VAMP4 τ before and during stimulation. (J) Average τ is displayed for all conditions. n = 301 nerve terminals from 7 independent experiments (syb2) and n = 256 nerve terminals from 17 independent experiments (VAMP4), Mann-Whitney test (I) and two-way analysis of variance (ANOVA) with Dunn’s multiple comparison test (J), ****P < 0.001. IRF, instrument response function.

  • Fig. 3 VAMP4 is retrieved constitutively from axons through the endolysosomal system.

    (A) Representative kymographs display the axonal mobility and directionality of traffic of VAMP4-EGFP and syb2-EGFP. (B and C) Frequency of anterograde, retrograde, or stationary trajectories (B) or the run length of retrograde trajectories (C) for both syb2-EGFP and VAMP4-EGFP. n = 25 (syb2) or n = 16 coverslips (VAMP4) from four independent preparations, ****P < 0.001 and **P = 0.0002, two-way ANOVA with Fisher’s least significant difference (B), and **P = 0.0012, unpaired t test (C). (D and E) Hippocampal neurons were transfected with syb2, synaptophysin, or VAMP4 tagged with an FT protein. (D) Representative images display FT protein fluorescence in either the blue (new protein), red (old protein), or merged channel. Scale bar, 5 μm. (E) Blue/red ratio of FT proteins. n = 24 (syb2 and synaptophysin) or n = 23 cells (VAMP4) from four independent preparations, ****P < 0.001 and ***P = 0.0004, Kruskal-Wallis test with Dunn’s multiple comparisons test. (F to H) Hippocampal neurons were transfected with mCherry-rab7 and either syb2-EGFP or VAMP4-EGFP. (F and G) Kymographs show the retrograde cotrafficking of syb2-EGFP and VAMP4-EGFP with mCherry-rab7. (H) Fraction of retrograde trajectories where syb2-EGFP or VAMP4-EGFP cotraffic with mCherry-rab7. n = 16 (syb2) or n = 18 (VAMP4) coverslips from four independent preparations, ****P < 0.001, Mann-Whitney test.

  • Fig. 4 Endolysosomal sorting of VAMP4 regulates its abundance in the SV pool.

    (A and B) Hippocampal neurons cotransfected with VAMP4-pHluorin and either wild-type (WT), constitutively active (Q67L), or dominant-negative (T22N) mCherry-rab7. Scale bar, 5 μm. Representative images (A) and bar graph (B) display synaptic VAMP4-pHluorin expression. n = 23 (mCherry), n = 12 (mCherry-rab7), n = 13 (Q67L), and n = 18 (T22N) coverslips from four independent preparations, ****P < 0.001, Kruskal-Wallis test with Dunn’s multiple comparisons test. (C) Effect of T22N mCherry-rab7 overexpression on sypHy synaptic expression. n = 21 (mCherry) or n = 16 (T22N mCherry-rab7) coverslips from four independent preparations, ***P < 0.0002, Mann-Whitney test. (D) Fusion events reported by VAMP4-pHluorin upon stimulation with 400 APs at 40 Hz under the conditions shown in (A) and (B), n = 23 (mCherry), n = 12 (mCherry-rab7), n = 13 (Q67L), and n = 18 (T22N) coverslips from four independent preparations, *P = 0.0129, Kruskal-Wallis test with Dunn’s multiple comparisons test.

  • Fig. 5 AP1-mediated sorting of VAMP4 to endolysosomes regulates its abundance in the SV pool.

    (A, C, and E) Representative images display hippocampal neurons transfected with VAMP4-EGFP and mCherry, α-synuclein A53T, or red fluorescent protein (RFP)–bassoon (Bsn) (A), VAMP4-EGFP–transfected neurons with and without 30 μM SMIFH2 (C) and VAMP4-EGFP cotransfected with either AP1 shRNA (AP1 KD) or a scrambled (Scr) control (E). Scale bars, 5 μm. (B, D, and F) Synaptic VAMP4 expression under all conditions described in (A), (C) and (E). (B) n = 40 (mCherry), n = 51 (α-synuclein A53T), and n = 53 (bassoon) cells from three independent preparations, ****P < 0.001 and **P = 0.0038, Kruskal-Wallis test with Dunn’s multiple comparisons test (P values adjusted). (D) n = 47 (control) and n = 61 (SMIFH) cells from three independent preparations (F), n = 18 coverslips (scrambled and AP1 KD) from four independent preparations, ***P = 0.0009 and **P = 0.0085, Mann-Whitney test. (G and H) Hippocampal neurons were transfected with either wild-type or VAMP4 L25A–FT. (G) Representative images display fluorescence in the blue (new protein), red (old protein), or merged channel. Scale bar, 5 μm. (H) Blue/red ratio [n = 30 (wild-type) and n = 39 (L25A) cells from four independent preparations, ****P < 0.001, Mann-Whitney test]. (I to K) Hippocampal neurons cotransfected with VAMP4-pHluorin and or shRNAs against either AP1 or CHC (or scrambled). Representative images (I) (scale bar, 5 μm) and quantification of VAMP4-pHluorin expression (J) in either scrambled or CHC KD synapses [n = 12 (scrambled), n = 14 (CHC KD) coverslips from three independent preparations, *P = 0.011, Mann-Whitney test]. (K) VAMP4-pHluorin fusion events [n = 12 (scrambled), n = 10 (AP1 KD), and n = 14 (CHC KD) coverslips from three independent preparations, ****P < 0.0001 and **P < 0.01, Kruskal-Wallis test with Dunn’s multiple comparisons test].

  • Fig. 6 VAMP4 KO synapses display a reduction in endolysosomal molecules and increased SV fusion.

    (A and B) Wild-type and VAMP4 KO brain lysates were immunoblotted for synaptic and endosomal proteins (all n = 3 animals, t test with Benjamini, Krieger, and Yekutieli correction for false discovery rate). MW, molecular weight. (C to F) MS analysis of synaptosomes from wild-type and VAMP4 KO brains (five animals for both wild-type and VAMP4 KO). (C) Unique proteins. (D) Volcano plot of significantly different proteins. Refer to data file S1 for full protein names in C and D. (E and F) GO cellular compartment analysis of increased/unique (E) and decreased/absent (F) proteins in VAMP4 KO synaptosomes. (G and H) Wild-type and VAMP4 KO cultures incubated with fluorescent-conjugated antibodies to synaptotagmin 1 (Syt1) were stimulated with 400 APs (40 Hz) and 50 μM TMR-dextran. (G) Representative images (Syt1, TMR-dextran, and merged). Scale bar, 5 μm. (H) TMR-dextran puncta per field of view normalized to Syt1 puncta. n = 17 coverslips from three independent preparations (both), *P = 0.0238, Mann-Whitney test. (I and J) Average time course of the sypHy response normalized to either the stimulation peak (I) or total SV pool [revealed by NH4Cl application (J)]. Stimulation indicated by bar. (K) Peak sypHy response as a fraction of the total SV pool. n = 11 coverslips from three independent preparations (both), *P = 0.043, unpaired t test.

  • Fig. 7 Increased Pr in hippocampal VAMP4 KO circuits.

    Neurotransmission at CA3-CA1 synapses was monitored using whole-cell patch-clamp recording in acute hippocampal slices from wild-type and VAMP4 KO mice. (A) Example mEPSC events. Frequency (B) and amplitude (C) of mEPSC events. n = 7 (wild-type) or n = 8 (KO) slices from four animals, Mann-Whitney test. (D) Paired-pulse ratio (PPR) of EPSCs as a function of the interstimulus interval (10 to 200 ms). n = 17 (wild-type) or n = 15 (KO) slices from either 10 or 9 animals, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, pulse number, two-way ANOVA with Fisher’s least significant difference, *P = 0.027, two-way ANOVA for genotype. (E) Evoked EPSC amplitude for slices stimulated with a 10-AP train (20 Hz, normalized to the first pulse). n = 10 (wild-type) or n = 9 (KO) slices from either seven or six animals, ****P < 0.0001, two-way ANOVA. (F and G) Slices were stimulated with 600 APs (40 Hz). Example EPSC amplitudes (F), average EPSC amplitude (normalized to peak response) (G). n = 10 (both) slices from either 6 or 8 animals, ****P < 0.0001, two-way ANOVA. (H) Pr of the CA3-CA1 synapse in wild-type and VAMP4 KO, calculated by dividing the amplitude of the first evoked EPSC by the effective RRP size. n = 10 (both) slices from either six or eight animals, **P = 0.039, Mann-Whitney test.

  • Fig. 8 VAMP4 conveys endolysosomal dysfunction into inhibition of SV fusion.

    (A to D) Primary cultures of hippocampal neurons transfected with sypHy or sypHy + VAMP4-FT (VAMP4 rescue) were incubated with 1 μM bafilomycin A1 and stimulated with two 20-Hz AP trains (2 and 45 s, indicated by bar), followed by NH4Cl solution. Average traces (A) and initial evoked sypHy response normalized to the total SV pool (B). Average traces (C) and sypHy response normalized to the initial response (Pr) (D). n = 14 (wild-type, KO/rescue) or n = 15 (KO) coverslips from eight independent preparations, (B) **P = 0.0015 and ***P = 0.0001, one-way ANOVA with Tukey’s multiple comparisons test, and (D) **P = 0.0077 and ****P < 0.0001, Kruskal-Wallis test with Dunn’s multiple comparisons test. (E to H) Identical experiments to those in (A) to (D) were performed in neurons expressing either mCherry-rab7T22N or treated with 30 μM SMIFH2. (E and F) Average traces. (G and H) Inhibition of the sypHy response by either mCherry-rab7T22N (G) or SMIFH2 (H) normalized to wild-type or VAMP4 KO controls. n = 16 (wild-type/T22N), n = 13 (KO/T22N), n = 22 (wild-type/SMIFH2), and n = 19 (KO/SMIFH2) coverslips from either four (G) or three (H) independent preparations, *P = 0.012 and ***P = 0.0003, unpaired t test.

  • Fig. 9 VAMP4 targeting to endolysosomes controls Pr.

    (A) Strong synaptic stimulation triggers ADBE, which recycles SV membranes and proteins, such as syb2 and VAMP4, following SV fusion. ADBE forms transient bulk endosomes from which vesicles bud to (i) refill the recycling SV pool (ii) or fuse with (mature to) endolysosomes for long-range trafficking and degradation. Proteins are sorted at bulk endosomes to either endolysosomes or SVs by different mechanisms. VAMP4 is predominantly sorted to endolysosomes via an AP1-dependent mechanism, whereas syb2 and, to a lesser extent, VAMP4 are sorted into SVs by different adaptor proteins. Both sorting pathways require clathrin for their function. (B) In wild-type nerve terminals, inhibition of either endolysosomal trafficking or AP1 function abolishes VAMP4 sorting to endolysosomes and results in its increased sorting in SVs. VAMP4 accumulation in the SV pool reduces SV fusion capacity and Pr, although increased numbers of VAMP4 molecules visit the surface during stimulation. (C) In VAMP4 KO terminals, loss of VAMP4 causes an increase in Pr under basal conditions due to loss of inhibitory control. In addition, blocking of endolysosomal trafficking and function does not affect Pr, highlighting the essential role of VAMP4 in converting endolysosomal dysfunction into changes of presynaptic release properties.

Supplementary Materials

  • Supplementary Materials

    Control of synaptic vesicle release probability via VAMP4 targeting to endolysosomes

    Daniela Ivanova, Katharine L. Dobson, Akshada Gajbhiye, Elizabeth C. Davenport, Daniela Hacker, Sila K. Ultanir, Matthias Trost, Michael A. Cousin

    Download Supplement

    The PDF file includes:

    • Figs. S1 to S7
    • Tables S1 and S2
    • Legend for data file S1
    • References

    Other Supplementary Material for this manuscript includes the following:

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article