Overlapping functions of stonin 2 and SV2 in sorting of the calcium sensor synaptotagmin 1 to synaptic vesicles

Abstract

Neurotransmission involves the calcium-regulated exocytic fusion of synaptic vesicles (SVs) and the subsequent retrieval of SV membranes followed by reformation of properly sized and shaped SVs. An unresolved question is whether each SV protein is sorted by its own dedicated adaptor or whether sorting is facilitated by association between different SV proteins. We demonstrate that endocytic sorting of the calcium sensor synaptotagmin 1 (Syt1) is mediated by the overlapping activities of the Syt1-associated SV glycoprotein SV2A/B and the endocytic Syt1-adaptor stonin 2 (Stn2). Deletion or knockdown of either SV2A/B or Stn2 results in partial Syt1 loss and missorting of Syt1 to the neuronal surface, whereas deletion of both SV2A/B and Stn2 dramatically exacerbates this phenotype. Selective missorting and degradation of Syt1 in the absence of SV2A/B and Stn2 impairs the efficacy of neurotransmission at hippocampal synapses. These results indicate that endocytic sorting of Syt1 to SVs is mediated by the overlapping activities of SV2A/B and Stn2 and favor a model according to which SV protein sorting is guarded by both cargo-specific mechanisms as well as association between SV proteins.

Keywords: calcium sensor; endocytosis; knockout mice; neurotransmission; synaptic vesicle protein sorting.

Fig. 1.

Exacerbated loss of Syt1 upon combined deletion of Stn2 and SV2A/B. (A) Levels of synaptic and endocytic proteins in total brain lysates from p12 control, SV2A/B DKO, and SV2A/B/Stn2 TKO mice probed by immunoblotting with specific antibodies. AP-2 α, adaptor protein 2α; CHC, clathrin heavy chain; Hsc70, heat shock cognate protein of 70 kDa; pan-Dyn, dynamin 1–3; Syp, synaptophysin; Syt1, synaptotagmin 1. The band remaining in DKO and TKO lysates decorated with pan-SV2 antibodies corresponds to the SV2C isoform. (B) Stn2 is up-regulated in the absence of SV2A/B (control, 100.0 ± 6.9%; DKO, 131.8 ± 7.2%; TKO, 4.7 ± 1.5%; n control/DKO = 6; n control/TKO = 5; **P < 0.01). (C) Syt1 levels are significantly reduced in DKO (42.3 ± 3.2%) and in TKO (29.9 ± 2.6%) compared with control (100.0 ± 6.0%; n control/DKO = 6; n control/TKO = 5; *P < 0.05, **P < 0.01, ***P < 0.001), and Syp levels are unaffected (control, 100.0 ± 7.6%; DKO, 105.3 ± 6.8%; TKO, 92.3 ± 10.21%; n control/DKO = 6; n control/TKO = 5). (D) Levels of clathrin (CHC), AP-2α, and dynamin 1–3 (Dyn) in total brain lysates derived from p12 control, DKO, and TKO mice (CHC—control, 100.0 ± 5.2%; DKO, 111.2 ± 14.0%; TKO, 106.9 ± 4.4%; AP2—control, 100.0 ± 7.1%; DKO, 101.5 ± 8.7%; TKO, 110.9 ± 9.3%; Dyn—control, 100.0 ± 7.7%; DKO, 81.5 ± 10.1%; TKO, 84.9 ± 6.5%; n control/DKO = 6; n control/TKO = 5). (E–J) Total levels of Syt1 (E and F), Syn1 (F and H), and Syb2 (I and J) in the hippocampus of control, DKO, and TKO mice revealed by immunostaining. These are representative confocal images of perfused sagittal brain sections. (Scale bar, 100 µm.) (F) Syt1 intensity levels are reduced in DKO and even further decreased in TKO mice compared with controls (control, 100.0 ± 5.6%; DKO, 71.4 ± 3.5%; TKO, 41.7 ± 3.0%; n control/DKO = 4; n control/TKO = 3, *P < 0.05, **P < 0.01). (H and J) Syn1 and Syb2 expression levels are not significantly different in DKO and TKO mice compared with controls (control Syn1, 100.0 ± 9.6%; DKO Syn1, 103.4 ± 18.1%; TKO Syn1, 124.5 ± 6.3%; control Syb2, 100,0 ± 19.0%; DKO Syb2, 95.7 ± 3.1%; TKO Syb2, 99.0 ± 8.4%; n WT/DKO = 4; n WT/TKO = 3). Regions of interest (ROI) in E, G, and I indicate areas taken for quantification. All data represent mean ± SEM.

Fig. S1.

Generation and characterization of SV2A/B DKO and SV2A/B/Stn2 TKO mice. (A) Scheme illustrating the generation of DKO and TKO animals as well as their corresponding controls. SV2A HET/SV2B KO mice (24) (HET/KO/WT) were bred with Stn2 KO mice (14) (WT/WT/KO) to generate triple heterozygous mice (HET/HET/HET), which were subsequently crossed to obtain mice lacking SV2B but heterozygous for SV2A and Stn2 (HET/KO/HET). These were crossed to generate DKO (KO/KO/WT), TKO (KO/KO/KO), as well as control mice (WT/KO/HET and WT/KO/WT) used in the current study. (B) Representative images of DKO and TKO animals and their corresponding littermates at p12. (C) Growth charts of DKO and TKO mice. Body weights of DKO and TKO do not differ significantly at any time point, whereas they are significantly reduced compared with controls (n control = 63, n DKO = 21, n TKO = 25; **P < 0.01, ***P < 0.001). (D) Electron micrographs of synapses from the proximal CA1 stratum radiatum area of control, DKO, and TKO mice (>70 synapses per genotype). (Scale bar, 1 µm.)

Fig. 2.

Combined deficiency of SV2 and Stn2 causes additive defects in Syt1 sorting. (A and B) Time course (A) and endocytic time constants (τ) (B) of Syt1-pHluorin endocytosis/reacidification in WT and Stn2 KO neurons coexpressing Syt1-pHluorin and scr or SV2A siRNA (KD) in response to 200 APs (40 Hz). Endocytosis is facilitated in the absence of either SV2A τ_WT,scr = 43.1 ± 3.6 s (τ_{WT,SV2A KD} = 34.5 ± 1.8 s) or Stn2 (τ_KO,scr = 35.1 ± 1.2 s), an effect further aggravated by depletion of both proteins (τ_{KO,SV2A KD} = 25.3 ± 1.0 s; *P < 0.05, ***P < 0.0001; n = 3; >950 boutons per condition). Data represent mean ± SEM. (C) Surface/total ratios of Syt1-pHluorin expressed in hippocampal wild-type (WT) or Stn2 KO neurons cotransfected with scr or SV2A siRNA (KD). The surface/total Syt1-pHluorin ratio was significantly increased in the absence of either SV2A or Stn2 (Stn2 WT, scr, 12.2 ± 0.1%; Stn2 WT, SV2A KD, 16.6 ± 0.1%; Stn2 KO, scr, 16.8 ± 0.1%) and further aggravated by depletion of both proteins (Stn2 KO, SV2A KD, 21.6 ± 0.1%; n = 5; >510 boutons per condition; *P < 0.05, ***P < 0.0001). (D) Total (red) and surface levels (green) of Syt1 in the hippocampus of p12 control, SV2A/B DKO, and SV2A/B/Stn2 TKO mice (green). (Scale bar, 100 µm.) ROIs in D indicate areas taken for quantification. (E) Elevated Syt1 surface levels in SV2A/B DKO and more pronouncedly in SV2A/B/Stn2 TKO mice (DKO, 129.6 ± 9.4%; TKO, 201.8 ± 22.2%) compared with controls (control, 100 ± 8.4%; n control/DKO = 4; n control/TKO = 3; *P < 0.05). Data represent mean ± SEM.

Fig. S2.

Impaired Syt1 kinetics and sorting in the absence of SV2A and Stn2. (A) Efficacy of SV2A KD in control hippocampal neurons. Representative confocal images of neurons coexpressing Syt1-pHluorin with scr siRNA (Upper panel) or SV2A siRNA (Lower panel) and immunostained with GFP (green) and SV2A (red) antibodies. (Scale bar, 10 µm.) (B) SV2A levels are significantly decreased in neurons treated with SV2A siRNA (KD), compared with the neurons treated with scr siRNA (scr, 100 ± 14.1%; SV2AKD, 36.1 ± 4.4%; n = 4; ***P < 0.0001). (C) Immunoblot of lysates from HEK293T cells overexpressing eGFP-SV2A and treated with scr or SV2A siRNA probed with SV2A antibody. (D) SV2A KD significantly reduces Syt1 total levels compared with control (scr, 100 ± 5.8%; SV2AKD, 79.1 ± 0.59%; *P < 0.05; quantification of 2A) and is rescued by overexpression of siRNA-resistant SV2A-GFP (scr, 104.7 ± 3.4%; SV2AKD, 110.7 ± 4.8%; **P < 0.01). (E) Elevated ratio of surface/total Syt1 in SV2A KD neurons compared with scr siRNA-treated controls (scr, 1.0 ± 0.14; SV2A KD, 1.65 ± 0.13; ***P < 0.001) is rescued by overexpression of siRNA-resistant SV2A-GFP (scr, 1.28 ± 0.06; SV2AKD, 1.34 ± 0.06; **P < 0.01; n = 3; >3,000 boutons per condition). (F) Accelerated retrieval of Syp-pHluorin in SV2A-deficient neurons stimulated with 200 APs at 40 Hz. (Inset) Endocytic time constants obtained by monoexponential fit [f(x) = A₁e^−x/t1 + y₀] (τWT scr = 39.8 ± 2.4 s; τWT SV2A KD =29.9 ± 1.4 s; **P < 0.001; mean ± SEM of n = 4 independent experiments; >1,900 boutons per genotype). (G) Surface accumulation of Syt1 in SV2A-deficient Stn2 KO neurons. Representative confocal images of Stn2 KO neurons coexpressing Syp-pHluorin (SpH, green) and either scr or SV2A siRNA (KD) and immunostained for total Syt1 (red) and surface Syt1 (Syt1 LD, blue). (Scale bar, 5 µm.) (H) Elevated ratio of surface/total Syt1 in Stn2 KO neurons depleted of SV2A (scr, 1.0 ± 0.1; SV2A KD, 1.8 ± 0.1; >1,000 boutons per condition; **P < 0.001).(I) Total Syt1 levels are reduced in SV2A-deficient Stn2 KO neurons (SV2AKD, 67 ± 5.6%), compared with control Stn2 KO neurons (scr, 100 ± 15.5%; *P < 0.05). (J) NH₄Cl pulse in SV2A-deficient neurons overexpressing Syt1-pHluorin reveals similar Syt1-pHluorin expression levels compared with scr siRNA-transfected neurons (scr, 21.9 ± 2.2 AU; SV2AKD, 22.0 ± 2.3 AU; n = 5; >510 boutons per condition). Data represent mean ± SEM.

Fig. 3.

SV2 regulates Syt1 sorting to SVs during neuronal activity. (A and B) Surface accumulation of Syt1 in SV2A-deficient hippocampal neurons. (A) Representative confocal images of mouse hippocampal neurons coexpressing Syp-pHluorin (SpH, green) with scr or SV2A siRNA (SV2A KD) and immunostained for total Syt1 (red) and surface Syt1 (Syt1 LD, blue). (Scale bar, 5 µm.) (B) Elevated normalized ratio of surface/total Syt1 in neurons treated with SV2A siRNA compared with scr siRNA-treated controls (set to 1; scr, 1.0 ± 0.08; SV2A KD, 1.62 ± 0.13; n = 3; >1,800 boutons per condition; **P < 0.001). Syp surface/total ratio is unaffected (scr, 1.0 ± 0.08; SV2A KD, 1.03 ± 0.09; n = 3; >1,700 boutons per condition). Surface accumulation of Syt1 is rescued in TTX-silenced neurons (scr, 1.0 ± 0.5; SV2AKD, 0.98 ± 0.06; n = 2; >4,500 boutons per condition). (C and D) Loss of SV2A selectively impairs Syt1-pHluorin (Syt1pH) sorting to SVs during repetitive rounds of exo-/endocytosis. Average traces in response to 200 APs (10 Hz) in the absence or presence of folimycin. Neurons were cotransfected with Syt1-pH and scr (C) or SV2A siRNA (KD, D). (E and F) Depletion of SV2A does not affect Syp-pHluorin (SypH) sorting fidelity during stimulation. Average traces of hippocampal neurons cotransfected with SypH and scr (E) or SV2A siRNA (F) in response to 200 APs (10 Hz) in the absence or presence of folimycin. (G) Reduced apparent number of exocytosed Syt1-pHluorin molecules in SV2A-deficient neurons (SV2AKD, 1.13 ± 0.10) compared with control (scr, 1.73 ± 0.15; **P < 0.001; n = 13; >630 boutons per genotype). The apparent number of exocytosed SypH molecules (H) is unaffected (scr, 3.0 ± 0.6; SV2AKD, 3.1 ± 0.4; n = 8; >300 boutons per genotype). Data represent mean ± SEM.

Fig. 4.

Combined deficiency of SV2 and Stn2 aggravates impairments in synaptic strength and short-term plasticity caused by deletion of SV2. (A and B) Reduced basal transmission at excitatory Schaffer collateral to CA1 pyramidal cell synapses in SV2A/B DKO (red trace) and SV2A/B/Stn2 TKO (blue trace) mice in response to increasing stimulus intensities. (A) fEPSP amplitudes are significantly reduced in DKO compared with control (***P < 0.001). Inset illustrates sample traces of maximal fEPSPs in DKO (red) and control (black). (B) fEPSP amplitudes in TKO are significantly lower compared with DKO (***P < 0.001). Inset illustrates sample traces of maximal fEPSPs in TKO (blue) and in DKO (red). (C–E) Altered paired-pulse facilitation (PPF) at excitatory Schaffer collateral to CA1 pyramidal cell synapses in DKO and TKO mice. (C) PPF measured at 10, 20, 50, 100, 200, and 500 ms interpulse intervals (ISIs) reveals increased facilitation in DKO compared with control (***P < 0.001). Inset depicts sample traces of fEPSPs at 50 ms ISI normalized to the first response of DKO (red) and control (black) synapses. (D) Significantly elevated PPF (fEPSP2/fEPSP1) at 10, 20, 50, 100, 200, and 500 ms ISI at synapses of TKO, compared with DKO (**P = 0.004). Inset illustrates sample traces of fEPSPs at 50 ms ISI normalized to the first response at DKO (red) and TKO (blue) synapses. (E) PPF at 50 ms ISI is significantly increased in DKO compared with control (***P < 0.001) and is further increased in TKO (***P < 0.001). Note that maximal facilitation of the second response is detected at 50 ms in control slices (control n = 20, n = 8; DKO n = 9, n = 4; TKO n = 14, n = 5). (F) Normal steady-state level of neurotransmission in DKO and TKO mice probed by high-frequency stimulation (20 Hz). Each value is an average of three consecutive responses except for n = 1. Only the first 200 out of 500 pulses are presented.

Fig. S3.

Increased facilitation in DKO and TKO is not caused by changes in presynaptic excitability, miniature EPSCs, or releasable SV pool size. (A) A representative trace of a maximal fEPSP recorded in CA1 stratum radiatum of a p12 mouse, illustrating the time points taken for measurements of presynaptic fiber volleys (black arrow) and fEPSP amplitudes (gray arrow). (B) Presynaptic excitability of controls, DKO, and TKO, measured as a relationship between the fiber volley amplitudes and stimulation intensities. Synapses of DKO (P = 0.7) and TKO (P = 0.6) mice show no significant differences in presynaptic excitability, compared with the control. (C–E) Unaltered action potential-independent neurotransmission in TKO mice. (C) Representative traces of mEPSCs recorded from CA1 neurons of control and TKO mice. (D) Frequency of miniature excitatory postsynaptic events is not changed in TKO compared with controls (P = 0.8). (E) Average amplitudes of mEPSCs are not affected in TKO mice (P = 0.7) compared with controls (control n = 4, n = 3; TKO n = 5, n = 3). (F and G) Graphs illustrating comparable basal neurotransmission (F) and short-term plasticity (G) at the synapses of two controls used in the current study (SV2A WT/SV2B KO/Stn2 WT:WT/KO/WT and SV2A WT/SV2B KO/Stn2 Het:WT/KO/HET). Relationships between fEPSPs and fiber volley amplitudes (F) or ISI and PPF are plotted (G). (H) Cumulative amplitudes in response to 500 pulses at 20 Hz show no differences in DKO (P = 0.96) and TKO (P = 0.50) compared with control synapses (control n = 19, n = 8; DKO n = 9, n = 4; TKO n = 14, n = 5). (I) Facilitation and subsequent depression of initially smaller fEPSPs in DKO and TKO synapses in response to a train of 500 action potentials applied at 20 Hz. The maximal facilitation peak (n of stimuli = 8–10) normalized to the control in DKO synapses reveals a 60% increase in facilitation (P < 0.001), whereas TKO synapses facilitate stronger than DKO (P = 0.011) as far as 200% compared with the facilitation peak in the control mice. After initial facilitation, amplitudes of fEPSPs subsequently depress, reaching a level that is still significantly higher in DKO and in TKO, compared with control. First pulse was taken as 100%, and each value except for n = 1 is an average of three consecutive responses. Only the first 200 out of 500 pulses are presented for simplicity (control n = 19, n = 8; DKO n = 9, n = 4; TKO n = 14, n = 5).

Fig. S4.

Schematic depicting Syt1 loss and missorting in the absence of Stn2, SV2, or both proteins. There are slightly reduced levels and partial missorting of Syt1 to the neuronal surface in Stn2 KO mice (2). SV2A/B DKO synapses show strongly reduced Syt1 levels and Syt1 missorting to the neuronal surface, resulting in its depletion from SVs. Loss of total Syt1 and depletion from SVs are further aggravated by deletion of both SV2A/B and Stn2 in TKO synapses.